Skip to main content

Full text of "Ontogeny and systematics of fishes : based on an international symposium dedicated to the memory of Elbert Halvor Ahlstrom"

See other formats 


f>]p  |::nrg|..:riPQ 

^l../- ^'.        J:.    .  .h./'il.  .:,..J;!,„<fh../ 


'Smn&d  on 

t.c>  tJae  M'enioiry  ol 

IBIbert  llilalvcr  A.jb,lstrO'B"i 


rbe  Sympo^wm  Vv^as  hdd  Awgust  IS-IS,  1983 


SporKKM'w]  by  tba 
JuitioDatl  jfi-iaiine  H;slKde!ii  Scrvia? 


ABiiiCi'itt-Hi',  Society  «.)';"  ]l':;:!i  i:hyoloi?j:^ti5  aad  HerfK'rtc^loBiBi'j 


nj 

_D 

■=0 

3 

r^ 

L 

1— 1 

D 

o 

m 

□ 

c,  / 


QNTOGENY  AND  SYSTEMATICS 
^  OF  FISHES 


Based  on 

An  International  Symposium  Dedicated 

to  the  Memory  of 

Elbert  Halvor  Ahlstrom 

The  Symposium  was  held  August  15-18,  1983,  La  JoUa,  CaUfomia 

Sponsored  by  the 

National  Marine  Fisheries  Service 

National  Oceanic  and  Atmospheric  Administration 

United  States  Department  of  Commerce 


\9 


■0 


4 

0 


5:^ 


Special  Publication  Number  1 
American  Society  of  Ichthyologists  and  Herpetologists 


Library  of  Congress  Catalogue  Card  Number:  84-72702 

ISSN  No.  0748-0539 

©  Copyright,  1984,  by  The  American  Society  of  Ichthyologists  and  Herpetologists 

Pnnted  by  Allen  Press  Inc..  Lawrence,  KS  66044  USA 


Preface 

The  National  Marine  Fisheries  Service  organized,  supported  and  conducted  an  international  symposium  entitled 
Ontogeny  and  Systematics  of  Fishes,  held  in  La  JoUa,  California  on  August  15-18,  1983,  and  dedicated  to  the  memory 
of  Elbert  Halvor  Ahlstrom.  Dr.  R,  Lasker  served  as  convener.  The  papers  presented  at  that  symposium  form  the  basis 
for  this  book,  which  is  published  by  the  American  Society  of  Ichthyologists  and  Herpetologists  as  their  Supplement  to 
Copeia,  Special  Publication  Number  1 .  Financial  support  was  provided  by  the  National  Marine  Fisheries  Service,  National 
Oceanic  and  Atmospheric  Administration,  U.S.  Department  of  Commerce. 

For  many  years.  Dr.  Ahlstrom  planned  to  write  a  book  on  larval  fishes  and  ways  in  which  they  contributed  to  systematics. 
A  few  years  before  his  untimely  death,  he  and  his  colleague  H.  G.  Moser  outlined  such  a  book  and  began  to  work  on 
the  initial  chapters.  Dr.  Ahlstrom  left  a  vast  store  of  notes,  data,  and  partly  completed  manuscnpts.  Dr.  Moser  realized 
that  much  of  the  significance  of  these  unique  and  important  data  would  be  lost  unless  they  were  brought  to  light.  He 
approached  colleagues  at  the  Southwest  Fisheries  Center  to  gather  a  group  of  larval  fish  workers  who  had  worked  closely 
with  Dr.  Ahlstrom,  and  who  were  given  access  to  his  notes,  to  collaborate  on  the  book.  From  this  initiative  a  plan 
developed  to  conduct  a  symposium  and  publish  the  results  in  a  book  to  accomplish  the  original  plan  of  Dr.  Ahlstrom 
and  honor  his  memory  as  one  of  the  nation's  foremost  fishery  scientists. 

A  symposium  steering  committee  was  formed  with  H.  G.  Moser  as  Chairman  and  consisted  of  D.  M.  Cohen,  M.  P. 
Fahay,  A.  W.  Kendall,  Jr..  W.  J.  Richards  and  S.  L.  Richardson.  The  steering  committee  first  met  in  Boulder,  Colorado 
to  develop  an  outline  for  the  symposium  and  book  and  invite  potential  contributors.  The  aim  was  to  present  the  current 
state  of  knowledge  of  early  life  history  of  fishes  and  apply  that  to  systematics.  Originally  it  was  intended  to  concentrate 
solely  on  the  marine  groups  with  which  Dr.  Ahlstrom  had  worked,  but  because  of  recent  advances  in  freshwater  and 
other  early  life  history  work,  the  plan  was  expanded  to  include  all  but  the  primitive  osteoglossomorphs.  Thus,  the  coverage 
was  to  start  with  the  elopomorphs. 

Following  the  Boulder  meeting,  potential  contributors  were  contacted  and  responded  enthusiastically.  The  Steering 
Committee  met  subsequently  in  Ocean  Springs,  Mississippi  and  Miami,  Florida  to  review  progress  and  refine  plans. 
Because  of  the  subject  matter  it  seemed  appropriate  that  the  American  Society  of  Ichthyologists  and  Herpetologists 
collaborate  in  publishing  the  papers  resulting  from  the  symposium.  C.  R.  Robins,  then  President  of  ASIH,  supported 
this  suggestion  and  assisted  in  many  ways.  Subsequent  to  the  symposium,  manuscripts  were  reviewed  and  edited  by  the 
Steering  Committee  of  the  Symposium,  which  served  as  an  editorial  committee  for  this  volume. 

The  Steering  Committee  thanks  all  of  the  authors  of  this  volume  among  whom  there  was  a  great  exchange  of  ideas 
and  generous  help.  Much  additional  assistance  was  provided  to  the  authors  and  is  here  acknowledged.  Institutional 
support  was  provided  by  the  National  Marine  Fisheries  Service  through  contributions  from  each  of  the  four  Fisheries 
Centers— Southwest,  Southeast,  Northwest  and  Alaska  and  Northeast.  Support  was  provided  by  the  National  Science 
Foundation  through  grants  DEB76-82279,  DEB78-26540;  the  National  Geographic  Society  by  grant  2535-82  from  the 
Committee  for  Research  and  Exploration;  the  Robert  E.  Maytag  Fellowship  at  the  University  of  Miami;  Natural  History 
Museum  of  Los  Angeles  County;  the  Australian  Museum  Trust,  the  Australian  Marine  Science  and  Technologies  Advisory 
Committee,  the  Commonwealth  Science  and  Industrial  Research  Organization  Science  and  Industry  Endowment  Fund, 
and  the  employers  of  the  contributors. 

The  following  individuals  supplied  specimens,  data,  technical  assistance,  publications,  and  reviewed  drafts  of  manu- 
scripts: M.  Allen.  R.  M.  Allen,  A.  Alvarino,  D.  Ambrose,  M.  E.  Anderson,  W.  D.  Anderson.  Jr..  F.  Balbontin.  C.  Baldwin, 
E.  K.  Balon,  P.  Berrien,  D.  Blood,  S.  Boardman,  S.  S.  Boggs,  E.  Bohlke,  M.  Bradbury,  J.  Brill,  D.  Brown,  J.  Bullock,  M. 
S.  Busby.  J.  A.  Cambray,  P.  Camus,  M.  H.  Carrington,  B.  Chemoff,  T.  A.  Clarke,  M.  Culbreth,  M.  Cluxton,  S.  Coombs, 

A.  S.  Creighton,  K.  Davis,  W.  P.  Davis.  C.  E.  Dawson.  M.  Dehaan,  N.  Demir,  A.  Desai,  H.  H.  DeWitt,  M.  DeWitt,  Y. 
Dotsu,  S.  D'Vincent,  B.  R.  Engstrand,  D.  Faber,  N.  R.  Foster,  P.  Fourmanoir,  C.  Frandsen,  H.  J.  Franke,  E.  Fridgeirsson, 
W.  George,  R.  H.  Gibbs,  G.  Gilmore,  D.  Gittings,  W.  Gladstone,  T.  Goh.  M.  F.  Gomon.  B.  Goldman,  A.  R.  Gosline, 
W.  A.  Gosline,  A.  E.  Gosztonyi,  P.  H.  Greenwood,  D.  Haggner,  G.  R.  Harbison,  G.  S.  Hardy,  K.  Hartel.  R.  Hartwick, 
T.  Hecht,  E.  Hubert.  J.  M.  Humphries.  J.  C.  Hureau.  T.  Iwamoto.  S.  Jewett,  P.  Keener,  S.  Kelley,  F.  Kirschbaum,  N. 
Komada,  Y.  Konishi,  D.  L.  Kramer,  J.  K.  Langhammer.  K.  Lazara,  K.  Lee,  S.  Lincoln,  J.  Lobon-Cervia,  V.  J.  Loeb,  G. 
Lundy,  N.  A.  Mackintosh,  F.  Mago-Leccia,  A.  M.  Martinez,  D.  McAllister,  M.  McCabe,  J.  McCosker.  R.  F.  McGinnis. 
R.  McMichael.  R.  Meier.  N.  Merrett.  J.  Michalski,  J.  Mighell,  R.  R.  Miller,  C.  Mills,  A.  Miskiewicz,  G.  E.  E.  Moodie, 
K.  H.  Moore,  K.  Mori,  J.  Moyer,  J.  A.  Musick,  T.  Nakata,  G.  Nelson,  J.  Nelson,  J.  Nichols,  J.  Nielsen,  T.  North,  S. 
Ochman,  G.  Patchell,  L.  R.  Parenti,  K.  Peters.  T.  Pomeranz,  S.  Poss,  L.  C.  Prescott,  J.  Quast,  J.  Randall,  K.  S.  Raymond, 

B.  Remington,  C.  S.  Richards,  T.  Roberts,  D.  E.  Rosen.  R.  Schoknecht,  A.  Sekerak,  T.  Senta,  J.  Shapiro.  J.  Shoemaker, 
P.  L.  Shafland,  M.  Shiogaki,  D.  L.  Schultz,  P.  H.  Skelton,  P.  E.  Smith,  J.  Song,  D.  E.  Snyder,  A.  Soeldner,  C.  Stehr,  D. 
Stein,  B.  Stender,  K.  Steward,  K.  Stoddard,  R.  E.  Strauss,  G.  Stroud,  K.  J.  Sulak,  A.  Suzumoto,  H.  Sweatman,  J.  N. 
Taylor,  V.  R.  Thomas,  G.  Theilacker.  R.  Thresher,  R.  Triemer,  D.  Tweedle,  J.  C.  Tyler,  F.  Utter,  F.  Van  Dolah,  R. 
Vari,  B.  Vinter.  L.  Vlyman,  R.  Wallus.  T.  Watanabe.  B.  A.  Watkins,  A.  Wheeler,  P.  Whitehead,  N.  Wilimovsky,  A.  B. 
Williams,  L.  Wood,  B.  L.  Yeager,  P.  Yuschak.  H.  Zadoretsky.  B.  J.  Zahuranec. 

Illustrators  deserve  special  praise  and  thanks.  B.  B.  Washington  illustrated  a  large  majority  of  the  specimens.  Other 
illustrators  include  G.  Mattson  who  served  in  this  capacity  with  Dr.  Ahlstrom  for  many  years.  B.  Y.  Sumida  and  H.  Orr 
at  the  Southwest  Fisheries  Center.  B.  Vinter  at  the  Northwest  and  Alaska  Fisheries  Center  and  J.  C.  Javech  at  the 


Southeast  Fisheries  Center.  The  original  illustrations  are  archived  at  the  Southeast  Fisheries  Center.  Miami  and  Southwest 
Fisheries  Center,  La  Jolla. 

During  the  final  editorial  processes,  J.  C.  Javech  and  B.  B.  Washington  mounted  illustrations  and  remade  many  that 
were  of  marginal  quality.  C.  Wolf  coordinated  and  reviewed  the  literature  cited  section  and  P.  Fisher  typed  the  literature 
cited  section  as  well  as  all  last  minute  editorial  changes. 

The  Editorial  Committee: 

H.  G.  Moser,  Editor  in  Chief 

W.  J.  Richards,  Managing  Editor 

D.  M.  Cohen 

M.  P.  Fahay 

A.  W.  Kendall,  Jr. 

S.  L.  Richardson 


CONTENTS 

Welcoming  Address.     By  /.  Barrett vii 

Frontispiece— Photograph  of  Elbert  Halvor  Ahlstrom.     By  /.  R.  Dunn viii 

Dr.  Ahlstrom.     By  R.  Lasker _ ix 

Introduction 

Ontogeny,  Systematics  and  Fisheries.     By  J.  H.  S.  Blaxter __ __ 1 

Ontogeny,  Systematics  and  Phylogeny.     By  D.  M.  Cohen 7 

Early  Life  History  Stages  of  Fishes  and  Their  Characters.     By  A.   W.  Kendall,  Jr..  E.  H.  Ahlstrom  and  H.  G. 

Moser 1 1 

Techniques  and  Approaches 

Early  Life  History  Descriptions.     By  E.  M.  Sandknop.  B.  Y.  Sumida  and  H.  G.  Moser _ 23 

Synopsis  of  Culture  Methods  for  Marine  Fish  Larvae.     By  J.  R.  Hunter 24 

Identification  of  Fish  Eggs.     By  A.  C.  Matarese  and  E.  M.  Sandknop 27 

Identification  of  Larvae.     By  H.  Powles  and  D.  F.  Markle 31 

Illustrating  Fish  Eggs  and  Larvae.     By  B.  Y.  Sumida,  B.  B.  Washington  and  W.  A.  Laroche 33 

Clearing  and  Staining  Techniques.     By  T.  Potthoff 35 

Radiographic  Techniques  in  Studies  of  Young  Fishes.     By  /.  W.  Tucker,  Jr.  and  J.  L.  Laroche 37 

Histology.     By  y.  y.  Govoni _  40 

Scanning  Electron  Microscopy.     By  G.  W.  Bochlert _ _ 43 

Developmental  Osteology.     By  J.  R.  Dunn 48 

Otolith  Studies.     By  E.  B.  Brothers __ 50 

Preservation  and  Curation.     By  R.  J.  Lavcnhcrg.  G.  E.  McGowen  and  R.  E.  Woodsum 57 

Development  and  Relationships 

Elopiformes:  Development.     By  W.  J.  Richards 60 

Notacanthiformes  and  Anguilliformes:  Development.     By  P.  H.  J.  Castle _ 62 

Elopiformes,  Notacanthiformes  and  Anguilliformes:  Relationships.     By  D.  G.  Smith 94 

Ophichthidae:  Development  and  Relationships.     By  M.  M.  Leiby 102 

Clupeiformes:  Development  and  Relationships.     By  M.  F.  McGowan  and  F.  H.  Berry 108 

Ostariophysi:  Development  and  Relationships.     By  L.  .A.  Fuiman 126 

Gonorynchiformes:  Development  and  Relationships.     By  H'.  J.  Richards 138 

Salmoniforms:  Introduction.     By  H '.  L.  Fink _ 1 39 

Esocoidei:  Development  and  Relationships.     By  F.  D.  Martin 140 

Salmonidae:  Development  and  Relationships.     By  A.  W.  Kendall.  Jr.  and  R.  J.  Behnke 142 

Southern  Hemisphere  Freshwater  Salmoniforms:  Development  and  Relationships.     By  R.  M.  McDowall _...  150 

Osmeridae:  Development  and  Relationships.     By  M.  E.  Hcarnc 153 

Argentinoidei:  Development  and  Relationships.     By  E.  H.  .Ahlstrom.  H.  G.  Moser  and  D.  M.  Cohen 155 

Stomiatoidea:  Development.     By  A'.  Kawaguchi  and  H.  G.  Moser 169 

Stomiiforms:  Relationships.     By  W.  L.  Fink                 1 8 1 

Families  Gonostomatidae,  Stemoptychidae.  and  Associated  Stomiiform  Groups:  Development  and  Relation- 
ships.    By  E.  H.  .Ahlstrom.  \V.  J.  Richards  and  S.  H.  Wcitzman 184 

Giganturidae:  Development  and  Relationships.     By  R.  K.  Johnson _ 199 

Basal  Euteleosts:  Relationships.     By  W.  L.  Fink _ 202 

Myctophi formes:  Development.     By  M.  Okiyama _ 206 

Myctophidae:  Development.     By  H.  G.  Moser.  E.  H.  Ahlstrom  and  J.  R.  Paxton 218 

Myctophidae:  Relationships.     By  J.  R.  Pa.xton.  E.  H.  Ahlstrom  and  //.  G.  Moser 239 

Scopelarchidae:  Development  and  Relationships.     By  R.  K.  Johnson 245 

Evermannellidae:  Development  and  Relationships.     By  R.  K.  Johnson 250 

Myctophiformes:  Relationships.     By  M.  Okiyama 254 

Gadiformes:  Overview.     By  D.  M.  Cohen 259 

Gadiformes:  Development  and  Relationships.     By  M.  P.  Fahay  and  D.  F.  Markle _ __ 265 


Gadidae;  Development  and  Relationships.     By  J.  R.  Dunn  and  A.  C.  Matarese 283 

Bregmacerotidae:  Development  and  Relationships.     By  E.  D.  Houde 300 

Ophidiiformes:  Development  and  Relationships.     By  D.  J.  Gordon.  D.  F.  Markle  and  J.  E.  Olney 308 

Lophiiformes:  Development  and  Relationships.     By  T.  W.  Pietsch _ 320 

Ceratioidei:  Development  and  Relationships.     By  E.  Bertelsen __ _ _ 325 

Atherinomorpha:  Introduction.     By  B.  B.  Collette 334 

Beloniformes:  Development  and  Relationships.     By  B  B.  Collette,  G.  E.  McGowen.  N.  V.  Parin  and  5.  Mito 335 

Atheriniformes:  Development  and  Relationships.     By  B.  N.  White,  R.  J.  Lavenberg  and  G.  E.  McGowen 355 

Cyprinodontiformes:  Development.     By  K.  W.  Able 362 

Lampriformes:  Development  and  Relationships.     By  J.  E.  Olney 368 

Mirapinnatoidei:  Development  and  Relationships.     By  E.  Bertelsen  and  TV.  B.  Marshall 380 

Beryciformes:  Development  and  Relationships.     By  M.  J.  Keene  and  A'.  A.  Tighe 383 

Zeiformes:  Development  and  Relationships.     By  A'.  A.  Tighe  and  M.  J.  Keene 393 

Gasterosteiformes:  Development  and  Relationships.     By  R.  A.  Fritzsche 398 

Scorpaeniformes:  Development.     By  B.  B.  Washington,  H.  G.  Moser,  W.  A.  Laroche  and  W.  J.  Richards 405 

Cyclopteridae:  Development.     By  A'.  W.  Able,  D.  F.  Markle  and  M.  P.  Fahay 428 

Scorpaeniformes:  Relationships.     By  B.  B.  Washington,  W.  N.  Eschmeyer  and  K.  M.  Howe 438 

Tetraodontoidei:  Development.     By  J.  A/.  Lets 447 

Balistoidei:  Development.     By  A.  Aboussouan  and  J.  M.  Lets _ 450 

Tetraodontiformes:  Relationships.     By  J.  M.  Lets - 459 

Percoidei:  Development  and  Relationships.     By  G.  D.  Johnson _ 464 

Serranidae:  Development  and  Relationships.     By  A.  W.  Kendall,  Jr. _ 499 

Carangidae:  Development.     By  W.  A.  Laroche,  W.  F.  Smith- Vaniz  and  S.  L.  Richardson 510 

Carangidae:  Relationships.     By  W.  F.  Smith-  Vaniz 522 

Mugiloidei:  Development  and  Relationships.     By  D.  P.  de  Sylva 530 

Sphyraenoidei:  Development  and  Relationships.     By  D.  P.  de  Sylva 534 

Polynemoidei:  Development  and  Relationships.     By  D.  P.  de  Sylva 540 

Labroidei:  Development  and  Relationships.     By  W.  J.  Richards  and  J.  M.  Leis 542 

Acanthuroidei:  Development  and  Relationships.     By  J.  M.  Leis  and  W.  J.  Richards 547 

Blennioidei:  Introduction.     By  R.  H.  Rosenblatt 551 

Schindlerioidei:  Development  and  Relationships.     By  W.  Watson.  E.  G.  Stevens  and  A.  C.  Matarese 552 

Trachinoidea:  Development  and  Relationships.     By  W.  Watson.  A.  C.  Matarese  and  E.  G.  Stevens _ 554 

Notothenioidea:  Development  and  Relationships.     By  E.  G.  Stevens.  W.  Watson  and  A.  C.  Matarese —  561 

Blennioidea:  Development  and  Relationships.     By  A.  C.  Matarese.  W.  Watson  and  E.  G.  Stevens 565 

Ammodytoidei:  Development  and  Relationships.     By  E.  G.  Stevens.  A.  C.  Matarese  and  W.  Watson 574 

Icosteoidei:  Development  and  Relationships.     By  A.  C.  Matarese.  E.  G.  Stevens  and  W.  Watson 576 

Zoarcidae:  Development  and  Relationships.     By  M.  E.  Anderson _ 578 

Gobioidei:  Development.     By  D.  Ruple 582 

Gobioidei:  Relationships.     By  D.  F.  Hoese 588 

Scombroidei:  Development  and  Relationships.     By  B.  B.  Collette,  T.  Potthoff,   W.  J.  Richards,  S.   Ueyanagi, 

J.  L.  Russo  and  Y.  Nishikawa 59 1 

Stromateoidei:  Development  and  Relationships.     By  M.  H.  Horn _ 620 

Gobiesociformes:  Development  and  Relationships.     By  L.  G.  Allen 629 

Callionymidae:  Development  and  Relationships.     By  E.  D.  Houde 637 

Pleuronectiformes:  Development.     By  E.  H.  Ahlstrom,  K.  Amaoka,  D.  A.  Hensley,  H.  G.  Moser  and  B.  Y.  Su- 

mida 640 

Pleuronectiformes:  Relationships.     By  D.  A.  Hensley  and  E.  H.  Ahlstrom - 670 

Literature  Cited _ 688 

Index _ - 746 

Photograph  of  Symposium  Attendees  760 


VI 


Welcoming  Address 

IzADORE  Barrett 
Director  of  the  Southwest  Fisheries  Center 

ON  behalf  of  the  National  Marine  Fisheries  Service's  Center  Directors,  sponsors  of  the  Symposium  on  the  Ontogeny 
and  Systematics  of  Fishes,  I  am  pleased  and  honored  to  welcome  you  to  La  JoUa.  We  are  here  to  honor  the  memory 
of  an  outstanding  biologist,  Elbert  Halvor  Ahlstrom,  known  to  his  friends  and  colleagues  as  Ahlie,  and  his  contributions 
to  fisheries  science. 

As  fishery  biologists  we  all  recognize  the  vital  importance  and  contributions  of  systematics  and  students  of  evolution 
to  the  development  of  fishery  science.  Less  well  known  or  appreciated  is  the  unique  role  and  interrelationship  of  the 
early  life  history  studies  of  fishes  and  the  assessment  of  the  role  of  ontogenetic  characters  in  fish  systematics.  This  was, 
of  course,  the  field  of  fisheries  research  to  which  Ahlie  dedicated  40  years  of  his  professional  life  and  where  he  initially 
evolved  the  special  methods  and  techniques  which  have  so  greatly  influenced  the  work  of  fishery  biologists  around  the 
world. 

I  know  that  I  speak  for  the  Directors  of  the  four  fisheries  centers— the  Northwest  and  Alaska  Fisheries  Center  in 
Seattle,  the  Southwest  Fisheries  Center  in  La  Jolla,  the  Northeast  Fisheries  Center  in  Woods  Hole,  and  the  Southeast 
Fisheries  Center  in  Miami  when  I  say  that  I  am  proud  that  the  National  Marine  Fisheries  Service  is  the  sponsor  of  this 
symposium.  I  believe  that  this  gathering  will  be  a  landmark  in  fisheries  science,  a  unique  event  which  has  brought 
together  eminent  scientists  from  10  countries  to  present  87  papers  reviewing  the  major  fish  groups,  with  particular 
attention  to  ontogenetic  characters  and  their  utility  in  assessing  phylogenetic  relationships.  I  fully  anticipate  that  the 
resulting  symposium  volume  which  will  be  based  on  the  papers  presented  here  will  stand  as  a  definitive  work  in  larval 
fish  biology  for  many  years  to  come. 

Again,  a  warm  welcome  to  all  of  you  and  especially  to  Marge  Ahlstrom  who  is  seated  in  the  audience  this  morning. 
I  hope  that  the  weather  and  circumstances  will  cooperate  and  that  your  stay  here  in  one  of  the  most  attractive  cities  of 
the  United  States  will  be  pleasant  and  productive. 

P.O.  Box  271,  La  Jolla,  California  92038. 


Dr.  Ahlstrom 
Reuben  Lasker 

MY  colleagues  have  entrusted  to  me  the  pleasant  task  and  distinct  privilege  of  saying  a  few  words  in  remembrance 
of  Dr.  Elbert  H.  Ahlstrom,  to  whom  this  symposium  is  dedicated.  Like  most  of  you  I  was  his  colleague  for  many 
years,  23  to  be  exact.  He  was  also  my  friend  and  mentor  to  whom  I  could  go  when  I  needed  advice  and  where  I  knew 
I  would  be  heard  as  an  individual  with  the  bond  of  common  scientific  endeavors. 

For  those  of  you  who  did  not  know  Dr.  Ahlstrom  1  would  like  to  capsulize  his  enormous  contribution  to  systematics 
and  fishery  science  by  outlining  what  I  believe  to  be  his  major  scientific  contributions.  Ahlie  realized  in  the  late  40's 
that  the  study  of  eggs  and  larvae  could  give  us  information  about  fish  populations  unobtainable  from  fishery  statistics, 
the  mainstay  of  fishery  science  at  that  time.  He  believed,  rightly,  that  the  ease  with  which  eggs  and  larvae  could  be  caught 
allowed  an  assessment  of  the  geographic  distribution  and  the  seasonal  extent  of  spawning  of  pelagic  species.  He  recognized 
that  any  assessment  of  a  fish  population  was  dependent  on  surrounding  that  population  in  time  and  space  and  that  this 
would  require  a  major  effort.  He  was  the  first.  I  believe,  to  determine  the  extent  of  a  major  pelagic  fish  population  using 
this  technique. 

The  simplicity  and  thoroughness  of  the  plankton  net  made  an  impression  on  him  and,  while  he  sought  to  improve 
collecting  techniques  constantly,  he  consistently  analyzed  the  errors  of  the  plankton  net  so  that  this  tool  could  be  used 
more  and  more  reliably.  Today,  it  is  still  one  of  the  most  powerful  collecting  and  assessment  tools  we  have,  largely 
because  of  his  diligence  and  persistence. 

The  scope  and  thoroughness  of  Dr.  Ahlstrom's  work  was  particularly  important.  His  taxonomic  skills  are  attested  to 
in  the  many  papers  he  wrote  and  which  stand  today  as  mainstays  of  the  systematic  and  fishery  literature.  He  liked  to  use 
the  title  "Kinds  and  abundance  of  fishes"  and  usually  provided  taxonomic  lists  in  these  of  several  pages  in  length.  His 
point,  of  course,  was  to  detail  the  complexity  and  uniqueness  of  particular  oceanic  regimes  and  to  set  the  ground  work 
for  ecological  research  which  inevitably  followed. 

Well,  what  of  his  other  attributes?  I  used  to  call  him  the  modem  Renaissance  Man  because  I  realized  whenever  I  had 
occasion  to  meet  him  socially  that  he  knew  almost  all  there  was  to  know  about  the  arts  and  the  sciences.  Of  his  fabulous 
classical  record  collection  1  recall  that  1  asked  him  once  if  he  really  listened  to  all  of  them.  His  reply  was  "we  used  to 
hear  each  one  once  a  year,  but  now,  since  the  collection  has  grown  so  large,  it's  once  every  two  years."  He  belonged  to 
the  San  Diego  Great  Books  Society,  and  read  them  all.  Engage  him  in  conversation  and  you  would  find  out  quickly  he 
knew  literature,  fine  wines,  photography  and  baseball,  to  name  a  few.  I  would  like  to  sum  up  this  brief  eulogy  by  pointing 
out  an  example  of  one  aspect  of  Ahlie  which  holds  my  greatest  admiration:  that  is,  his  dedication  to  work.  One  incident 
during  our  relationship  illustrates  the  point  I  wish  to  make. 

When  Science  Fairs  started  to  become  the  vogue  in  San  Diego,  Dr.  Ahlstrom  was  asked  to  host  a  group  of  young 
Science  Fair  participants  to  teach  them  something  about  oceanography.  He  arranged  to  take  out  the  old  Bureau  of 
Commercial  Fisheries  ship,  the  Black  Douglas,  for  a  day  to  illustrate  collecting  methods  at  sea.  In  fact,  the  day  was 
beautiful,  but  there  was  a  swell  upon  the  sea  and  no  sooner  did  we  get  out  of  the  harbor  than  almost  everyone,  except 
Ahlie  and  some  of  the  seasoned  veterans,  felt  the  effects  of  a  rather  pronounced  roll  for  which  the  Black  Douglas  was 
famous,  even  in  the  calmest  of  seas.  Dr.  Ahlstrom  proceeded  with  his  typical  dedication  to  illustrate  Nansen  bottles, 
plankton  nets,  and  bathythermographs  to  the  group  of  Science  Fair  students  who  were  becoming  less  and  less  interested 
and  more  and  more  seasick. 

Ahlie  continued  with  a  single-mindedness  of  purpose  and  a  dedication  that  was  so  characteristic  of  him.  Without  his 
noticing,  a  caucus  was  held  by  these  young  students  and  a  representative  meekly  asked,  "Dr.  Ahlstrom,  may  we  please 
go  home?" 

Two  versions  of  what  happened  next  were  told  to  me  later.  The  first  was  that  Ahlie  responded  immediately  to  the 
problem  and  ordered  the  ship  to  port.  Another  version  was  that  Ahlie  continued  until  he  was  finished,  made  sure  he 
had  a  proper  sample,  and  then  ordered  the  ship  into  port.  I'm  afraid  I  can't  tell  you  which  is  correct— I  was  in  a  bunk, 
seasick!  I  meant  this  story  as  a  small  illustration  of  Dr.  Ahlstrom's  dedication  to  his  work. 

He  was  a  dedicated  scientist  who  had  an  insatiable  curiosity  about  the  biotic  world  and  who  was  convinced  that  what 
he  was  doing  was  important  and  would  advance  fishery  science.  This  symposium  is  one  piece  of  evidence  that  he  was 
right. 

Now  the  question  must  be  asked— how  is  it  that  Ahlie  could  be  so  dedicated  to  work  and  yet  have  found  time  to 
become  a  true  example  of  a  Renaissance  man,  with  a  deep  knowledge  of  art,  wine,  architecture,  photography,  sports, 
and  much  more?  I  pondered  this  with  admiration  for  many  years  and  I  think  I  have  the  answer.  He  was  one  of  those 
rare  individuals  who  never  cease  learning,  because  he  had  a  true  scholar's  love  for  learning.  I  like  Robert  Whittenton's 
description  of  Sir  Thomas  More  when  I  think  of  Ahlie:  he  was,  like  More,  "a  man  for  all  seasons." 

Southwest  Fisheries  Center,  P.O.  Box  271,  La  Jolla,  California  92038. 


Photograph  of  Elbert  Halvor  Ahlstrom,  by  J.  R.  Dunn. 


INTRODUCTION 


Ontogeny,  Systematics  and  Fisheries 
J.  H.  S.  Blaxter 


IN  the  inter-war  years  work  on  fish  eggs  and  larvae  was  Umited 
to  studies  on  horizontal  and  vertical  distribution  with  a  view 
to  completing  our  knowledge  of  the  early  life  history  of  different 
species.  Resources  for  research  were  then  much  more  limited 
than  they  are  today  and  most  work  was  done  on  the  important 
food  fishes.  In  the  1 950's  a  great  expansion  took  place  as  fisheries 
biologists  realised  how  much  a  study  of  early  life  history  would 
be  a  key  to  solving  some  of  their  problems.  This  expansion  took 
place  on  a  broad  geographical  and  mtemational  front,  but  great 
credit  must  be  given  to  the  foresight  and  imagination  of  E.  H. 
Ahlstrom.  who  built  up  a  team  of  biologists  at  La  Jolla  who 
then  and  subsequently,  played  a  major  role  m  leading  and  de- 
veloping this  field  with  special  reference  to  the  fisheries  of  the 
California  Current. 

In  the  last  two  decades  the  output  of  publications  has  risen 
at  an  exponential  rate  as  evidenced,  for  example,  by  the  62 
papers  in  the  1973  Early  Life  History  Symposium  held  in  Oban 
(Blaxter,  1974)  and  the  139  papers  in  the  1979  Symposium  at 
Woods  Hole  (Lasker  and  Sherman.  1981).  Furthermore,  in  a 
selected  hibhography  of  pelagic  fish  and  larva  surveys  prepared 
by  Smith  and  Richardson  (1979),  some  1200  papers  are  listed, 
most  of  them  published  in  the  last  30  years.  Ahlstrom  was 
certainly  a  major  catalyst  in  this  reaction,  but  it  is  sad  to  record 
that  his  obituary  appeared  in  the  Proceedings  of  the  1979  Sym- 
posium, although  he  was  still  alive  and  present  at  the  meeting 
itself  to  impart  his  wisdom  and  expertise. 

It  is  proposed  to  discuss  the  post-war  advances  in  our  knowl- 
edge of  early  life  history  stages  under  five  headings:  (1)  as  they 
impinge  on  systematics  and  taxonomy.  (2)  the  success  and  role 
o{  experimental  work  in  tanks  and  of  modelling,  (3)  the  scaling- 
up  of  tank  studies  to  large  enclosures  and  embayments,  (4)  the 
application  oi sea  surveys  to  test  models,  to  investigate  the  stock- 
recruitment  relationship  and  to  measure  spawning  stock  bio- 
mass,  and  (5)  \he  future. 

Systematics  and  Taxonomy 

A  number  of  techniques  have  been  developed  to  help  in  the 
identification  and  classification  of  fish  larvae.  Since  the  devel- 
opment of  the  skeleton  and  meristic  characters  are  now  so  im- 
portant in  identification,  techniques  of  clearing  and  staining  or 
x-radiography  have  become  standard  methods  for  examining 
the  internal  osteology  of  larvae  (Ahlstrom  and  Moser,  1981). 
Morphometries  and  body  pigmentation  are  also  important  and 
are  used  extensively  by  Russell  (1976)  in  his  monograph  on  fish 
eggs  and  larvae  of  the  N.E.  Atlantic. 

Rearing  experiments  have  shown  that  the  sequence  of  de- 
velopmental events  may  also  be  specific  in  character.  For  ex- 
ample the  development  of  the  acoustico-latcralis  system  and 
swimbladder  in  herring  as  shown  by  Allen,  Blaxter  and  Denton 
(1976)  is  a  long-drawn-out  affair  and  quite  different  from  that 
of  the  larval  anchovy  as  described  by  O'Connell  ( 1 98 1  a)  or  the 
menhaden  or  sprat.  There  are  several  larval  features,  such  as 


the  swimbladder  and  other  internal  organs,  or  features  of  the 
labyrinth,  which  would  help  in  the  separation  of  similar-looking 
species  if  only  they  were  not  obscured  by  fixation. 

Often  the  taxonomist  (or  fisheries  biologist)  resorts  to  count- 
ing menstic  characters  such  as  vertebrae,  fin  rays,  scales  or  gill 
rakers.  Yet  many  of  these  characters  have  been  shown  by  ex- 
periment to  be  labile  and  to  respond  to  environmental  condi- 
tions during  early  development.  The  earlier  work,  mainly  on 
freshwater  species  such  as  the  sea  trout,  was  summarised  by 
Taning  (1952).  Since  then  a  range  of  further  studies  by  Fahy, 
Lindsey  (e.g.,  see  Fahy,  1982)  and  others  have  confirmed  the 
earlier  experiments,  showing  that  temperature,  salinity  and  oxy- 
gen level  influence  meristic  counts  and  that  there  is  a  critical 
period  when  this  influence  operates.  Little  work  has  been  done 
on  marine  species  although  Hempel  and  Blaxter  (1961)  showed 
that  temperature  and  salinity  both  influence  myotome  and  ver- 
tebral counts  in  herring  (the  species  in  which  stock  separation 
by  meristic  counts  has  been  most  widely  applied). 

It  seems  likely  that  any  environmental  variable  which  influ- 
ences the  relationship  between  differentiation  and  growth  will 
affect  the  meristic  count  by  determining  the  amount  of  embry- 
onic tissue  which  is  present  when  the  differentiation  into  skeletal 
units  lakes  place.  The  larval  taxonomist  needs  to  be  cautious 
in  interpreting  small  differences  in  meristic  values,  especially 
when  they  are  related  to  clines  or  other  types  of  geographical 
distribution.  That  is  not  to  say,  however,  that  there  is  no  un- 
derlying genetic  mechanism.  The  environment  acts  as  a  "fine- 
tuning"  mechanism.  Whether  this  fine-tuning  is  accidental  or 
adaptive  might  well  be  worth  discussion  at  the  symposium. 

A  warning  also  needs  to  be  directed  at  morphometries.  Rear- 
ing experiments  in  different-sized  tanks  by  Theilacker  ( 1 980b) 
show  the  influence  of  space  on  growth  rates.  Compansons  of 
reared  and  wild  fish  larvae,  especially  of  herring  by  Blaxter 
(1976),  show  that  tank-reared  fish  are  often  shorter  and  fatter 
than  their  wild  counterparts  at  the  same  developmental  stage. 
There  seems  to  be  an  interplay  between  diet  and  activity  which 
is  enhanced  by  the  confinements  of  the  rearing  tank.  This  makes 
it  difficult  to  extrapolate  growth  criteria  from  tanks,  such  as 
condition  factor,  to  establish,  for  example,  the  nutritional  status 
of  larvae  at  sea  (Fig.  1). 

A  further  and  serious  problem  identified  by  the  handling  and 
use  of  live  larvae  is  the  shrinkage  caused  by  capture  and  fixation. 
A  number  of  workers  such  as  Blaxter  (1971),  Schnack  and  Ro- 
senthal (1978),  Theilacker  (1980a)  and  Bailey  (1982)  have  ad- 
dressed this  problem  but  the  most  significant  findings  are  those 
of  Hay  (1981)  on  Pacific  herring.  Feeding  larvae  from  rearing 
experiments  were  released  into  the  mouth  of  a  plankton  net  at 
sea  and  then  fixed  by  various  techniques  after  capture.  Shrinkage 
in  body  length  ranged  from  a  mere  5%  to  a  massive  43%  de- 
pending on  the  technique.  Extensive  voiding  of  gut  contents  also 
occurred.  The  implications  of  these  results  in  morphometric  or 
feeding  studies  will  not  be  lost  on  the  present  audience. 


1 


ONTOGENY  AND  SYSTEM ATICS  OF  FISHES -AHLSTROM  SYMPOSIUM 


•20 


.16- 


(J 


•12 


•08- 


HERRING 


WILD 


12  16 

LENGTH  (mm) 


"20 


Fig.  1 .  Comparison  between  range  of  condition  factors  (C.F.)  as  dry 
weight/length^  of  wild  herring  caught  at  sea  by  plankton  net  and  reared 
herring  larvae  near  starvation  (from  Blaxter,  1976). 


Finally,  the  ageing  of  larvae  by  daily  ring  formation  in  the 
otoliths  should  be  mentioned.  This  technique  was  pioneered  by 
Brothers  et  al.  (1976)  on  anchovy  larvae  and  California  grunion 
following  Pannella's  suggestion  that  daily  increments  were  being 
laid  down  in  the  sagittae  of  some  temperate  fish  species.  The 
findings  were  validated  by  rearing  larvae  in  tanks  and  sampling 
the  population  at  intervals  of  1-7  days.  Struhsaker  and  Uchi- 
yama  (1976)  supported  these  results  from  their  work  on  the 
Hawaiian  nehu  and  subsequently  the  technique  was  widely 
adopted  in  fisheries  laboratories.  Attempts  by  Geffen  (1982)  to 
manipulate  ring  formation  in  cod,  herring,  plaice,  salmon  and 
turbot  larvae  by  varying  the  photoperiod,  temperature  and  feed- 
ing regimes  did  not  lead  to  any  consistent  result  — the  ring  de- 
position was  frequently  not  daily  and  the  main  determinant  in 
herring  and  turbot  seemed  to  be  growth  rate— the  higher  the 
growth  rate,  the  higher  the  rate  of  ring  deposition.  Bailey  (1982), 
however,  found  otolith  rings  deposited  daily  over  a  10-day  pe- 
riod in  post  yolk-sac  Pacific  hake  larvae  reared  in  tanks.  Sea- 
caught  larvae  with  more  than  about  30  increments  were  less 
satisfactory  because  of  the  appearance  of  different  types  of  ring 
and  it  was  not  certain  whether  they  were  daily.  Dale  (1984)  in 
a  recent  study  of  reared  Atlantic  cod  otoliths  using  electromi- 
croscopy,  found  daily  rings  in  a  12L/12D  cycle  but  not  in  the 
dark.  Daily  ring  deposition  only  continued,  however,  for  a  few 
days  post-hatching. 

Although  the  ageing  of  anchovy  and  grunion  from  daily  rings 
seems  reliable,  further  validation  experiments  are  required  at 
sea.  This  is  conceptually  difficult  on  a  wild  stock  of  larvae  of 
mixed  age  and  it  is  notoriously  difficult  to  remain  over  a  single 
population  of  larvae  for  many  days.  Mass  release  of  reared 
larvae  into  the  sea  remains  an  ambitious  possibility.  Perhaps 
best  of  all  such  a  release  should  be  into  some  large  enclosure 
system  initially  free  of  a  larval  population.  Validation  experi- 
ments must  also  test  the  more  unusual  environmental  condi- 


tions which  apply  in  high  latitudes  where,  for  example,  daylight 
prevails  over  the  full  24  hours. 

Experimental  Work 

The  functional  anatomy  approach  to  taxonomy  so  elegantly 
described  in  a  recent  review  by  Moser  (1981)  shows  the  extent 
to  which  structure  can  be  used  to  deduce  function.  The  inter- 
action of  this  approach  with  that  of  the  experimentalist  has 
yielded  much  useful  information. 

Since  the  1950's  increasing  success  in  rearing  marine  fish 
larvae  may  have  provided  the  taxonomists  with  help  as  well  as 
some  doubts  as  described  in  the  last  section.  It  has  also  led  to 
a  wide  literature  on  the  physiology,  behaviour  and  physiological 
ecology  of  larvae  (and  the  use  of  larvae  in  pollutant  bioassay) 
as  biologists  seized  the  opportunity  to  exploit  such  new  and 
valuable  material.  Perhaps  the  most  credit  should  be  given  to 
Shelboume  (1964)  for  his  extensive  and  painstaking  rearing  ex- 
periments on  plaice,  and  later  sole,  at  Port  Erin,  Isle  of  Man. 
These  experiments  undoubtedly  led  to  the  present  wide  practice 
of  marine  finfish  aquaculture  with  the  expanding  commercial 
use  of  turbot,  sole,  bass,  bream  and  gilthead. 

Rearing  may  still  be  considered  as  something  of  an  art  and 
is  often  most  successful  in  the  hands  of  dedicated  people  with 
a  "feel"  for  what  is  right  or  wrong.  Undoubtedly  a  breakthrough 
was  made  in  finding  suitable  food  for  larvae.  It  is  significant 
that  both  plaice  and  sole  can  take  Anemia  nauplii  from  first 
feeding  as  can  some  races  of  herring.  This  resulted  in  another 
U.K.  focus  for  rearing  at  Aberdeen,  and  later  Oban,  developed 
by  Blaxter  (1968)  on  the  herring.  Species  with  smaller  larvae 
(with  smaller  mouths)  were  only  successfully  reared  when  Las- 
ker's  group  at  La  Jolla  (Lasker  et  al.,  1970;  Theilacker  and 
McMaster,  1971;  Hunter,  1976)  developed  the  use  of  the  rotifer 
Brachionus plicatilis  and  the  naked  dinoflagellate  Gymnodmium 
splendens  as  small  food  items  for  early-stage  larvae  of  species 
like  northern  anchovy  and  jack  mackerel.  About  the  same  time 
Howell  (1973)  also  used  Brachionus  to  rear  turbot  larvae  at  Port 
Erin. 

Subsequently  a  number  of  factors  have  been  identified  to  add 
to  our  corpus  of  knowledge  on  rearing.  These  include  the  need 
for  good  water  quality,  with  the  interesting  idea  of  "green  water" 
culture  of  larvae  in  fairly  high  densitiesofC/j/oreZ/a  which  seems 
to  damp  out  fluctuations  in  metabolites,  and  perhaps  enhance 
oxygenation  as  well  as  providing  secondary  feeding  for  the  larvae 
(e.g.,  Houde,  1977;  Morita,  1984).  Adequate  light  for  visually- 
feeding  larvae  and  the  need  to  prevent  excessive  bunching  of 
larvae  or  their  prey  are  also  important,  as  is  the  quality  of  the 
food.  Success  or  failure  may  now  depend  on  the  fatty-acid  profile 
of  the  Anemia  nauplii  which  are  still  used  by  most  workers  in 
the  later  stages  of  rearing.  Artificial  diets  of  encapsulated  or 
particulate  food  are  also  being  developed  but  have  yet  to  be 
introduced  as  a  standard  technique  for  early  rearing. 

Before  turning  to  the  extrapolation  and  application  of  exper- 
imental data  to  modelling,  mention  must  be  made  of  Haydock's 
(1971)  and  Leong's  (1971)  work  on  the  induction  of  spawning 
in  the  croaker  and  anchovy  by  pre-treatment  with  an  appro- 
priate photoperiod  followed  by  hormone  injection.  This  has 
been  applied  subsequently  to  the  menhaden  by  Hettler  (1981), 
and  to  many  other  species,  and  has  become  a  standard  method 
for  workers  requiring  eggs  over  long  periods  or  at  a  specific  time. 

We  now  have  the  widest  knowledge  of  the  development,  be- 
haviour and  physiology  of  both  anchovy  and  herring  larvae  (see 
Fig.  2)  but  there  are  several  species  such  as  cod,  jack  mackerel, 
mackerel,  plaice  and  turbot  which  run  them  a  close  second. 


BLAXTER:  ONTOGENY,  SYSTEMATICS,  FISHERIES 


Lateral  line 


Respiration 


Red  muscle 


Reynolds  number 
(Re)  and 


hydrodynamic 


Viscous 


regimes 
Digestive  tract 


Re<lO 


Time  to  50% 
starvation 


Larval  period 


First  feeding 


"^ 1 

Photopic  vtslon         I 

Lens  retractor  muscle 
Improved  accommodation^  —  " 

^ *-'- 


Threshold 

tof 
schooling 


Initial  swim  bladder  Inflation 
Olel  vertical  movements 


,  ---^T- 


Functional  eye  First  rods 


Increase  in  number  of  neuromasts 


Scotopic  vision 


Rod  recruitment  continues 


Many  rods 


Canal  formation 


First        Epidermis 
RBC's      thickens 


Many 
RBC  s 


Cutaneous  respiration 


Superficial 
1   layer 


Transition, 
"?0<R'e'<2°° 

Functional  gut 


Movable  lower  Jaw 


2.5  days 

:t 


I      I 
5 


Gill  respiration 


Scale  formation 


Midline 

2-3  layers       3-4  layers 


7-8  layers 


Re>200 


Stomach  forms 
.It- 


Expandable  mouth 
3.3  days  4  days 


T^ 1 1 1 — I 1 1 1 1 1 1 1 1 1 1 1 — 1- 

15  20  25  30 

Length  (mm) 


Juvenile 
period 


Filler  feeding 


15  days 

* 


1 — I 1 1 1 1 1 — 

2.5  5   10    15     20     25     30 


35      40      45 

Days  at  16°  C 


— 1 1 1 1 1 1 1 — 

50     55     60     65    70    75    80 


Fig.  2.     Events  during  development  of  the  northern  anchovy.  RBC  =  red  blood  cells.  Time  to  50%  starvation  is  number  of  days  to  starvation 
at  which  50%  of  the  fish  died  (from  Hunter  and  Coyne.  1982). 


Much  of  this  work  is  summarised  by  Theilacker  and  Dorsey 
(1980). 

Over  the  past  few  years  the  assembly  of  much  basic  data  has 
allowed  the  current  vogue  for  modelling  to  be  applied  to  fish 
larvae.  Modelling  is  an  attempt  to  synthesise  and  simplify  basic 
data  usually  in  mathematical  form.  Mathematical  models  are 
often  iterative  and  they  have  the  value  of  being  in  a  form  suitable 
for  computers.  Laurence  (1981)  has  recently  reviewed  modelling 
work  on  fish  larvae  and  the  complexity  and  type  of  interaction 
is  shown  in  Fig.  3.  The  main  problem  addressed  has  been  that 
of  feeding.  The  earlier  models  of  Blaxter  (1966),  Rosenthal  and 
Hempel  ( 1 970),  Blaxter  and  Staines  ( 1 97  1 )  and  Hunter  (1972) 
estimated  the  feeding  efficiency  of  larvae,  the  volume  of  water 
searched  in  unit  time  and  the  density  of  food  required  to  give 
good  survival  and  growth.  More  sophisticated  models  have  now 
been  developed  (e.g.,  Jones  and  Hall,  1 974;  Beyer  and  Laurence, 
1981)  and  Vlymen's  (1977)  model  allows  for  the  prey  species 
being  non-randomly  distributed. 

The  need  for  larvae  and  their  prey  to  co-exist  temporally  was 
spelled  out  by  Gushing  ( 1 975)  in  his  match-mismatch  hypothesis. 
Thus  the  timing  of  reproduction  appears  to  have  evolved  to 
synchronise  the  larval  stages  with  the  main  phase  of  the  annual 
production  cycle.  Spawning  is  probably  controlled  in  most  tem- 
perate fish  species  by  photoperiod  and  temperature  which  are 
not  the  only  determinants  of  plankton  production.  Hence  a 
match  or  mismatch  is  possible  between  this  production  and  the 
presence  of  fish  larvae  with  a  resulting  influence  on  year  class 
strength. 


An  early  paradox  existed  in  that  the  density  of  the  larger 
micro-zooplankton  such  as  copepod  nauplii  required  for  good 
growth  and  survival  in  tanks  was  of  the  order  of  1  organism/ 
ml.  Such  densities  are  rarely  found  in  the  sea  as  judged  from 
normal  plankton  sampling.  This  led  to  the  suggestion  of  micro- 
scale  patchiness  of  food  in  the  sea,  which  might  occur  at  inter- 
faces such  as  steep  thermoclines  and  at  tide-  and  wind-induced 
fronts.  The  integrity  of  such  microscale  patchiness  would  not, 
of  course,  be  obvious  using  nets  sampling  large  volumes. 

This  led  Lasker  (1975)  to  bioassay  samples  of  water  taken  at 
different  depths  and  places  off  the  Califomian  coast,  using  an- 
chovy larvae  both  hatched  and  tested  on  board  ship.  Chloro- 
phyll-rich layers  with  very  high  densities  oi  Gymnodinium  were 
found  near  the  thermocline.  The  bioassay  showed  good  larval 
feeding  in  these  water  samples,  suggesting  that  patchiness,  in- 
deed, might  be  a  valid  concept.  This  was  to  some  extent  con- 
firmed by  later  findings  that  stable  weather  conditions  (which 
maintained  the  thermocline)  favoured  good  year  classes  of  an- 
chovy larvae  off  the  Califomian  coast  (Lasker,  1981).  Owen 
( 1 980)  has  subsequently  shown  from  samples  taken  by  plankton 
pumps  and  water  bottles  that  patchiness  of  microzooplankton 
such  as  copepod  nauplii  and  tintinnids  and  various  protozoan 
species  and  phytoplankton  (some  of  which  are  known  to  be  the 
food  of  anchovy  larvae)  exist  off  the  Peruvian  and  Califomian 
coasts  on  the  scale  of  a  few  centimetres  up  to  one  metre  (see 
Fig.  4).  Only  Houde  and  Schekter  ( 1978)  have  attempted  to  rear 
larvae  in  simulated  food  patches  and  found  that  survival  of  sea 
bream  was  similar  when  they  were  exposed  to  3  h  of  food  per 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


PHYSICAL  A  CHEMICAL  INFLUENCES 

CIRCULATION 
DIFFdSlON 

80UNDARV  EXCHANGE 
DISCONTINUITY  EVENTS  (STORMSj 
PRIMARY  PRODUCTION  & 

ORGANIC  RECYCLING 
POLLL  TION  OR  TOXICITY 
ABIOTIC  FACTORS  (TEMPERATURE 

SALINITY.  OXYGEN) 


IMPORTANT  I  ARVAL  FISH  INTERACTIONS 


LARVAl 
PREDATORS 


MOSTLY 
UNIDENTIFIED 


RECRUITMENT 
ASSESSMENTS 

USHERY  MODELS 

ECOSYSTEM  MODELS 

MANAGEMENT  STRATEGY 


LARVAL  MORTALITY  DETERMINED  DIRECTLY  IN  LINKAGES  A  A  B 


o 
o 


tl'RREMLY  rWPHASlZED  STL:nrFS 
STtlDIES  FOR  EMPHASIS 


SI  I   DthS  ME   IMMFniAlE 
I  ESSEk   IMPORT  ASCE 


DIRECl   HACnONAI    IINKAGES 
I  ISKAGES  OE  I  ESSER  IMPORTANCE 


Fig.  3.     A  generalised  scheme  for  the  main  interactions  between  larval  fish  and  their  biotic  and  abiotic  environment,  providing  a  basis  for 
modelling  (from  Laurence,  1981). 


day  as  when  fed  at  the  same  food  level  continuously.  Clearly, 
expeiiments  need  to  be  devised  to  test  the  effect  of  spatial  rather 
than  temporal  food  patchiness. 

The  evidence  is  thus  accumulating,  but  very  slowly,  that  lai^al 
survival  may  depend  on  the  extent  and  stability  of  microscale 
food  patches  or  interfaces,  at  least  in  some  areas.  It  may  be  that 
the  rather  high  food  densities  required  in  small-scale  tank  rear- 
ing do  indeed  apply  to  conditions  in  the  sea  and  that  such 
densities  are  only  found  in  patches. 

SCALING-UP 

Two  major  areas  may  be  identified  where  rearing  work  has 
been  extended  into  large-scale  containers.  The  first  of  these  are 
the  large  onshore  enclosures  and  embayments  used  by  the  pres- 
ent generation  of  Norwegian  biologists;  the  second  are  the  deep- 
water  plastic  bags  used  by  Scottish  workers  in  Loch  Ewe  on  the 
Scottish  West  Coast.  The  Norwegians  have  achieved  remarkable 
growth  and  survival  rates  for  herring  and  cod  larvae,  as  high  as 
30-70%  survival  from  hatching  to  metamorphosis,  in  shallow 
4,000-60,000  m^  enclosures  (Oieslad  and  Moksness,  1981; 
Kvenseth  and  Oiestad,  1984).  The  Loch  Ewe  bags,  which  are 
deep  cylinders,  of  about  300  m\  have  been  used  for  rearing 
herring  and  cod,  but  with  much  less  success  than  the  Norwegians 
(Gamble  et  al.,  1981;  Gamble  and  Houde,  1984). 

Possibly  volume  itself  is  important,  or  more  likely  the  ratio 
between  volume  and  wall  area.  The  interface  between  wall  and 
sea  water  is  not  a  natural  one  for  fish  larvae,  feeding  may  be 
difficult  al  the  interface,  and  food  may  aggregate  there  in  an 
inaccessible  form.  Morita  (1984)  reports  that  Pacific  herring 
larvae  have  recently  been  reared  in  20  m'  tanks  with  a  46% 
survival  from  hatching  to  a  mean  length  of  about  7  cm  in  1 1 2 
days.  This  spectacular  result  may  have  been  partly  a  feature  of 
a  fairly  large  onshore  tank  but  also  the  "green  water"  technique 


mentioned  earlier.  Hunter  (1984)  suggests  that  the  high  survival 
in  some  large  tank  or  enclosure  experiments  is  achieved  by  the 
elimination  of  predators.  To  the  present  author  a  combination 
of  optimal  feeding  conditions  and  low  predation  seems  to  be 
the  likely  cause. 

The  events  have  been  described  so  far  in  a  topsy-turvy  way, 
in  that  sea  surveys  have  always  been  the  most  widely-adopted 
approach  to  problems  associated  with  the  early  life  history  of 
fish.  The  experimental  and  enclosure  studies  are  the  icing  on 
the  research  cake,  although  both  Norwegian  and  Japanese  work- 
ers are  seriously  considering  the  possibility  of  restocking  de- 
pleted inshore  fisheries  or  topping-up  poor  year-classes  of  cod 
and  herring  by  releasing  reared  late-stage  larvae  or  O-group 
juveniles. 

Sea  Surveys 

These  are  expensive  in  terms  of  ship-time  and  manpower. 
Originally  designed  to  advance  our  knowledge  of  spawning 
grounds,  larval  drift,  and  horizontal  and  vertical  distribution, 
they  are  often  now  linked  to  more  practical  aims.  Nevertheless, 
superb  time-series  exist  for  areas  like  the  California  Current  and 
North  Sea  as  a  result  of  the  patience  and  foresight  of  earlier 
workers  like  Ahlstrom  and  later  workers  like  Smith  and  Saville 
(see  review  by  Smith  and  Richardson,  1977).  Sea  surveys  have 
always  been  a  rich  ground  for  innovative  science,  in  terms  of 
sampling  techniques,  interpretation  and  usage.  Experimenters 
and  modellers  have  provided  a  great  boost  for  this  work,  allow- 
ing new  interpretations  to  be  made  and  new  hypotheses  to  be 
tested. 

No  more  mention  will  be  made  of  the  matrix-filling  role  of 
sea  surveys— namely  the  completion  of  details  of  life  history, 
which  is  still  taking  place  and  has  been  much  aided  by  the  vast 
improvement  in  egg  and  larval  identification  in  the  past  two 


BLAXTER:  ONTOGENY,  SYSTEMATICS,  FISHERIES 


CONCENTRATION  (no/X) 


20  0 

D                        1000                    2000 

3000 

<■ 

^ 

X 

J 

20  4 

"-^^ 

J 

e 

Prorocentrum  —          ^  ^ — ■C'     ' 

_-^-^ 

^ 

I 

?0  8 

J^^    \             — 

-Nit/schia 

1— 

—  "" "  "            ^ — \ 

a. 

-             <                                          J^ 

UJ 

Q 

212 

f^"^'^"^"'--- 

V           L____ 

"jt 

216 

PHYTOPLANKTON 

25 


20  0 


20  4 


20  8 


21   2 


CONCENTRATION  (no /i) 

50  75 


100 


216- 


Fig.  4.  Vanation  in  concentration  of  microplankton  in  samples  from  20  cm  depth  intervals  in  the  chlorophyll  maximum  layer  over  the  coastal 
shelf  of  the  Southern  California  Bight  dunng  March.  1976.  Prorocentrum,  tintinnids  and  copepod  nauplii  are  all  food  items  for  larval  anchovy 
(from  Owen.  1980). 


- 

• •r- 

) 

Tintinn 

r 

t 

{ Nauplior 

s-^r         copepods 

/ 

- — Noctiluca 

MICRO-ZOOPLANKTON 

decades.  Improvement  in  plankton  nets  and  young  fish  trawls 
means  that  vertical  profiling  and  quantitative  sampling  have 
finally  come-of-age.  This  ability  to  sample  quantitatively  is  the 
single  most  important  advance  in  allowing  larval  populations 
to  be  assessed  reliably  and  for  allowing  models  to  be  tested.  The 
outcome  is  two-fold.  The  door  is  open  for  biomass  estimates  of 
spawning  stock  from  egg  and  larval  surveys  and  for  testing  the 
possible  factors  in  the  stock-recruitment  relationship.  Each  of 
these  will  be  considered  in  the  final  part  of  this  paper. 

/.  Biomass  estimation.  — ¥or  many  years  population  dynami- 
cists  lacked  good  information  on  the  absolute  size  of  the  spawn- 
ing stock  and  regulation  was  largely  achieved  by  minimum  mesh 
and  landing  sizes.  Of  late,  as  a  result  of  catastrophic  declines  in 
some  species,  whole  fisheries  have  been  closed  or  controlled  by 
quotas  and  total  allowable  catch  (TAC).  The  use  of  TAC's  has 
been  greatly  aided  by  virtual  population  analysis  and  also  by 
sonar-based  fish  counting  surveys;  these  give  an  estimate  of  total 
stock  size,  the  reliability  of  which  depends  on  the  extent  of  the 
survey,  the  ability  to  identify  the  species  in  question  and  the 
precision  of  the  calibration  of  target  strength. 

To  supplement  the  results,  estimates  of  spawning  stock  size 
have  been  made  on  an  ad  hoc  basis  by  counting  eggs  and  larvae 
and  converting  them  into  the  parental  spawning  stock  biomass 
by  a  knowledge  of  fecundity,  age  distribution  and  sex  ratio.  Some 
of  the  pioneering  work  was  done  by  Sette  and  Ahlstrom  (1948) 
on  Califomian  pilchard  and  Simpson  (1959)  on  North  Sea  plaice. 
Saville,  Baxter  and  McKay  (1974)  counted  the  demersal  eggs  of 
the  herring  on  the  small  spawning  ground  of  Ballantrae  Bank 
in  the  Clyde.  This  was  later  extended  by  Saville  and  McKay 
(see  Saville,  1981)  to  herring  larval  surveys  in  the  North  Sea 
and  off  the  Scottish  west  coast.  The  biomass  of  Pacific  hemng 
is  now  routinely  assessed  from  the  intertidal  egg  deposition  along 
the  coast  of  Canada  and  the  USA  as  described  in  the  recent 
Nanaimo  Herring  Symposium  (Hay,  1984;  Haegele  and 
Schweigert,  1984).  Similar,  but  ad  hoc.  data  are  available  for 
the  northern  anchovy  from  the  work  of  Smith  (1972),  Parker 
( 1 980)  and  Picquelle  and  Hewitt  (1983),  for  the  Atlantic  mack- 
erel from  Lockwood,  Nichols  and  Dawson  (1981)  and  Berrien, 
Naplin  and  Pennington  (1981)  and  for  North  Sea  cod  from  Daan 


(1981).  Some  of  these  data  give  absolute  measures,  some  relative 
ones  from  year-to-year,  often  related  to  biomass  estimates  by 
other  means. 

This  survey  technique  has  notable  disadvantages.  It  must  be 
done  at  a  limited  time  of  year  and  is  obviously  easiest  to  interpret 
for  one-off  spawners.  The  survey  must  be  done  rapidly  and  as 
near  the  spawning  season  as  possible  to  overcome  any  errors 
caused  by  mortality  between  spawning  and  sampling.  Although 
it  can  be  applied  to  a  closed  fishery,  the  age  structure  of  the 
population  is  required  to  compute  the  aggregate  fecundity,  hence 
scientific  sampling  of  the  adults  is  required. 


2.  Stock-recruitment.— The  relationship  between  the  size  of  the 
spawning  stock  in  any  year  and  the  number  of  recruits  it  supplies 
to  the  fishery  subsequently  is  vital  information  for  the  regulation 
of  fisheries.  This  is  specially  true  where  recruitment  overfishing 
is  prevalent  as  in  the  clupeoids.  Over  many  years  a  stock-re- 
cruitment relationship  may  be  obtained  empirically  in  any  fish- 
ery, but  this  is  time-consuming  and  usually  contains  inexplicable 
features.  While,  as  might  be  expected,  low  spawning  stock  leads 
to  low  recruitment,  high  spawning  stocks  may  also  give  unex- 
pectedly low  recruitment,  as  the  result  of  density-dependent 
effects.  Alternatively  spawning  stocks  of  a  given  size  can  yield 
enormously  different  brood  strengths,  of  the  order  of  10-100 
times,  in  a  quite  unpredictable  way. 

It  is  not  surprising  that  the  underlying  causes  of  the  control 
of  brood  strength  are  of  much  interest  to  fisheries  biologists  and 
have  received  the  attention  of  experimentalists  and  modellers. 
Most  marine  fish  have  a  very  high  fecundity,  of  the  orders  of 
tens  of  thousands  to  a  few  million.  From  such  a  starting  point 
mortality  must  be  very  high  and  it  is  surprising  that  brood 
strength  variations  are  not  even  more  variable  than  is  actually 
the  case.  What  then  do  we  know  of  the  mortality  rate  of  eggs 
and  larvae  in  the  sea?  Are  there  critical  periods  when  it  is  es- 
pecially high?  What  are  the  causes  of  mortality? 

Hjort's  original  hypothesis,  now  some  70  years  old,  expressed 
the  view  that  a  critical  period  existed  after  yolk  resorption  as 
the  larvae  sought  external  food  sources.  This  hypothesis  was 
supported  by  earlier  rearing  experiments  in  which  very  high 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


mortalities  occurred  at  first  feeding.  Measurements  of  mortality 
rates  of  eggs  and  larvae  at  sea  tend  to  show  a  high  but  continuing 
mortality  of  perhaps  5-20%  per  day.  The  results  of  sea  surveys 
are,  however,  often  difficult  to  interpret  because  of  the  need  to 
sample  within  a  discrete  larval  population  over  a  long  time. 
May  (1974),  in  his  review  of  this  subject,  concluded  that  star- 
vation at  the  end  of  the  yolk-sac  stage  may  often  have  a  major 
influence  on  brood  strength  but  that  mortality  from  fertilization 
to  the  O-group  stage  is  the  ultimate  determinant. 

The  results  of  modelling  and  the  tests  of  the  patchiness  hy- 
pothesis which  have  already  been  discussed  support  the  idea 
that  first  feeding  is  a  critical  time,  although  not  having,  neces- 
sarily, the  dominant  effect  claimed  by  Hjort.  Experimenters  and 
modellers  have  also  derived  further  concepts  for  testing.  The 
major  sources  of  mortality  are  identified  as  starvation  and  pre- 
dation.  Starvation,  of  course,  only  operates  from  the  end  of  the 
yolk-sac  stage.  Blaxter  and  Hempel  (1963)  used  the  expression 
"point-of-no-retum"  to  express  the  point  at  which  larvae,  as  a 
result  of  starvation,  are  too  weak  to  feed  even  if  food  becomes 
available.  Sometimes  called  "ecological  death"  or  "irreversible 
starvation"  this  is  a  useful  concept  for  assessing  the  chances  of 
larval  survival  under  different  conditions.  For  larvae  in  a  good 
nutritional  state  the  time  to  the  point-of-no-retum  may  be  only 
1-2  days  in  a  small  newly  feeding  larva  like  the  anchovy,  but 
2-3  weeks  in  a  well  grown  flatfish  larva  like  the  plaice  (see 
Theilacker  and  Dorsey,  1980).  Implicit,  also,  in  the  concept  is 
that  larvae  can  live  for  some  time  after  the  point-of-no-retum. 
During  this  time  they  may  be  especially  liable  to  capture  by  nets 
and,  without  adequate  knowledge,  a  false  impression  might  be 
obtained  of  the  size  or  nutritional  state  of  the  larval  population. 

The  assessment  of  nutritional  state  of  larvae  has  been  of  wide 
interest  in  recent  years,  in  the  hope  of  relating  this  to  brood 
strength.  Initially  Blaxter  ( 1 965)  measured  the  condition  factors 
of  tank-reared  herring  larvae  after  varying  periods  of  starvation 
and  then  later  compared  the  results  with  the  condition  factors 
of  sea-caught  herring  larvae  (Blaxter,  1971).  It  was  found  that 
most  sea-caught  larvae  had  much  lower  condition  factors  than 
starving  tank-reared  larvae  and  it  became  apparent  that  the 
extrapolation  of  tank  criteria  to  the  sea  was  invalid  because  the 
tank  larvae  were  short  and  fat  compared  with  wild  larvae  (see 
Fig.  1).  This  means  that  condition  factor  comparisons  of  wild 
larvae  are  only  valid  on  a  relative  basis  from  year-to-year  or 
place-to-place  (e.g.,  Chenoweth,  1970;  Vilela  and  Zijlstra,  1971) 
and  only  then  if  one  can  be  satisfied  that  shrinkage  after  capture 
is  consistent.  The  problems  of  tank ;  sea  comparisons  and 
shrinkage  are  unfortunately  likely  to  be  the  most  serious  in  long 
clupeoid  larvae  to  which  these  experiments  have  been  applied. 
No  one  has  checked  their  validity  in  the  more  common  type  of 
larvae  with  a  shorter  body  form. 

These  problems  led  to  work  at  Oban  and  La  Jolla  on  histo- 
logical criteria  for  assessing  starvation  (Ehrlich  et  al.,  1976; 
O'Connell,  1976;  Theilacker,  1978).  O'Connelfs  work  on  an- 
chovy larvae  deserves  special  mention.  He  found  from  screening 
the  state  of  the  body  organs  such  as  pancreas  and  gut  that  these 
showed  increasing  signs  of  degeneration  as  starvation  pro- 
ceeded. On  applying  his  criteria  to  sea-caught  anchovy  larvae 
O'Connell  (1981b)  found  evidence  for  quite  a  high  percentage 
of  larvae  suffering  from  advanced  starvation  and  considerable 
differences  in  the  incidence  of  starvation  in  closely  adjacent 
areas.  This  method  is  now  being  applied  by  Theilacker  on  jack 
mackerel  larvae  from  year-to-year  and  is  likely  to  be  adopted 
on  a  routine  basis. 


The  other  cause  of  mortality,  predation,  has  recently  become 
fashionable  following  the  work  of  Eraser,  Lasker,  Lillelund  and 
Theilacker  and  subsequently  Kuhlmann,  von  Westemhagen  and 
Rosenthal,  Bailey,  Purcell  and  several  other  workers  (See  re- 
views of  Hunter,  1981,  1984).  Copepods,  euphausiids,  amphi- 
pods  and  chaetognaths  are  all  implicated  but  perhaps  medusae 
are  the  most  voracious  group  of  predators  (Bailey  and  Batty, 
1983),  especially  for  inshore  spawners  like  Pacific  herring.  Pre- 
dation, of  course,  operates  from  the  moment  of  spawning  and 
Hunter  and  Kimbrell(  1980)  and  MacCall  (1980),  in  particular, 
have  discussed  the  incidence  of  density-dependent  cannibalism 
of  spawning  anchovies  on  their  own  eggs  and  larvae.  It  is  gen- 
erally thought  that  strong  selection  pressure  exists  for  fast  growth 
which  will  take  larvae  speedily  through  the  more  vulnerable 
early  stages.  Larvae  have  been  shown  experimentally  to  be  less 
vulnerable  when  they  are  larger,  their  escape  speeds  are  higher 
and  their  recovery  from  a  predator  attack  (for  predators  of  a 
given  size)  more  likely.  As  Hickey  (1979,  1982)  has  shown,  an 
efficient  wound-healing  mechanism  exists,  allowing  larvae  to 
recover  from  bites,  stings  and  other  forms  of  damage.  The  high 
survival  rates  of  larvae  reared  in  the  absence  of  predators  (Kven- 
seth  and  Oiestad,  1984;  Morita,  1984)  suggest  strongly  that 
predation  is  a  major  source  of  mortality  in  the  sea.  Although  it 
is  difficult  to  assess  the  relative  importance  of  starvation  and 
mortality  in  any  larval  population,  it  is  also  clear  that  the  two 
must  interact  in  the  sense  that  starving  larvae  will  be  more 
susceptible  to  predation. 

The  Future 

In  this  paper  modelling  has  been  only  briefly  discussed.  The 
method  is  now  widely  used  for  setting  up  hypotheses  about 
feeding,  starvation,  predation,  cannibalism  and  other  factors 
associated  with  the  stock-recruitment  relationship  and  biomass 
estimation.  This  approach  is  likely  to  continue  as  a  basis  for 
sea  surveys.  It  seems  uncertain  whether  biomass  will  be  routinely 
estimated  by  egg  and  larval  surveys  except  perhaps  in  Pacific 
herring  and  northern  anchovy.  The  cost  is  too  high  and  sonar 
surveys,  if  the  problems  can  be  ironed  out,  seem  to  be  a  better 
bet. 

Experimental  data  on  predation  still  need  to  be  collected  and 
few  correlations  exist  between  predator  populations  and  egg  and 
larval  mortality  in  the  sea.  In  fact  mortality  studies  on  eggs  and 
larvae  in  the  sea  in  general  need  to  be  perfected  since  the  prob- 
lems of  following  discrete  populations  and  of  ageing  larvae  are 
still  not  fully  solved.  At  least  one  source  of  information  is  largely 
untapped  and  that  is  the  explanation  for  the  high  survival  rates 
of  larvae  in  large  enclosures.  In  particular  the  distribution  of 
the  larvae  and  their  food  in  these  enclosures  is  not  known  and 
may  throw  light  on  the  validity  of  the  patchiness  hypothesis. 
Information  on  frontal  systems,  and  interfaces  as  a  result  of  tide, 
wind,  upwelling  and  thermo— and  halo— clines  is  now  quickly 
being  assembled  by  hydrographers  and  marine  biologists.  The 
larval  biologists  should  be  ready  to  exploit  the  results. 

It  will  be  apparent  to  the  audience  how  far  research  into  the 
early  life  history  of  fish  has  advanced  in  the  last  30  years.  A 
major  force  has  been  the  work  off"the  Califomian  coast  generated 
by  Ahlstrom  and  his  recruits  at  La  Jolla.  It  is  therefore  very 
fitting  that  this  symposium  should  be  dedicated  to  his  memory. 

Scottish  Marine  Biological  Association,  Dunstaffnage 
Marine  Research  Laboratory,  P.O.  Box  3,  Oban, 
Argyll,  Scotland. 


Ontogeny,  Systematics,  and  Phylogeny 
D.  M.  Cohen 


THE  work  of  Ahlie  and  his  students  and  colleagues  has  brought 
to  the  fore  great  amounts  of  descriptive  information  about 
the  early  life  history  (ELH)  stages  of  fishes  gathered  over  many 
years.  These  data  are  of  broad  provenance,  many  being  the 
results  of  original  research  by  the  Ahlstrom  school,  others  being 
taken  from  the  literature.  Only  a  scientist  with  Ahlie's  capabil- 
ities—an extensive  knowledge  of  fishes  and  their  ontogeny,  a 
fine  sense  of  order  in  nature,  and  a  critical  intellect— could  per- 
ceive pattern  in  the  bewildering  diversity  represented  by  the 
early  life  history  stages  of  fishes.  As  would  any  good  scientist, 
Ahlie  questioned  the  meaning  of  these  patterns,  and  it  is  chiefly 
to  further  this  inquiry  that  this  symposium  was  convened. 

Most  students  of  comparative  fish  ontogeny  know  more  about 
adult  fishes  than  ichthyologists  who  study  adults  know  about 
larval  fishes;  they  have  to.  Ahlie  stated  in  his  lectures.  "Larval 
taxonomy  is  just  an  adjunct  to  adult  taxonomy  and  you  have 
to  start  with  the  adults  to  know  the  larvae."  Early  on  he  dis- 
covered that  data  from  early  life  history  studies  did  not  always 
confirm  classifications  based  on  adults  alone.  We  all  want  to 
know  which  data  sets  most  closely  approximate  phylogenetic 
relationships;  how  apparent  conflicts  best  can  be  resolved;  how 
the  data  of  ontogeny  can  be  integrated  into  the  overall  field  of 
fish  systematics?  Answering  these  questions  is  not  easy,  espe- 
cially within  the  framework  dictated  by  the  widespread  adoption 
of  new  methodologies  in  systematics,  which  claim  to  require 
more  stringent  evaluation  of  characters  than  has  been  heretofore 
customary.  Many  traditional  character  suites  are  being  rejected 
for  purposes  of  elucidating  phylogenies,  and  new  data  are  needed 
for  testing.  Our  purposes  m  this  volume  are  to  state  the  bases 
for  what  has  come  to  be  called  larval  fish  taxonomy  and  to 
consider  the  systematics  of  various  groups  of  fishes  in  terms  of 
the  rich  and  virtually  untapped  store  of  data  offered  by  the  study 
of  early  life  history  stages. 

My  own  objectives  in  the  present  paper  are  several.  First  of 
all.  I  want  to  indicate  the  reasons,  some  obvious,  some  not,  for 
the  nearly  exclusive  use  of  adult  fishes  in  systematics,  which  has 
prevailed  until  very  recently.  Secondly,  I  will  briefly  discuss  the 
conceptual  and  methodological  framework  of  classification 
within  which  early  life  history  data  is  being  used.  Finally,  I  will 
comment  on  the  possible  importance  of  early  life  history  data 
for  the  study  of  phylogeny  with  special  reference  to  fishes. 

Why  Has  There  Been  So  Little  Use  of 
ELH  Stages  in  Fish  Systematics? 

The  fact  that  most  fish  classifications  are  based  entirely  or 
chiefly  on  the  structure  of  adults  was  a  source  of  concern  to 
Ahlie  and  remains  so  to  many  of  us,  although  this  Symposium 
is  an  indication  of  positive  change.  I  discuss  below  what  may 
be  some  of  the  reasons  for  a  long  preoccupation  with  adults. 

In  the  first  place,  zoologists  have  been  studying  adults  for  a 
longer  period  of  time  than  they  have  early  life  history  stages. 
Although  the  dim  beginnings  of  classification  are  often  placed 
with  Aristotle,  it  was  the  great  naturalists  Aldrovandi.  Belon. 
Gesner.  and  Rondelet  who  in  their  cataloging  of  nature  provided 
our  earliest  adult  fish  classifications.  Several  technological  de- 
siderata would  have  prevented  the  study  of  early  life  history 
stages  during  the  1 6th  century  when  these  early  scientists  were 


at  work.  Even  though  lenses  had  been  known  for  a  long  time, 
appropriate  microscopes  were  not  invented  until  the  1 7th  and 
18th  centuries  (Singer,  1959)  when  another  requisite  advance 
occurred,  the  use  of  alcohol  and  other  fluids  as  a  preservative 
for  zoological  specimens  (Singer,  1950).  Techniques  for  clearing 
flesh  and  staining  bone  and  cartilage  are  modem  acquisitions, 
as  is  the  use  of  x-ray  photographs  (Ahlstrom  and  Moser.  1981). 
The  invention  of  fine-mesh  towing  nets  did  not  occur  until  1 846 
(Sverdrup.  Johnson,  and  Fleming,  1942),  deferring  until  rela- 
tively recent  times  the  availability  of  suitable  collections  of  early 
life  history  stages  for  scientific  study. 

The  rearing  of  early  stages  is  another  valuable  component  of 
the  study  of  larval  fish  taxonomy,  and  although  fish  culture  is 
an  ancient  art,  the  staging  of  fry  and  their  preservation  and 
microscopic  study  is  technology-dependent  and  relatively  re- 
cent. 

Lack  of  information  on  metamorphosis  or  of  congruence  of 
larval  and  adult  stages  has  also  delayed  the  adoption  of  early 
life  history  stages  information  into  classification  schemes.  Of 
course  not  many  kinds  of  fishes  demonstrate  an  ontogenetic 
change  as  sudden  and  dramatic  as  do  the  eels,  but  the  fact  that 
this  particular  transformation  was  not  described  until  1897 
(Grassi  and  Calandruccio)  indicates  the  long  advance  start  held 
by  the  use  of  adult  stages.  Even  more  recent  have  been  discovery 
of  the  Anoplogaster-Caulolepis  relationship  (Grey,  1955a),  the 
Gibberichthys-Kasidoron  relationship  (de  Sylva  and  Eschmeyer, 
1977),  the  Giganturidae-Rosauridae  relationship  (Johnson,  this 
volume),  and  the  as-yet-unpublished  identity  of  larval  forms 
such  as  Svetovidovia.  These  and  other  examples  are  described 
in  this  volume.  And  indeed,  even  when  the  study  of  the  devel- 
opmental biology  of  vertebrates  commenced,  early  emphasis  in 
the  mid- 18th  century  was  on  classical  embryology,  the  describ- 
ing of  processes  and  structures  rather  than  on  comparing  them 
(Rostand,  1964).  Not  until  the  early  years  of  the  present  century 
when  fishery  scientists  began  to  use  larval  fishes  in  their  inves- 
tigations of  commercial  species  and  required  identifications  were 
serious  efforts  made  to  compare  data  (Ahlstrom  and  Moser, 
1981). 

Until  Ahlie  commenced  his  now  famous  courses  on  larval 
fishes,  there  were  few  places  where  a  student  could  learn  about 
them;  hence,  there  are  only  rare  instances  of  attention  being 
paid  to  any  potential  value  they  might  have  in  solving  problems 
in  systematics.  By  now,  in  contrast,  there  are  courses  and  sem- 
inars available  in  a  number  of  universities  on  the  study  of  ELH 
stages  of  fishes. 

Another  phenomenon  that  I  believe  has  inhibited  the  use  of 
early  life  history  stages  in  fish  systematics  is  what  I  call  the 
curatorial  mind  set.  Many  curators  of  adult  fish  collections  are 
wary  of  microscopic  specimens  stored  in  vials.  Although  these 
collections  occupy  small  space,  their  maintenance  and  docu- 
mentation are  labor-intensive  and  their  use  is  foreign  to  most 
ichthyologists.  There  are  many  excellent  collections  of  larval 
fishes,  but  they  are  mostly  in  fishery,  environmental  and  marine 
biology  laboratories— organizations  that  have  no  institutional 
commitment  to  long-term  collection  storage.  Collections  that 
document  important  publications  or  have  potential  value  in 
systematics  should  ultimately  be  deposited  in  a  museum  that 


8 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


has  a  mandate  to  guarantee  long-term  archival  storage  and  easy 
access.  Several  such  institutions  that  presently  house  larval  fish- 
es or  are  willing  to  do  so  are  the  Zoological  Museum  of  the 
University  of  Copenhagen,  which  maintains  the  extensive 
worldwide  collections  taken  during  the  Dana  Expeditions,  as 
well  as  ones  documenting  the  earlier  classical  studies  on  larval 
fishes  by  Johannes  Schmidt  and  his  students,  the  Harvard  Mu- 
seum of  Comparative  Zoology,  the  Smithsonian  Institution,  and 
the  Natural  History  Museum  of  Los  Angeles  County.  If  collec- 
tions of  ELH  stages  are  to  realize  their  full  potential  in  system- 
atics,  then  it  is  timely  for  ichthyoplankton  specialists  to  offer 
good  developmental  series,  especially  illustrated  ones,  and  for 
museum  curators  to  accept  them. 

Fossils  have  been  studied  for  clues  to  the  major  classification 
of  fishes  since  the  days  of  Louis  Agassiz  (Patterson,  1981a)  and 
to  the  extent  that  they  were  available  have  been  widely  consid- 
ered as  important  adjuncts  or  indeed  prerequisites  to  compre- 
hending the  phylogeny  of  particular  groups.  Although  this  view 
is  now  receiving  heavy  criticism  (Patterson,  1981b),  the  fact 
remains  that  it  did  exist  for  many  years  and  may  have  detracted 
from  the  potential  contribution  of  the  non-fossil  suites  of  char- 
acters carried  by  early  life  history  stages.  Even  so,  students  of 
fossils  and  of  larvae  share  a  preoccupation  with  the  caudal  fin 
skeleton,  a  structure  that  is  often  well  preserved  in  fossils  and 
can  be  studied  in  two  dimensions  and  which,  during  the  course 
of  ontogeny,  exposes  a  wealth  of  information  of  great  value  to 
the  systematist. 

Because  adult  stages  have  been  the  chief  source  of  characters 
used  in  fish  systematics,  a  perception  has  arisen  that  these  char- 
acters are  in  some  way  more  useful  or  more  indicative  of  a 
phylogenetic  classification  than  are  the  characters  of  early  life 
history  stages.  How  did  such  a  view  arise?  For  many  years, 
systematists  tended  to  concentrate  on  the  search  for  conserva- 
tive, "non-adaptive"  characters  (labeled  the  Darwin  Principle 
by  Mayr,  1969).  They  discarded  not  only  ones  that  they  believed 
were  directly  affected  by  the  environment  but  also  ones  that 
appeared  to  smack  of  convergence.  It  seemed  reasonable  and 
proper,  for  example,  to  group  together  for  phylogenetic  purposes 
fishes  with  one  spine  and  five  soft  rays  in  the  pelvic  fin  because 
the  character  was  apparently  conservative,  non-adaptive,  and 
non-convergent.  On  the  other  hand,  it  seemed  wrong  to  group 
together  all  fishes  with  canine  teeth  because  the  character  was 
apparently  non-conservative,  adaptive,  and  surely  convergent. 
With  regard  to  larval  fishes,  Moser  (1981)  recently  discussed 
the  occurrence  of  a  large  number  of  apparently  highly  adaptive 
larval  characters  distributed  across  a  broad  taxonomic  spec- 
trum. He  states,  "Marine  teleost  larvae  have  evolved  an  enor- 
mous array  of  morphological  specializations,  such  that  it  seems 
to  me  we  are  looking  at  a  distinct  evolutionary  domain  quite 
separate  from  that  of  the  adults.  It  is  reasonable  to  assume  that 
these  remarkable  structural  specializations  are  adaptive  and  re- 
flect each  species'  solution  to  the  challenge  of  survival  in  a 
complex  and  demanding  environment."  My  point  here  is  that 
if  a  systematist  rejected  adaptive  characters  (and  many  did), 
then  he  would  have  been  unlikely  to  use  ELH  stages,  and  this 
may  be  another  reason  why  they  have  not  received  sufficient 
attention. 

How  Systematists  Do  Their  Work 

Even  if  systematists  agreed  among  themselves  about  their 
immediate  goals  and  how  best  to  achieve  them,  the  task  of  this 


Symposium  would  be  daunting.  But  contemporary  systematists 
do  not  agree  on  either  objectives  or  methodology.  The  concepts 
that  purport  to  link  systematics  to  phylogeny  are  being  actively 
reassessed,  and  it  is  within  the  context  of  rapidly  changing  ideas 
in  systematics  that  our  presentations  and  discussions  will  occur. 

There  are  basically  three  conceptual  methods  now  being  used 
by  systematists,  and  although  the  bare  bones  of  these  methods 
are  easily  comprehended,  in  practice  they  become  more  complex 
and  their  independence  from  each  other  less  clear.  The  interested 
reader  who  is  as  yet  unaware  of  the  intense  debate  both  between 
and  within  the  several  schools  of  systematic  classification  is 
referred  to  the  pages  of  the  journal  Syslonatic  Zoology  for  many 
articles  and  references  as  well  as  ones  cited  in  this  section.  A 
recent  description  and  comparison  of  the  three  methods  is  given 
by  Mayr  (1981),  who  lists  many  important  references.  Although 
1  do  not  propose  to  use  very  much  space  here  on  a  redundant 
treatment,  1  will  briefly  describe  each  method  and  comment  on 
its  strengths  and  weaknesses. 

The  theoretically  simplest  method  (or  methods— there  is  more 
than  one  algorithm,  and  there  is  disagreement  on  which  is  best) 
is  called  phenetics  or  numerical  taxonomy  and  is  described  in 
detail  by  Sokal  and  Sneath  (1963)  and  Sneath  and  Sokal  (1973). 
It  is  based  on  overall  similarity.  Many  unweighted  characters 
are  used  to  generate  clusters  of  OTUs  (operational  taxonomic 
units),  which  may  be  anything  from  individuals,  populations, 
or  species  to  orders,  classes,  or  phyla.  The  hierarchically  ar- 
ranged clusters,  which  lack  a  time  dimension,  are  called  phe- 
nograms.  Neither  homology  nor  the  fossil  record  are  considered 
in  selecting  characters.  Each  member  of  a  cluster  bears  a  closer 
resemblance,  although  not  necessarily  genealogical  relationship, 
to  other  members  of  its  cluster  than  it  does  to  members  of  other 
clusters.  Some  pheneticists  claim  that  if  a  sufficient  number  of 
characters  is  analyzed,  any  influence  of  convergence  becomes 
dampened  and  the  phenogram  will  express  phylogenetic  rela- 
tionships. Unfortunately,  there  seems  to  be  no  good  way  to 
ascertain  how  many  characters  are  needed.  Other  pheneticists 
do  not  ascribe  phylogenetic  significance  to  their  clusters  and 
merely  claim  to  be  representing  overall  similarity.  Replicability 
of  results  is  the  chief  objective.  Many  classifications  that  purport 
to  be  based  on  the  methods  of  cladistics  or  evolutionary  clas- 
sification, upon  close  scrutiny  appear  to  be  basically  phenetic. 
There  are  apparently  few  fish  classifications  using  ELH  char- 
acters, which  are  explicitly  based  on  phenetic  methods.  One 
example  is  a  paper  on  Northeast  Pacific  cottid  genera  (Rich- 
ardson, 1981a)  which,  according  to  the  author,  was  not  entirely 
satisfactory  for  phyletic  purposes.  Ichthyologists  who  restrict 
their  data  sources  for  a  phenetic  analysis  to  a  single  life  history 
stage  should  consider  a  study  by  Michener  (1977),  who  gener- 
ated four  different  phenetic  classifications  of  a  group  of  bees 
based  on  different  life  history  stages  or  character  suites. 

A  second  method  is  called  cladistics  or  phylogenetic  system- 
atics, and  although  it  has  been  more  or  less  on  the  scene  for 
many  years,  it  is  only  since  the  revision  and  translation  into 
English  of  its  original  presentation  (Hennig,  1950,  1966)  that  it 
has  gained  wide  currency  and  is  now  used,  either  explicitly  or 
implicitly,  by  many  systematic  ichthyologists  all  around  the 
world  but  particularly  in  North  America  and  western  Europe. 
A  recent  guide  to  the  method  is  a  book  by  Wiley  (1981),  and 
the  reader  is  advised  to  consult  also  Brundin  ( 1 966)  for  a  notably 
lucid  interpretation.  Cladistics  requires  a  stringent  evaluation 
of  characters.  Primitive  or  generalized  ones  (called  plesiomor- 


COHEN:  ONTOGENY,  SYSTEMATICS,  PHYLOGENY 


phic)  for  the  group  being  analyzed  are  discarded  for  purposes 
of  generating  a  phylogenetic  classification;  only  derived  char- 
acters (apomorphic)  are  of  value,  and  monophyletic  groups  are 
defined  by  the  degree  to  which  they  share  such  characters  (syn- 
apomorphy).  The  distribution  of  derived  character  states  among 
a  monophyletic  assemblage  of  taxa  is  analyzed  and  used  to 
generate  an  hierarchically  arranged  chart  called  a  cladogram,  in 
which  each  node  or  branching  point  on  the  diagram  gives  rise 
to  two  branches  that  are  interpreted  as  genealogical  lineages  and 
are  called  sister  groups.  In  instances  in  which  the  data  do  not 
allow  the  unambiguous  definition  of  two  branches,  more  are 
often  used.  Each  member  of  a  monophyletic  group  is  more 
closely  related  genealogically  to  other  members  of  its  group  than 
it  is  to  members  of  other  groups.  More  than  one  cladogram  can 
be  generated  with  the  same  data  set,  and  the  most  parsimonious, 
that  is,  the  one  requiring  the  fewest  evolutionary  steps,  is  taken 
as  the  most  natural  or  best.  According  to  Panchen  ( 1 982),  prob- 
lems in  logic  invalidate  the  use  of  parsimony  in  cladistics.  Not 
all  cladists  agree  about  precisely  what  a  cladogram  represents, 
but  some  interpret  it  directly  as  a  phylogenetic  classification. 
One  of  the  greatest  problems  in  using  cladistics  is  the  difficulty 
in  evaluating  character  states  for  primitiveness  or  degree  of 
derivation.  Two  methods  have  been  used;  one  involves  onto- 
genetic stages  and  will  be  discussed  later  in  this  paper.  A  second 
method,  called  out-group  comparison  (Wiley,  1981,  gives  a  good 
description),  is  the  most  subjective  part  of  the  entire  cladistic 
procedure  and  to  a  certain  degree  may  involve  circular  reason- 
ing. A  practical  problem  that  cladistics  has  not  yet  conquered 
is  that  of  naming,  for  classifications  must  be  used  by  many  who 
have  no  interest  in  theory,  and  naming  categories  on  a  strictly 
genealogical  basis  raises  many  problems,  as  does  the  practice 
followed  by  some  cladists  of  naming  all  branching  points.  Some 
attributes  of  ELH  stages  that  might  be  considered  unsuitable 
for  use  in  evolutionary  classification  are  available  for  use  in 
cladistics.  One  example  concerns  character  stages  that  are  in- 
terpreted as  being  highly  adaptive  rather  than  conservative.  If 
polarity  can  be  ascertained,  then  so-called  adaptive  characters 
are  available.  Rates  and  sequences  of  ontogenetic  change  also 
constitute  potentially  valuable  character  suites. 

The  third  method,  presently  called  evolutionary  classification, 
is  more  difficult  to  define  and  discuss.  It  has  a  long  history  and 
an  extensive  literature  (Mayr,  1981).  The  methods  of  evolu- 
tionary classification  are  eclectic  and  generally  more  subjective 
than  those  of  phenetics  and  cladistics.  They  do  not  easily  lend 
themselves  to  overall  generalization.  Characters  are  selected  and 
weighted  by  paying  particular  attention  to  homology  and  con- 
vergence; to  the  extent  that  they  are  available,  evidence  from 
embryology  and  palaeontology  are  also  used.  Primitive  char- 
acters are  admitted  to  the  system.  Data  are  used  from  ecolog- 
ically oriented  facets  of  evolution  such  as  selection,  competition, 
predation,  and  ecological  biogeography.  Historical  biogeogra- 
phy,  rate  of  evolution,  and  genetics  are  also  considered.  An 
hierarchical  classification  is  derived,  which  has  an  inferred  time 
axis  and  which  may  generally  reflect  genealogical  relationships. 
However,  degree  of  phenetic  difference  in  selected  characters, 
which  is  interpreted  as  reflecting  degree  of  genetic  difference, 
may  be  considered  along  with  branching  pattern  in  converting 
a  strict  genealogy  into  a  classification.  Patterson  (1981b)  has 
discussed  and  criticized  such  procedure.  Whatever  may  be  phy- 
letic  relationships,  the  definition  of  taxa  is  essentially  subjective, 
and  each  member  of  a  group  is  not  necessarily  more  closely 


related  genealogically  to  other  members  of  its  group  than  it  is 
to  members  of  a  different  group.  The  test  for  goodness  of  a 
classification  is  pragmatic;  if  it  has  high  predictive  value  it  is 
good.  (By  prediction  is  meant  the  degree  to  which  a  classification 
encompasses  additional  data.)  In  commenting  on  evolutionary 
systematics  Panchen  (1982)  writes  that  it,  "has  always  been 
somewhat  ad  hoc  in  its  procedure,  yielding  good  results  with 
competent  taxonomists  and  bad  with  incompetent  ones.  The 
standard  warks  [sic]  on  procedure  .  .  .  are  to  some  extent  ra- 
tionalizations of  a  tradition  that  is  too  largely  intuitive." 

As  a  summary,  I  have  tried  to  compare  in  Table  1  some  of 
the  techniques,  objectives,  and  assumptions  of  the  three  meth- 
ods. Phenetics  requires  the  fewest  assumptions  but  would  seem 
to  offer  the  systematist  a  classification  with  the  least  information 
value.  Cladistics  has  the  most  constraints,  so  many  and  so  strin- 
gent in  fact,  that  they  may  limit  its  practical  use,  although  the 
method  is  particularly  valuable  in  indicating  areas  for  which 
additional  or  more  suitable  data  are  required.  Misuse  of  cla- 
distics may  soon  rival  the  long-time  abuse  by  systematists  of 
parametric  statistics.  Evolutionary  classification  tries  to  include 
the  most  information  from  the  most  sources,  but  the  methods 
for  doing  so  are  not  very  well  formalized.  Cladists  treat  their 
method  of  classification  as  a  general  theory  of  biology  (Nelson 
and  Platnick,  1981),  a  forcing  function  among  all  evolutionary 
phenomena,  which  must  therefore  comply  with  a  parsimonious 
model  derived  entirely  from  character  state  analysis.  Evolu- 
tionary classification,  on  the  other  hand,  incorporates  infor- 
mation from  a  wide  variety  of  biological  phenomena  and  to 
that  extent  is  forced,  rather  than  forcing.  Predictability,  as  a  test 
of  goodness  for  a  classification,  is  more  pragmatic  and  logically 
less  satisfying  than  is  parsimony.  Perhaps  an  important  question 
for  theoretical  systematists  to  consider  is  the  formulation  of 
comparable  definitions  for  replicability,  parsimony,  and  pre- 
dictability. 

Ontogeny  and  Fish  Phylogeny 

Louis  Agassiz,  who  fought  the  idea  of  organic  evolution,  pro- 
posed a  "threefold  parallelism"  of  arranging  organisms  in  a 
series  or  classification.  His  three  parallels  were  palaeontology, 
what  we  would  now  consider  to  be  homology,  and  ontogeny. 
Even  though  he  failed  to  interpret  the  parallels  as  evidence  for 
evolution,  his  keen  perception  of  the  fact  that  they  do  exist  in 
nature  and  are  somehow  interrelated  has  elicited  extensive  com- 
ment and  reinterpretation  (see  especially  Gould,  1977)  and  is  a 
suitable  point  of  departure  for  addressing  the  importance  of 
ontogeny  as  a  source  of  information  about  homology,  the  bio- 
genetic law,  developmental  stages  as  alternatives  to  outgroup 
comparisons  in  cladistics,  paedomorphosis,  and  the  application 
of  life  history  stages  to  phylogenetic  inquiry. 

If  characters  are  the  meat  and  muscle  of  classification,  then 
homology  surely  shapes  the  skeleton  on  which  phylogenetic  clas- 
sifications are  arranged.  The  worth  of  any  allegedly  phylogenetic 
classification  is  no  better  than  the  degree  to  which  homology 
has  been  assessed,  and  how  to  do  this  is  a  major  problem  for 
the  systematist.  Like  the  weather,  everyone  talks  about  homol- 
ogy but  does  nothing  about  it— or  almost  nothing.  The  concept, 
which  is  so  pervasive  in  the  study  of  phylogeny  and  in  evolution, 
has  been  with  us  since  pre-Darwinian  times,  although  not  always 
in  the  way  that  we  understand  it  today.  The  great  comparative 
anatomist  Owen  defined  it  in  1866  as  follows;  "A  'homologue' 


10 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  1 .    Comparison  of  Three  Methods  Used  in  Biological  Classification. 


Evolutionary' 


Character  weighting 
Convergence 
Homology 
Fossil  History 
Eco-evolutionary  Data 
Rale  of  Evolution 
No.  of  Characters 
No.  of  Specimens 
Branches  from  a  Node 
End  Product 


Test  of  Goodness 


No 

Not  Considered 

Not  Considered 

Not  Considered 

Not  Considered 

Not  Considered 

Many 

Few 

Two  to  Many 

Perhaps  a  Phylogeny 


Replicability 


Yes 

Important 
Important 
Not  Important 
Not  Important 
Not  Important 
One  to  Medium 
Few  to  Many 
Two  when  Possible 
Phyiogenetic  Classification 
Based  on  Genealogy 

Parsimony 


Yes 

Important 

Important 

Important 

Important 

Important 

One  to  Medium 

Few  to  Many 

Two  to  Many 

Phyiogenetic  Classification  Based 

on  Genealogy  and  Degree  of 

Difference 
Predictability 


is  the  same  pail  or  organ  in  different  animals  under  every  variety 
of  form  and  function."  He  goes  on  to  note,  however,  that  some 
earlier  workers  defined  the  concept  as  we  now  define  analogy. 
But  our  problem  remains  identical  with  that  of  Owen— how  to 
define  same.  In  a  recent  discussion  of  homology,  Patterson  (1982) 
proposed  similarity  in  ontogeny  as  part  of  a  test  of  homology. 
But  the  use  of  similarity  in  development  to  help  define  Owen's 
"same"  is  tautological. 

Palaeontologists  proceed  in  a  basically  circular  fashion  in  their 
use  of  homology.  They  depend  upon  a  time  series  to  trace  the 
history  of  transformed  states  of  a  presumably  homologous  char- 
acter along  a  sequence  that  is  interpreted  as  a  genealogy.  But  of 
course  the  characters  are  considered  homologous  because  they 
are  part  of  a  genealogy.  Whether  they  admit  to  it  or  not,  most 
systematists  use  pure  phenetics  in  the  search  for  homology,  and 
although  this  common  sense,  intuitive,  non-scientific  approach 
works  much  of  the  time,  still,  many  systematists  have  misin- 
terpreted as  homologues  characters  that  are  actually  analogous 
and  have  filled  the  literature  with  many  misdiagnosed  conver- 
gences. In  comparative  vertebrate  anatomy  and  systematics,  the 
convention  has  grown  up  that  certain  organ  systems  are  more 
conservative  than  others  and  therefore  provide  a  better  method 
for  detecting  homologies.  The  nervous  system  is  generally  con- 
sidered the  best,  the  skeleton  the  next  best,  followed  by  viscera 
and  muscles,  with  the  integument  the  least  good.  In  fishes,  for 
example,  Freihofer  (1963,  1970)  has  used  the  patterns  of  the 
ramus  lateralis  accessorius  and  ramus  canalis  lateralis  nerve 
systems  relative  to  elements  of  the  skeleton  to  propose  groupings 
of  fishes.  But  even  here  the  possibility  of  convergence  cannot 
be  ignored  (Gosline,  1968),  and  again  the  problem  of  circularity 
arises  because  many  ichthyologists  define  osteological  features 
on  the  basis  of  their  topographic  relation  to  elements  of  the 
nervous  system.  Another  example  relates  to  homologies  of  pho- 
tophore  series  in  lantemfishes  as  determined  by  studies  of  their 
innervation  (Ray,  1950).  Here  also,  the  conclusions  based  on 
this  method  appear  to  be  equivocal  (Moser  and  Ahlstrom,  1 972). 

A  direct  method  for  demonstrating  the  homology  of  structures 
would  be  to  trace  them  back  during  development  to  their  an- 
lagen.  De  Beer  (1951)  has  commented  on  the  apparent  failure 
of  experimental  embryology  to  validate  this  approach.  Even  so, 
a  survey  of  the  development  of  bony  structure  during  fish  on- 
togeny presented  by  Dunn  ( 1 983b)  lists  some  observed  instances 
of  losses,  gains,  and  modifications,  chiefly  in  the  caudal  fin  skel- 
eton, which  interpret  homologies  in  adult  structure;  unfortu- 


nately, these  instances  are  too  few.  Ahlie  had  a  long  interest  in 
the  caudal  fin  skeleton,  particularly  of  flatfishes,  and  the  com- 
pletion of  his  work  by  colleagues  hopefully  will  constitute  an 
additional  contribution  to  the  use  offish  ontogeny  in  identifying 
homologous  structures. 

The  concepts  of  ontogeny  and  homology  are  intimately  as- 
sociated in  the  idea  that  the  study  of  early  life  history  stages  of 
an  organism  will  reveal  its  adult  ancestral  stages— ontogeny  re- 
capitulates phylogeny— as  proposed  by  Ernst  Haeckel  in  the 
latter  half  of  the  19th  century.  Taken  at  its  most  extreme,  the 
biogenetic  law  has  been  interpreted  as  meaning  that  an  entire 
genealogy  is  encapsulated  in  an  ontogenetic  series.  If  adults  of 
extant  species  of  a  group  were  to  be  matched  up  with  their  closest 
approximations  in  an  ontogenetic  series,  homology  would  un- 
fold before  our  eyes.  Of  course  its  value  to  us  in  unraveling 
phylogeny  would  be  redundant,  because  phylogeny  would  be 
there  as  well.  It  was  soon  evident  however  that  the  biogenetic 
model  is  far  too  crude  to  approximate  nature.  The  embryologist 
von  Baer  had  previously  formulated  four  "laws"  or  general 
propositions  about  embryology  that  have  been  restated  in  var- 
ious forms  by  many  authors  and  applied  to  the  interpretation 
of  phylogeny.  The  following  are  taken  from  De  Beer  ( 1 95  1 ):  ( I ) 
In  development  from  the  egg  the  general  characters  appear  be- 
fore the  special  characters.  (2)  From  the  more  general  characters 
the  less  general  and  finally  the  special  characters  are  developed. 
(3)  During  its  development,  an  animal  departs  more  and  more 
from  the  form  of  other  animals.  (4)  The  young  stages  in  the 
development  of  an  animal  are  not  like  the  adult  stages  of  other 
animals  lower  down  on  the  scale,  but  are  like  the  young  stages 
of  those  animals.  These  propositions  are  useful  generalizations 
and  we  can  all  think  of  obvious  instances  of  fish , ontogeny  that 
can  be  interpreted  by  one  or  more  of  them .  Consider  for  example 
the  bilaterally  symmetrical  larvae  of  flatfishes,  the  early  presence 
and  subsequent  loss  of  a  swimbladder  in  stromateoids  (Horn, 
1970a),  the  sequence  of  fusions  during  ontogeny  in  the  caudal 
fin  skeleton  of  myctophids  (Ahlstrom  and  Moser,  1976),  the 
ontogeny  of  the  upper  jaw  bones  and  dentition  in  notosudids 
(Berry,  1964a),  and  the  presence  of  a  pectoral  fin  in  larval  Tac- 
tosloma  and  its  loss  in  adults  (Ahlstrom,  lecture  notes).  On  the 
other  hand,  a  plethora  of  early  life  history  stages  of  fishes  man- 
ifests character  states  that  represent  morphological  specializa- 
tions occurring  early  in  development.  Consider  the  egg  stages 
of  macrourids  with  their  hexagonal  patterns,  atherinomorphs 
with  their  filaments,  and  argentinoids  with  their  pustules.  Other 


COHEN:  ONTOGENY,  SYSTEMATICS,  PHYLOGENY 


11 


instances  for  which  it  is  difficult  to  accept  that  ontogeny  has 
recapitulated  phylogeny  include  the  leptocephalus  of  eels,  the 
stalked  eyes  of  assorted  larval  bathylagids,  myctophids  and  Idi- 
acanthus.  the  elongated  guts  of  larval  melanostomiatids,  the 
extensive  armature  of  many  spiny-rayed  fishes  during  their  lar- 
val stages,  and  the  produced  fin  rays  found  in  many  kinds  of 
larval  fishes.  Examples  of  all  of  these  are  illustrated  and  de- 
scribed in  this  volume.  With  regard  to  proposition  three  in 
particular,  Ahlie  often  pointed  out  instances  of  fishes  that  were 
easily  distinguished  as  larvae  but  became  more  similar  in  ap- 
pearance as  adults;  one  example  is  Bathylagiis  milleh  and  B. 
pacificus;  Myctophum  aurolaternalum  and  other  myctophid 
species  is  another.  Von  Baer's  propositions  as  applied  to  phy- 
logeny are  tidy  and  appealing  but  are  completely  operative  only 
under  the  rather  special  condition  that  major  evolutionary 
changes  (except  for  paedomorphosis)  are  restricted  to  the  adult 
stage  (Gould,  1977;  Fink.  1982). 

For  cladistic  analysis,  the  polarization  of  characters  through 
direct  observation  of  their  transformation  during  ontogeny  has 
been  discussed  by  Nelson  (1978)  and  others  as  an  alternative 
to  the  often  unsatisfactory  indirect  method  of  outgroup  com- 
parison. Such  use  of  ontogeny,  which  depends  on  von  Baer's 
first  three  propositions,  has  been  analyzed  by  Henning  (1966), 
who  noted  its  uncertainty.  As  examples  from  fish  ontogeny  given 
above  indicate,  ontogeny  could  replace  or  corroborate  outgroup 
comparison  but  only  to  the  extent  that  the  biogenetic  law  is 
valid  for  a  particular  situation.  Patterson's  (1982)  statement, 
"that  ontogeny  is  the  decisive  criterion  in  determining  polarity," 
would  seem  to  be  based  on  limited  acquaintance  with  ELH 
stages. 

Paedomorphosis  refers  to  the  presence  in  adults  of  larval  char- 
acters (De  Beer,  1951)  and  has  been  variously  considered  as 


insignificant  to  very  important  in  evolution.  For  fishes  at  least, 
I  think  the  latter  is  the  case.  As  one  example,  small  adult  size 
could  be  considered  a  particularly  widely  distributed  neotenic 
character.  In  his  discussion  of  paedomorphosis  and  cladistics. 
Fink  (1982)  remarked  that  it  is  difficult  to  identify  this  phe- 
nomenon without  paired  taxa,  but  surely  this  is  not  always  true. 
Although  the  relationships  of  the  curious  little  fish  Schindleria 
are  unknown,  it  would  be  difficult  to  deny  that  it  has  many 
neotenic  characters  (Watson,  Stevens  and  Matarese,  this  vol- 
ume). On  a  larger  scale  paedomorphosis  may  have  been  im- 
portant in  establishing  novel  phyletic  lines  as  well  as  isolated 
species  or  genera,  and  the  study  of  ELH  stages  will  be  essential 
in  detecting  these  divergences. 

I  end  this  essay  by  noting  that  the  most  important  use  of  all 
for  information  about  fish  ontogeny  may  be  providing  characters 
for  charting  fish  phylogeny  rather  than  theories  about  phylogeny. 
Distinguishing  and  identifying  species  for  purposes  of  fish  bi- 
ology and  management  has  been  the  chief  use  for  what  is  called 
larval  fish  taxonomy,  and  the  large  resulting  literature  is  sum- 
marized in  this  volume.  Many  of  the  same  descriptive  data  are 
of  apparent  value  for  purposes  of  grouping  similar  species  or 
other  taxa  for  phyletic  purposes.  Published  examples  of  syn- 
thesis are  far  fewer  than  of  descriptions,  but  accounts  using  each 
of  the  three  methodologies  previously  described  are  available, 
either  cited  in  this  volume  or  presented  here  as  original  research. 
ELH  characters  can  meet  many  methodological  constraints  and 
will  be  used  increasingly  by  ichthyologists.  To  what  advantage 
remains  to  be  seen,  but  the  prognosis  is  good. 

Life  Sciences  Division,  Los  Angeles  County  Museum  of 
Natural  History,  900  Exposition  Boulevard,  Los 
Angeles,  California  90007. 


Early  Life  History  Stages  of  Fishes  and  Their  Characters 
A.  W.  Kendall,  Jr.,  E.  H.  Ahlstrom  and  H.  G.  Moser 


Patterns  of  Teleost  Early 
Life  History 

IN  discovering  that  Atlantic  cod  lay  free-floating  planktonic 
eggs  which  develop  into  pelagic  larvae,  G.  O.  Sars,  in  1865 
(see  Hempel,  1979;  Ahlstrom  and  Moser,  1981)  had  also  come 
upon  an  example  of  the  widespread  life  history  pattern  of  marine 
fishes.  Most  marine  fishes,  regardless  of  systematic  affinities, 
demersal  or  pelagic  habits,  coastal  or  oceanic  distribution,  trop- 
ical or  boreal  ranges,  spawn  pelagic  eggs  that  are  fertilized  ex- 
ternally and  float  individually  near  the  surface  of  the  sea  (Fig. 
5).  These  eggs  range  from  about  0.6  to  4.0  mm  in  diameter 
(mode  about  1  mm)  and  generally  are  spherical.  Within  a  species 
there  is  little  variation  in  egg  characters  such  as  size,  number 
and  size  of  oil  globules,  and  pigmentation  and  morphology  of 
the  developing  embryo.  Development  time  is  highly  tempera- 
ture dependent  and  also  species-specific.  The  eggs  hatch  into 
relatively  undeveloped  yolk-sac  larvae  which  swim  feebly  and 


rely  on  their  yolk  for  nourishment  while  their  sensory,  circu- 
latory, muscular,  and  digestive  systems  develop  to  the  point 
that  they  can  feed  on  plankton.  Even  these  yolk-sac  larvae  have 
characters  (pigment  patterns,  body  size  and  shape,  myomere 
number)  that  reflect  their  heritage.  After  the  yolk  is  utilized, 
they  develop  transient  "larval"  characters  such  as  pigment  pat- 
terns and,  in  some,  specialized  head  spines  and  fin  structures 
that  are  apparently  adaptive  for  this  phase  of  their  life  history. 
During  this  period  more  characteristics  of  the  adult  (e.g.,  me- 
ristic  characters)  gradually  develop.  At  the  end  of  the  larval 
stage,  they  may  go  through  an  abrupt  transformation  to  the 
juvenile  stage,  particularly  if  they  move  from  a  pelagic  to  de- 
mersal habitat,  or  the  transformation  may  be  gradual.  In  some 
fishes,  there  is  a  prolonged  and  specialized  stage  between  the 
larval  and  juvenile  stages.  These  pelagic  (often  neustonic)  forms 
eventually  transform  into  demersal  juveniles.  The  juvenile  stage 
is  characterized  by  specimens  having  the  appearance  of  small 
adults— all  fin  rays  and  scales  are  formed,  the  skeleton  is  almost 


EGGS 


YOLK 
SAC 


PRE 

FLEXION 


FLEXION 


POST 
FLEXION 


> 

< 

> 

m 


JUVENILE 


Fig.  5.     Early  life  history  stages  of  Trachurus  symmelricus  from  Ahlstrom  and  Ball  (1954), 


KENDALL  ET  AL.:  ELH  STAGES  AND  CHARACTERS 


13 


END  POINT  EVENTS 

TERMINOLOGY 

Primary  developmental 
stages 

Transitional  stages 
Subdivisions 


OTHER 
TERMINOLOGIES 

Hubbs,  1943,1958 

Sette,  1943 

Nikolsky,  1963 

Hattori,  1970 

Balon,  1975  (phases) 

Snyder,  1976,1981  (phases)      \ 


E  q  q 

Larva 

Juvenile 

1                          '                          ' 

1 

Yolk  sac 
larva 


Transforma 
lion  larva 


Early 

Middle 

Late 

Piftlexion 

Flexion 

PnMflexinn 

larva 

larva 

larva 

Pelagic  or 

special  juven 


' 

1 

E  m  b 

y  o 

Proiarva 

Post  1  a  rv  a 

Prejuvenile 

' 

1                        1 

Larva 

Post  larva 

Embryo 


Prelarva 


Cleavage  egg 

Embryo 

Eleulhero 
embryo 

Protoptery- 
qiolarva 

Pterygiolarva 

1 
1 
1 
1 

1                            1                             1                            1 

Protolarva 

M  e  s  0- 
larva 

# 

M  e  t  a  1  a 

V  a 

Fig.  6.    Terminology  of  early  life  history  stages. 


completely  ossified,  the  larval  pigment  pattern  is  overgrown  or 
lost  and  replaced  by  dermal  pigment  similar  to  that  of  the  adults, 
and  the  body  shape  approximates  that  of  the  adults. 

Although  this  is  the  most  frequently  observed  life  history 
pattern,  there  are  many  variations  (see  Breder  and  Rosen,  1 966) 
often  related  to  increased  parental  investment  in  individual 
progeny  with  a  concomitant  decrease  in  fecundity  and  larval 
specializations.  There  is  scant  information  on  the  young  of  many 
deep-sea  fishes,  and  this  may  be  due  in  part  to  life  history 
strategies  that  do  not  include  eggs  and  larvae  that  occur  in  the 
epipelagic  zone  (where  most  of  the  collecting  is  done).  Marshall 
(1953)  discussed  life  history  adaptations  of  these  fish  such  as 
the  production  of  few,  large  yolky  eggs  that  hatch  into  relatively 
advanced  larvae.  These  young  may  remain  far  below  the  more 
productive  surface  layers,  and  thus  not  be  susceptible  to  most 
sampling  procedures.  Markle  and  Wenner  (1979)  cite  evidence 
for  demersal  spawning  of  two  species  of  groups  (Alepocephal- 
idae,  Zoarcidae)  that  are  seldom  collected  in  the  plankton  as 
larvae. 

Many  coastal  marine  and  nearly  all  freshwater  fishes  lay  de- 
mersal eggs  which  are  generally  larger  than  the  I  mm  mode  of 
pelagic  eggs.  In  such  fish  development  from  hatching  through 
juvenile  stage  is  direct  and  the  larvae  gradually  attain  adult 
characters  of  shape,  pigmentation,  and  meristic  features.  The 
demersal  eggs  frequently  are  adhesive  and  laid  in  some  sort  of 
nest.  Parental  care  of  the  nest  is  observed  in  many  species,  and 
this  care  may  extend  to  the  larvae  after  hatching  (e.g.,  mouth 
brooding  in  cichlids,  ariids).  Parental  care  takes  another  form 
in  Sehastes.  where  development  through  the  yolk-sac  stage  takes 


place  in  the  ovary  and  first-feeding  larvae  are  extruded.  Vivi- 
parity, in  which  nourishment  is  supplied  by  maternal  structures, 
has  evolved  many  times  (e.g.,  poeciliids,  some  zoarcids,  em- 
biotocids),  whereby  the  larval  stage  is  bypassed  and  the  fish  are 
extruded  ("bom")  as  juveniles  (Wourms,  1981). 

Early  Life  History  Stages 

Between  spawning  and  recruitment  into  the  adult  population, 
most  fishes  undergo  dramatic  changes  in  morphology  and  hab- 


Table  2.    Examples  of  Characters  of  Pelagic  Eggs  that  May  Be 
Useful  for  Systematic  Studies  of  Certain  Fishes. 


Character  slates 


Systematic  groups 


Egg  size 

Egg  shape 

Envelope 
sculpturing 


Oil  globule 
position 

Embryonic 
characters 


<  1  mm->5  mm 
>3  mm->5  mm 

Round  — oblong 


Varying  distances  between 

pores 
Varying  length/density  of 

filaments 

Anterior  to  posterior  in 
yolk  sac 

Slate  of  development  of 
various  organs/organ  sys- 
tems at  various  develop- 
mental mileposts 


Pleuronectidae 
Anguilliformes 

Engraulidae 
Ostraciontidae 

Gadidae 

Atheriniformes 
(Exocoetidae) 

Perciformes 
Gadidae 


14 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  7.  Examples  of  features  of  yolk-sac  larvae  of  teleosts.  (A-C).  Paracallionymus  costatus.  A.  soon  after  hatching  0.98  mm  NL;  B.  1.8  mm 
NL;  C.  1.9  mm  NL.  From  Brownell  (1979).  Features  demonstrated  in;  (A)  include  the  small  size  of  the  larva,  the  lack  of  an  oil  globule,  the 
segmented  yolk,  and  the  dorsally  arranged  melanophores;  (B)  demonstrates  the  migration  of  melanophores  ventrally  and  the  formation  of  the 
anus  producing  a  preanal  finfold;  (C)  demonstrates  further  ventral  migration  of  melanophores,  beginning  of  larval  pectoral  fin  formation,  the 
decrease  in  yolk-sac  size,  and  beginning  of  pigment  in  the  eye;  (D)  Diplodus  sargus.  2.4  mm  NL.  From  Brownell  (1979).  Single  pigmented  oil 
globule  posterior  in  the  unsegmented  yolk  and  a  short  preanal  finfold  are  demonstrated;  (E)  Trachurus  I.  capensis.  2.2  mm  NL.  From  Brownell 
(1979).  Single  pigmented  oil  globule  anterior  in  segmented  yolk  with  moderately  long  preanal  finfold  demonstrated;  (F)  Cololabis  saira.  5.1  mm 
SL.  (original).  Well-developed,  heavily  pigmented  yolk-sac  larva  at  hatching  with  notochord  flexion  beginning  and  some  caudal  rays  formed;  (G) 
Argentina  silus.  1.1  mm.  Redrawn  from  Schmidt  (1906c).  A  large  but  poorly  developed  yolk-sac  larva  at  hatching  with  a  large  oil  globule;  and 
(H)  Hippoglossus  slenolepis.  9.5  mm.  From  Pertseva-Ostroumova  (1961).  A  large  but  poorly  developed  yolk-sac  larva  at  hatching  with  no  oil 
globule. 


its.  As  mentioned  earlier,  at  hatching,  particularly  in  marine 
fishes  with  pelagic  eggs,  the  fish  is  in  an  extremely  undeveloped 
state  and  then,  as  a  free-living  individual,  it  gradually  develops 
the  adult  characters.  This  process  is  continuous,  but  there  are 
morphological  and  ecological  mileposts  that  are  significant  in 
the  life  of  the  fish  and  which  allow  us  to  subdivide  this  process 
so  that  we  can  communicate  results  of  our  studies  and  compare 
different  fishes  at  the  same  moment  in  development. 

Fish  early  life  history  has  been  and  continues  to  be  studied 
from  a  number  of  different  perspectives  (Ahlstrom  and  Moser, 
1976).  Some  studies  deal  directly  with  embryology  and  later 
ontogeny,  others  emphasize  functional  morphology  of  larval 
structures,  apply  larval  features  to  taxonomic  and  systematic 
studies,  investigate  the  ecology  of  eggs  and  larvae,  or  use  these 
stages  to  address  fishery-related  problems  such  as  assessment 
of  spawning  stock  size  and  recruitment  success.  All  of  these 


studies  have  in  common  the  need  to  subdivide  early  life  history 
and  communicate  information  based  on  processes  and  events 
occurring  during  these  subdivisions.  As  with  any  communica- 
tion, it  is  vitally  important  to  use  terms  that  are  clearly  defined 
and  this  is  particularly  true  with  the  diverse  disciplines  that  are 
involved  in  larval  fish  studies.  Historically,  several  disciplines 
have  used  different  names  for  the  same  stage,  or  subdivided 
development  differently  [see  Okiyama  (1979a)  and  Fig.  6  in  this 
paper].  This  has  led  to  confusion  rather  than  communication. 
Several  criteria  seem  appropriate  for  defining  stages  of  de- 
velopment to  be  used  by  students  of  any  discipline.  The  variety 
of  developmental  patterns  should  be  recognized  and  the  defi- 
nitions should  apply  to  as  many  patterns  as  possible.  Thus, 
stages  should  be  based  on  very  widespread,  fundamental  fea- 
tures of  development.  The  stages  should  have  some  significance 
in  the  life  history  of  the  fish,  both  morphologically  and  func- 


KENDALL  ET  AL.:  ELH  STAGES  AND  CHARACTERS 


15 


From  demersal  eggs 


From  pelagic  eggs 


Clupea  harengus  harengus 

egg  diameter  =  1.2-1. 5mm 
NL  at  hatclning  =  4.9mm 


Etrumeus   teres 
egg  diameter  =  1.3mm 
NL  at  hatching  =  4.8mm 


Krevanoski  1956 


Mito  1961 


O 


Mukhacheva  and  Zviagina  1960 


Gadus  macrocephalus 

egg  diameter  =  0.8-1. 4mm 
NL  at  hatching  =  3.6mm 


Colton  and  IWarak  1961 


Gadus  morhua 
egg  diameter  =  1.1 -1.9mm 
NL  at  hatching  =  3.6mm 


Lepidopsetta  bilineata 
egg  diameter  =  1.02-1. 09mm 
NL  at  hatching  =  3.9mm 


Isopsetta  isolepis 
egg  diameter  =  0.90-0. 99mm 
NL  at  hatching  =  2.9mm 


Pertseva-Ostroumova  1961 


Richardson  et  al  1980 


Fig.  8.     Newly  hatched  yolk-sac  larvae  of  related  fishes  with  pelagic  and  demersal  eggs  of  comparable  sizes. 


tionally,  such  as  a  particular  type  of  nourishment  or  locomotion. 
Also  the  endpoints  for  the  stages  should  be  easily  observed  and 
sharply  defined. 

The  most  general  scheme  of  terminology  of  early  development 
of  fishes  includes  (Fig.  5): 

The  "egg  stage"  (spawning  to  hatching).  The  egg  stage  is  used 
in  preference  to  the  embryonic  stage  because  there  are  characters 
present  during  this  stage  other  than  just  embryonic  characters 
(e.g.,  those  associated  with  the  egg  envelope). 

The  "larval  stage"  (hatching  to  attainment  of  complete  fin 
ray  counts  and  beginning  of  squamation).  One  of  the  funda- 
mental events  in  development  of  most  fishes  is  the  flexion  of 
the  notochord  that  accompanies  the  hypochordal  development 
of  the  homocercal  caudal  fin.  It  is  convenient  to  divide  the  larval 
stage  on  the  basis  of  this  feature  into  "preflexion."  "flexion," 
and  "postflexion"  stages.  The  flexion  stage  in  many  fishes  is 
accompanied  by  rapid  development  of  fin  rays,  change  in  body 
shape,  change  in  locomotive  ability,  and  feeding  techniques. 

The  "juvenile  stage"  (completion  of  fin  ray  counts  and  be- 
ginning of  squamation  until  fish  enters  adult  population  or  at- 
tains sexual  maturity). 

Transitional  stages  can  also  be  recognized:  the  "yolk-sac  larval 
stage"  (between  hatching  and  yolk-sac  absorption);  and  the 
"transformation  stage"  (between  larva  and  juvenile).  Meta- 
morphosis occurs  during  this  stage  and  is  considered  complete 
when  the  fish  assumes  the  general  features  of  the  juvenile. 

The  life  histories  of  some  fishes  include  other  specialized 
ontogenetic  stages  that  have  received  various  names.  In  some 
cases,  these  are  the  generic  names  under  which  these  stages  were 


described  before  they  were  recognized  as  larvae  of  other  species 
(e.g.,  the  leptocephalus  stage  of  Anguilliformes,  the  scutatus 
stage  of  Anlennarius.  the  vexillifer  stage  of  Carapidae.  and  the 
kasidoron  stage  of  Gihhertchthys).  In  other  cases,  consistent  fea- 
tures of  development  of  a  group  permit  useful  subdivisions  of 
stages  (e.g..  in  leptocephali  the  engyodontic  and  euryodontic 
stages). 

The  Egg  Stage 

Hempel  (1979)  reviewed  the  egg  stage  relative  to  fisheries 
investigations.  Ahlstrom  and  Moser  (1980)  presented  a  concise 
review  of  the  range  of  characters  observed  in  pelagic  fish  eggs, 
particularly  those  useful  in  identifying  eggs  in  plankton  samples. 
Sandknop  and  Matarese  in  this  volume  also  discuss  this  subject 
in  detail.  The  characters  that  have  proven  useful  for  egg  iden- 
tification include  egg  size  and  shape,  size  of  perivitelline  space, 
yolk  diameter  and  character  (homogeneous  or  segmented),  num- 
ber and  size  of  oil  globules,  texture  of  the  egg  envelope  (smooth 
or  with  protrusions),  pigment  on  the  yolk  and  embryo,  and 
characters  of  the  developing  embryo  (relative  rate  of  develop- 
ment of  various  parts,  body  shape,  number  of  somites)  (Table 
2). 

The  egg  stage  has  been  subdivided  by  a  number  of  workers 
(e.g.,  Apstein,  1909).  Fishery  biologists  need  to  determine  the 
age  of  eggs  at  the  time  of  collection  for  production,  drift,  and 
mortality  estimates.  Embryologists  have  designated  stages  to 
coincide  with  significant  developmental  features.  While  the  stages 
of  fishery  biologists  are  designed  to  divide  the  embryonic  stage 
into  several  easily  recognized  portions,  embryologists  are  more 


16 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  3.    Examples  of  Use  of  Characters  of  Early  Life  History  Stages  in  Taxonomic  and  Systematic  Studies.  X  indicates  range  of  stages 

and  taxonomic  levels  at  which  characters  vary.  (X)  indicates  infrequent  state. 

Developmental  stage 


Character 


Lar\ac 

Taxonomic  level 

Yolk- 
sac 

Pre- 
flexion 

Flexion 

Post- 
flexion 

Trans- 
forma- 
tion 

Rpfprenf**^ 

Spe- 
cies 

Genus 

Family 

Order 

IX \.  IK  1  \,lkVV«^ 

Egg 

Keyed  to  Table  4 

(X) 

X 

X 

(X) 

X 

20 

X 

X 

(X) 

(X) 

X 

20,29 

(X) 

X 
X 

X 
X 

X 

X 

(X) 

X 

2,38 

1,  2,  11,  19,  24,  27,  39 

(X) 

X 

X 

X 

11,  19,24,27,39 

X 

X 

X 

X 

X 

1,2,3.5,  11.  15,  17,  19, 
20,25.27.28.33,34 

X 

X 

(X) 
X 

X 

X 

(X) 

X 

27,38 
19 

X 

X 

X 

X 

X 

X 

X 

2,  3,4.  5.  10,  11,  13,  14, 
19,20,23.  24,25,  26,  27, 
28,  29,  31»,  33,  37,  40 

X 

X 

X 

X 

X 

28,33,35,36,38 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

1,2,3,4,8,9,  11,  13,  14, 
15,  17,  19,  20,21,22, 
25,27,28,29,33,36, 
38,  39,  40 

X 

X 

X 

X 

X 

X 

X 

9,  11,23,24,25,27,36. 
38,40 

X 

X 

X 

X 

X 

1,9,  14,23,27,29,33 

X 

X 
X 

X 
X 

X 
X 

14,  27,  29 

27 

X 

X 

X 

8.  10,  14 

X 

X 
X 

X 

X 

X 
X 

36 

8,  14,  15.33 

X 

X 

X 

X 

X 

20,  33,  38.  39 

(X) 

X 

X 
X 
X 

X 

X 
X 
X 

X 

X 

X 
X 

X 

X 

X 

X 

X 

X 

14,20,29,33,38 

8,  10.  14,20,33 
14,20,33 
10,  14,20 
29 

X 

X 

X 

6,  19,20,30,32 

(X) 

X 
X 

X 

(X) 
X 

X 
X 
X 

X 
X 

X 
X 

X 

X 

7,  16,  19,23.29,33,40 

11,27 

12,  14,21 

X 

X 

X 

X 

X 

X 

X 

X 

10,  11,22,23,29,30,39 

(X) 

X 

X 

X 

X 

13,  14,20,26,27,34 

Meristic  characters 
Fin  spines/soft  rays 
Principal  caudal  rays 
Pelvic  fin 
Dorsal/anal  fin 
Pectoral  fin 
Vertebrae 

Branchiostegals 
Gill  rakers 

Larval  characters 

Body  shape 


Snout  shape 
Pigment  patterns 


Head  spines 

Fin  ray  elongation 
Fin  ray  ornamentation 
Fin  ray  serration 
Pinfold  size/shape 
Preanal  finfold 
Pectoral  size  shape 
Larval  gut 

Shape 

Length 
Larval  eye 

Shape 

Stalked 

Choroid  tissue 

Migration 

Other  characters 
Egg  characters 
Osteological  development 
Scale  formation 
Photophore  formation 
Size  at  developmental  stage 
Fin  development  sequence 


•  Emphasis  on  oil  globule  placement  in  yolk-sac  larvae. 


interested  in  tracing  the  sequence  of  development.  The  em- 
bryologist's  approach  will  probably  provide  more  useful  infor- 
mation for  systematic  investigations. 

Although  excellent,  early  descriptive  work  was  done  on  teleost 
embryology  (e.g.  Wilson,  1891),  comparative  research  on  de- 
velopment needs  to  be  done  to  allow  an  evaluation  of  its  value 
to  syslematics,  a  subject  that  has  proven  so  fruitful  among  in- 
vertebrates. It  appears,  from  the  characters  that  have  been  stud- 
ied in  greatest  detail,  that  convergence  may  overshadow  phy- 
letically  significant  information.  For  instance,  the  egg  envelope 
sculpturing  on  Pleuronichthys,  a  pleuronectiform,  was  found 


even  on  scanning  electron  microscope  examination  to  be  quite 
similar  to  that  on  Synodus,  a  myctophiform  (Sumida  et  al., 
1979).  Phylogenetically  diverse  fishes  often  have  round  pelagic 
eggs,  about  1  mm  in  diameter,  with  a  single  oil  globule.  Demersal 
eggs  from  equally  diverse  fishes  are  generally  larger  than  I  mm 
and  develop  a  vitelline  circulatory  system.  Yolk  segmentation 
seems  to  be  a  character  of  more  primitive  fishes,  but  some 
carangids  and  other  perciforms  have  yolks  that  are  secondarily 
segmented  in  an  evolutionary  sense.  Detailed  studies  are  needed 
to  sort  out  these  and  other  features  of  the  teleost  egg  and  its 
embryonic  development  in  a  systematic  context. 


KENDALL  ET  AL.:  ELH  STAGES  AND  CHARACTERS 


17 


Table  4. 


Some  Contributions  in  Which  Ontogenetic  Characters  have  been  used  to  Examine  Systematic  Relationships  (Updated  from 

Ahlstrom  and  Moser,  1981). 


References 

Dale 

Ciroup  dealt  with 

Egg 

Stages 

Ur- 
vac 

Juv 
ad 

Larval  characters  showing 
relationships 

No. 

Among 
spe- 
cies 

Among 
genera 

Among 
subfam- 

or          Among 
families      orders 

1,3.5 

Ege.  V. 

1930,53,57 

Paralepididae 

— 

+ 

+ 

X 

X 

2 

Bertelsen.  E. 

1951 

Ceratioidei 

— 

+ 

+ 

X 

X 

X 

4 

Bertelsen.  E.,  and  N.  B.  Marshall 

1956 

Minpinnati 

— 

+ 

+ 

X 

X 

X 

6 

Pertseva-Ostroumova.  T.  A. 

1961 

Pleuronectidae 

+ 

+ 

+ 

X 

X 

7 

Berry,  F.  H. 

1964a 

Mar.  teleosts 

— 

+ 

— 

X 

g 

Pertseva-Ostroumova,  T.  A. 

1964 

Myctophidae 

— 

+ 

— 

X 

9 

Gutherz,  E.  J. 

1970 

Bothidae 

— 

+ 

— 

X 

10,  14 

Moser.  H.  G.,  and  E.  H.  Ahlstrom 

1970,  74 

Myctophidae 

— 

+ 

+ 

X 

X 

X 

11 

Mead.  G.  W. 

1972 

Bramidae 

— 

+ 

+ 

X 

X 

12 

Ahlstrom,  E.  H. 

1974 

Stemoptychidae 

— 

+ 

+ 

X 

13 

Johnson.  R.  K.. 

1974b 

Scopelarchidae 

— 

+ 

+ 

X 

X 

15 

Okiyama.  M. 

1974a 

Myctophiformes 

— 

+ 

— 

X 

X 

16 

Potthofr.  T. 

1974 

Scombndae 

— 

+ 

+ 

X 

17 

Richards,  W.  J.,  and  T.  Potthoff 

1974 

Scombridae 

— 

+ 

+ 

X 

18 

Aboussouan.  A. 

1975 

Carangidae 

— 

+ 

— 

X 

19 

Ahlstrom,  E.  H..  J.  L.  Butler,  and 
B.  Y.  Sumida 

1976 

Stromateoidei 

+ 

+ 

+ 

X 

X 

X 

20 

Ahlstrom.  E.  H..  and  H.  G.  Moser 

1976 

Mar.  teleosts 

+ 

+ 

+ 

X 

21 

Ahlstrom.  E.  H.,  H.  G.  Moser,  and 
M.  J.  OToole 

1976 

Myctophidae 

— 

+ 

+ 

X 

22 

Bertelsen.  E.,  G.  Krefft,  and  N.  B. 
Marshall 

1976 

Notosudidae 

— 

+ 

± 

X 

X 

23 

Futch.  C.  R. 

1977 

Bothidae 

— 

+ 

— 

X 

X 

24 

Moser.  H.  G.,  E.  H.  Ahlstrom,  and 
E. Sandknop 

1977 

Scorpaemdae 

— 

+ 

± 

X 

X 

X 

25 

Okiyama,  M.,  and  S.  Ueyanagi 

1978 

Scombridae 

— 

+ 

— 

X 

X 

26 

Powlcs.  H..  and  B.  W.  Stender 

1978 

Sciaenidae 

— 

+ 

± 

X 

27 

Kendall.  A.  W..  Jr. 

1979 

Serranidae 

— 

+ 

+ 

X 

X 

28 

Ueyanagi,  S.,  and  M.  Okiyama 

1979 

Scombridae, 
Istiophoridae 

— 

+ 

+ 

X 

29 

Amaoka.  K. 

1979 

Pleuronectiformes 
(in  part) 

— 

+ 

— 

X 

X 

30 

Dotsu.  Y. 

1979 

Gobiidae 

+ 

+ 

— 

X 

31 

Suzuki.  K..  and  S.  Hioki 

1979a 

Percoidei 

+ 

+ 

— 

X 

X 

32 

Mito.  S. 

1979a.  b 

Mar.  teleosts 

+ 

— 

— 

X 

X 

33 

Okiyama.  M. 

1979b 

Myctophoidei 

— 

+ 

— 

X 

34 

Potthoff.  T..  W.  J.  Richards,  and 
S.  Ueyanagi 

1980 

Scombrolabracidae 

— 

+ 

+ 

X 

X 

35 

Zahuranec,  B.  J. 

1980 

Myctophidae 

( Na  nnobrach  lu  m) 

— 

+ 

+ 

X 

X 

36.37 

Richardson.  S.  L. 

1981a,c 

Cottidae 

— 

+ 

+ 

X 

38 

Washington,  B.  B. 

1981 

Cottidae 

— 

+ 

X 

X 

39 

Johnson.  R.  K. 

1982 

Scopelarchidae 
Evermannellidae 

— 

+ 

+ 

X 

X 

X 

40 

Kendall,  A.  W.,  Jr.,  and  B.  Vinter 

1984 

Hexagrammidae 

- 

+ 

+ 

X 

X 

The  Yolk-sac  Larval  Stage 

At  hatching,  larvae  can  be  at  various  states  of  developmenl, 
dependent  to  a  large  degree  on  the  size  of  the  yolk  (Fig.  7). 
Larvae  from  eggs  with  small  yolks  are  less  developed  at  hatching 
than  those  that  hatch  from  eggs  with  larger  yolks.  Since  the  bulk 
of  maiine  fish  spawn  eggs  that  are  about  I  mm  in  diameter  and 
have  a  narrow  perivitelline  space,  the  yolk  is  only  slightly  less 
than  I  mm.  Larvae  from  such  eggs  generally  lack  a  functional 
mouth,  eye  pigment,  and  differentiated  fins.  They  possess  a  large 
yolk  sac  relative  to  the  size  of  the  lai~va  which  supplies  nour- 
ishment while  the  larvae  develop  to  become  self-feeding.  Newly 
hatched  larvae  from  demersal  eggs  are  generally  further  ad- 


vanced in  development  than  lai^ae  from  pelagic  eggs  of  com- 
parable size  (Fig.  8).  In  these  and  other  fish  with  large  eggs, 
hatching  may  be  delayed  until  the  yolk  sac  is  absorbed  and  the 
larvae  are  ready  to  feed  at  hatching,  having  bypassed  the  yolk- 
sac  larval  stage.  The  delayed  absorption  of  yolk  reaches  an  ex- 
treme in  fishes  such  as  salmonines  in  which  the  yolk-sac  larva 
transforms  directly  into  a  juvenile;  Hubbs  (1943)  proposed  the 
term  "alevin"  be  applied  to  this  yolk-sac  larval  stage. 

At  hatching,  locomotion  and  orientation  of  most  yolk-sac 
larvae  are  aided  by  a  continuous  median  finfold  (dorsal,  caudal, 
anal)  and  larval  pectoral  fins.  During  egg  development,  many 
fish  embryos  develop  melanophores  that  originate  in  the  neural 


18 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


,r ..'—v  ^„.-«n.-)-'-T' ^ 


Fig.  9.  Examples  of  teleost  larvae  illustrating  extremes  of  some  systematically  useful  larval  characters.  (A)  Myctophum  aurolaternatum.  26.0 
mm  (Moser  and  Ahlstrom,  1974).  Note  stalked  oval  eye  with  choroid  tissue,  trailing  gut,  and  dorsal  fin  developing  in  finfold;  (B)  Epinephelus 
sp..  8.4  mm  (Kendall,  1979).  Note  elongate,  serrate  dorsal  and  pelvic  spines;  (C)  Adioryx  (Holocentrus)  vexillarius.  8.5  mm  (McKenney,  1959). 
Note  head  spines;  and  (D)  Lopholatilus  chamaeleonticeps,  6.0  mm  (Fahay  and  Berrien,  1981).  Note  spines  on  head  and  body. 


crest  and  are  generally  aligned  along  the  dorsal  surface  of  the 
embryo.  During  the  yolk-sac  stage,  these  melanophores  move 
laterally  and  ventrally  to  establish  the  beginning  of  the  larval 
pigment  pattern.  Orton  (1953a)  describes  these  events  in  detail 
in  Sardinops  sagax.  This  realignment  may  begin  during  the  late 
embryonic  stages,  before  hatching.  Some  species  hatch  with  few 
if  any  melanophores,  and  when  they  first  appear,  they  are  in 
ventral  positions.  Apparently,  the  pigment  cells  migrate  before 
pigment  formation  occurs. 

The  presence  and  position  of  oil  globules  in  yolk-sac  larvae 
vary  and  can  be  of  diagnostic  value.  In  fishes  with  single  oil 
globules,  it  can  be  far  forward  (e.g.,  labrids,  most  carangids, 
muUids,  and  lethrinids),  in  the  middle  of  the  yolk  sac  (e.g..  some 
clupeids,  serranids,  and  argentinids),  or  more  usually  near  the 


rear  of  the  yolk  sac.  The  shape  and  relative  size  of  the  yolk  sac 
itself  are  variable  and  provide  additional  taxonomic  characters. 
In  summary,  although  the  yolk-sac  stage  starts  at  hatching 
and  ends  when  the  yolk  is  absorbed,  fish  are  at  different  stages 
of  development  with  regard  to  such  features  as  pigmentation, 
eye  development,  and  fin  formation  during  this  stage.  The  strik- 
ing pigment  rearrangements  that  occur  during  this  stage  provide 
further  emphasis  that  the  yolk-sac  stage  is  a  transitional  stage 
between  the  egg  and  larval  stages. 

The  Larval  Stage 

During  the  larval  stage  many  ontogenetic  changes  occur  (Mos- 
er. 1981).  Some  of  these  relate  directly  to  the  development  of 
the  adult  form  while  other  changes  and  structures  are  specialized 


KENDALL  ET  AL.:  ELH  STAGES  AND  CHARACTERS 


19 


B 


Fig.  10.  Apparent  convergence  in  siphonophore-mimicking  appendages  on  larval  fish.  (A)  Loweina  rara.  17.6  mm.  Note  lower  pectoral  fin 
ray  (Moser  and  Ahlstrom,  1970);  (B)  Carapussp.,  3.8  mm  (Padoa,  1956j).  Note  elongate  dorsal  fin  ray;  (C)  Exterilium  larva,  64  mm.  Note  trailing 
gut  (Moser,  1981);  (D)  Lopholus  sp.,  12.t  mm.  Note  elongate  dorsal  and  pelvic  ray  (Sanzo.  1940);  and  (E)  Arnoglossus  japonkus,  30.5  mm. 
Note  elongate  dorsal  ray  (Amaoka,  1973). 


and  of  presumed  functional  significance  primarily  for  planktonic 
existence  (Fig.  9).  These  latter  features  are  of  particular  interest 
in  systematic  studies  of  larval  fish  ontogeny.  They  include  pig- 
ment pattern,  larval  body  shape,  armature  on  head  bones,  and 
precocious  (early  forming),  elongate,  or  serrate  fin  spines.  The 
sequence  and  way  of  developing  adult  structures,  such  as  the 
skeleton  and  fin  rays,  are  also  useful  larval  characters.  All  of  the 
characters  of  the  larvae— whether  they  are  specialized  larval 
characters  or  merely  characters  observable  in  the  larvae— may 
have  potential  systematic  value  at  some  taxonomic  level;  how- 
ever, the  usefulness  of  most  of  the  characters  has  not  been  eval- 
uated (Tables  3  and  4). 

Among  the  most  taxonomically  useful  larval  characters,  gen- 
erally at  the  specific  or  generic  level,  is  the  pigment  pattern. 
Usually,  each  species  has  a  distinct  larval  pigment  pattern.  In 
some  the  number  and  placement  of  individual  melanophores 


are  diagnostic,  while  in  others  the  location,  shape,  and  size  of 
groups  of  melanophores  are  key  characters.  At  a  higher  taxo- 
nomic level,  in  the  myctophiforms  for  example,  the  peritoneal 
pigment  blotches  seem  to  indicate  relationships  on  a  suborder- 
family  level.  Problems  associated  with  the  usefulness  of  pigment 
patterns  include  1 )  the  widespread  distribution  of  some  patterns, 
and  2)  the  variable  state  of  melanophore  contraction  on  larvae 
of  the  same  species.  An  example  of  the  first  problem  is  the 
frequent  occurrence  of  a  row  of  small  melanophores  along  the 
ventral  midline  from  just  behind  the  anus  to  the  tip  of  the  tail. 
Another  example  is  a  pigmented  area  midlaterally  on  the  caudal 
peduncle  which  occurs  in  numerous  groups.  A  ventral  spot  at 
the  junction  of  the  cleithra  is  also  quite  common.  These  are  just 
a  few  examples  of  widespread,  presumably  convergent  pigment 
patterns  that  limit  the  usefulness  of  pigment  in  systematic  stud- 
ies of  larvae.  The  causes  for  the  observed  differences  in  degree 


20 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  11.     Liopropoma  sp.,  11.0  mm.  Collected  by  G.  R.  Harbison, 
16  May  1981,  6°31.8'S,  150°21.8'E.  Note  elongate  dorsal  spines. 


of  contraction  of  melanophores  are  not  well  understood  al- 
though they  may  be  partially  related  to  ambient  light  intensity. 
The  relative  size  and  placement  of  melanophores  are  genetically 
determined  and  therefore  useful  in  a  systematic  context,  while 
the  degree  of  contraction  seems  to  be  physiologically  deter- 
mined. 
In  general,  the  body  shape  and  size  at  various  stages  of  de- 


velopment are  characteristic  of  larvae  at  the  generic  or  familial 
level,  although  subtle  differences  in  body  shape  may  be  char- 
acteristic of  species.  Size  at  stage  of  development  can  be  envi- 
ronmentally modified  (e.g.,  by  temperature  or  food)  to  some 
extent,  but  is  primarily  genetically  determined.  There  appears 
to  be  some  convergence  in  larval  body  shape,  such  as  on  a  long 
tubular  body  in  several  divergent  groups  (e.g.,  Clupeiformes, 
Argentinidae,  Blennioidea),  just  as  there  is  on  the  "herring" 
morph  of  adults. 

A  valuable  and  fairly  widespread  set  of  larval  characters  con- 
cerns the  development  of  spines  and  armature  on  bones  of  the 
head  and  cleithral  region.  Such  armature  has  provided  diag- 
nostic larval  characters  as  well  as  material  for  systematic  infer- 
ence at  levels  from  species  to  order.  Larval  head  armature  ap- 
pears to  be  a  mark  of  the  Acanthopterygii.  Only  a  few  scat- 
tered examples  of  such  armature  appear  in  fishes  which  have 
only  soft  rays  as  adults  (e.g.,  Sudis).  Within  the  spiny-rayed 
fishes,  beryciforms  are  quite  heavily  armed  with  spines  on  many 
head  bones.  Perciforms  usually  do  not  have  spines  on  the  pa- 
rietals  but  the  supraoccipital  is  armed  in  some.  The  Scorpaeni- 
formes  are  just  the  opposite:  they  tend  to  have  head  armature 
that  includes  spines  on  the  parietals  but  do  not  have  spines  on 
the  supraoccipital. 

Nowhere  are  larval  specializations  more  evident  or  varied 
than  in  the  fins.  Elongation  of  particular  spines  or  soft  rays  or 
enlargement  of  whole  fins  are  frequently  seen.  Such  elongations 
have  been  described  for  rays  of  the  dorsal,  pelvic,  pectoral,  and 
caudal  fins;  thus  they  occur  with  both  spines  and  soft  rays.  In 
some,  these  long  rays  may  bear  pigmented  "bulbs"  or  appear 
like  flagellae.  Such  specialized  rays  are  produced  in  the  dorsal, 
pectoral,  or  pelvic  fins  of  taxonomically  diverse  fishes.  The  ex- 
tended gut  of  "exlerilium"  ophidioid  larvae  (Fraser  and  Smith, 
1974)  and  the  serial  pigment  pattern  of  some  leptocephali  (Smith, 
1979)  may  give  the  same  appearance  to  potential  predators  as 
these  elongate  rays.  All  of  these  structures  may  be  mimicking 
siphonophores:  a  remarkable  example  of  convergence  (Fig.  10 
and  1 1 ).  Elongate  fin  spines  are  heavy  and  armed  with  serrations 
in  some.  Elongated  rays  are  often  precocious  in  development, 
with  some  even  forming  in  the  egg.  These  fin  characters  seem 
to  vary  at  the  family-species  levels.  Other  characters  associated 
with  fin  development  include  the  sequence  of  formation  and 
movement  and  loss  of  whole  fins  or  some  of  the  rays.  Dorsal 
and  anal  fins  move  forward  along  the  body  during  larval  de- 
velopment in  elopiform  and  clupeiform  fishes.  They  develop  in 
"streamers"  in  the  finfold  of  argentinoids  and  attach  to  the  body 
proper  just  before  or  during  transformation.  The  shape  of  the 
finfold,  presence  or  absence  of  a  preanal  finfold,  and  shape  of 
the  pectoral  fin  base  provide  additional  characters  at  the  family- 
genus  level. 

Gut  characters  offish  larvae  include  length  and  shape  as  well 
as  the  development  of  a  protruding,  trailing  hindgut  in  some. 
In  fishes  with  pholophores,  their  placement  and  sequence  of 
development  are  excellent  characters  at  the  subfamily-species 
levels.  The  eye  of  a  larva  is  specialized  in  a  number  of  ways. 


Fig.  12.  Examples  of  special  juvenile  stages.  (A)  Hexagrammos  lagocephalus.  28.0  mm.  A  neustonic  or  epipelagic  form  of  a  species  that  is 
demersal  as  an  adult  (from  Kendall  and  Vinter,  1984);  (B)  Forapiger  longirosths.  17  mm.  A  spiny  form  that  lives  on  tropical  reefs  as  an  adult 
(from  Kendall  and  Goldsborough,  1 9 1 1 );  (C)  Sehaslolobus  altivetis,  26.8  mm.  A  barred  pelagic  form  of  a  species  that  is  demersal  on  the  continental 
slope  as  an  adult  (from  Moser  et  al.,  1977);  (D)  Oncorhynchus  kisulch.  37  mm.  The  freshwater  alevin  or  parr  stage  of  an  andromous  salmonid 
(from  Auer,  1982);  and  (E)  Kali  macrodon.  45  mm.  The  juvenile  of  a  bathypelagic  species.  Originally  described  as  Gargaropteron  pterodactylops 
(see  Johnson  and  Cohen,  1974). 


KENDALL  ET  AL.:  ELH  STAGES  AND  CHARACTERS 


21 


22 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Its  size  and  rate  of  development  are  useful,  as  well  as  whether 
it  is  round  or  oval.  Some  fish  larvae  have  eyes  borne  on  stalks 
that  reach  an  extreme  in  Idiacanthus,  while  others  develop  an 
area  of  choroid  tissue.  Migration  of  the  eye  in  flatfish  larvae 
from  a  symmetrical  position  to  one  side  of  the  head  is  well 
known.  The  sequence  of  development  of  ossified  structures  is 
proving  to  be  a  powerful  tool  in  systematic  studies  offish  larvae. 
The  losses  and  fusions  of  bones,  which  are  generally  assumed 
based  only  on  adult  material,  can  and  should  be  tested  using 
developmental  studies.  The  caudal  fin  skeleton  has  provided 
excellent  developmental  characters  to  be  used  for  systematic 
inferences,  mainly  at  the  order-generic  levels.  The  development 
of  scales  has  been  little  studied  but  may  prove  valuable,  espe- 
cially in  fishes  with  precocious  scales  (e.g.,  some  anthiins,  hol- 
ocentrids). 

The  Transformation  Stage 

Between  the  larval  and  juvenile  stages,  there  is  a  transitional 
stage  which  may  be  abrupt  or  prolonged  and  which,  in  many 
fish,  is  accompanied  by  a  change  from  planktonic  habits  to 
demersal  or  schooling  pelagic  habits  (Fig.  12).  In  some  fishes 
migration  to  a  "nursery"  ground  occurs  during  or  just  before 
this  stage.  Morphologically  the  transformation  stage  is  charac- 
terized by  a  change  from  larval  body  form  and  characters  to 
juvenile-adult  body  form  and  characters.  At  the  end  of  this  stage 
the  fish  generally  looks  similar  to  the  adult,  with  major  differ- 
ences only  in  pigmentation  patterns.  Two  ontogenetic  processes 
occur  during  this  stage  of  transition  between  the  larva  and  ju- 
venile: I )  loss  of  specialized  larval  characters,  and  2)  attainment 
of  juvenile-adult  characters.  Changes  that  occur  during  this  stage 
include  pigment  pattern,  body  shape,  fin  migration  (e.g.,  in  clu- 
peids  and  engraulids),  photophore  formation,  loss  of  elongate 
fin  rays  and  head  spines  (e.g.,  in  epinepheline  serranids  and 
holocentrids),  eye  migration  (pleuronectiforms),  and  scale  for- 
mation. 

In  several  groups,  where  the  transformation  stage  is  pro- 
longed, the  fish  have  developed  specializations  that  are  distinct 
from  both  the  larvae  and  juveniles.  This  stage  has  been  desig- 
nated the  prejuvenile  stage  (Hubbs,  1943).  The  specializations 
generally  involve  body  shape  and  pigmentation.  In  many,  the 
morph  resembles  a  herring-like  fish  and  is  apparently  adapted 
for  neustonic  life.  The  dorsal  aspect  of  the  fish  is  dark  green  or 
blue  and  the  lateral  and  ventral  is  silvery  or  white.  The  body 
tends  to  be  herring  shaped  and  the  mouth  terminal.  Fins  are 
generally  unpigmented.  Such  a  stage  is  present  m  Gadiformes 
(Urophycis),  Beryciformes  (Holocentrus),  Perciformes  (e.g.,  Po- 
malomus,  MuUidae,  Mugilidae)  and  Scorpaeni formes  (e.g., 
Scorpaenichthys,  Hexagrammos).  In  other  fishes,  such  as  some 
myctophiforms  and  carapids,  the  prolonged  transformation  stage 
may  have  distinctive  body  and  fin  shapes. 

Implications  of  Larval  Fish 
Morphology 

When  studying  the  appearance  of  larval  fishes,  one  is  im- 
mediately struck  with  their  diversity  and  morphological  dissim- 
ilarity to  adults.  This  dissimilarity  led  early  workers  to  establish 


names  for  several  of  these  forms,  not  realizing  that  they  were 
the  young  stages  of  known  adults.  After  establishing  the  identity 
of  many  fish  larvae  in  a  variety  of  groups,  we  hypothesize  that 
the  larvae  of  all  species  are  recognizably  distinct.  The  use  of 
diversity  of  larval  form  in  vertebrate  systematics  was  discussed 
some  time  ago  by  Orton  (1953b,  1955c,  1957)  and  in  this  vol- 
ume we  examine  this  use  in  detail  in  numerous  groups  of  fishes. 

Why  are  the  larvae  so  diverse?—  Despite  the  tremendous  mor- 
tality associated  with  living  in  the  planktonic  realm  during  the 
larval  period,  survival  must  be  sufficient  to  maintain  the  species 
and  provide  a  dispersal  mechanism  for  it.  To  different  degrees, 
various  taxa  apparently  rely  on  survival  and  longevity  of  in- 
dividual larvae.  The  amount  of  reliance  is  presumably  related 
to  fecundity  and  importance  of  dispersal  and  colonization  to 
the  taxon.  A  number  of  structures  have  evolved  that  would  be 
expected  to  enhance  larval  survival  in  the  plankton.  Practically 
no  experimental  work  has  been  done  to  investigate  the  function 
of  larval  structures,  but  some  structures  probably  assist  flotation 
and  feeding  while  others  decrease  predator  mortality.  Conver- 
gence on  characters  that  are  apparently  functionally  important 
to  larval  survival  in  the  plankton  is  seen.  These  specializations 
develop  in  conjunction  with  the  basic  ontogeny  of  the  taxon. 
In  studying  systematics  using  larval  fishes,  both  the  basic  pattern 
of  development  and  the  specialized  structures  must  be  analyzed. 

Why  are  these  larvae  so  morphologically  unlike  the  adults?— 
Most  larvae  are  adapted  to  survive  in  an  ecological  realm  (gen- 
erally the  plankton)  that  is  far  different  from  that  of  the  adult. 
These  are  small  organisms,  compared  to  adults,  and  they  live 
in  the  plankton,  having  to  find  and  capture  food  there  and  avoid 
becoming  food.  They  float  and  migrate  vertically  in  a  milieu 
that  may  be  moving  much  faster  than  they  are.  During  this 
larval  period,  these  fish  undergo  extreme  changes  in  morphology 
yet  remain  a  functioning  (eating,  avoiding  predators)  organism 
and  eventually  end  up  in  a  suitable  nursery  area  for  the  juvenile 
stage. 

How  then  can  larval  morphology  help  us  understand  the  evolu- 
tion of  these  fishes?—  Mler  recognizing  that  each  species  has  a 
morphologically  distinctive  larva,  generally  we  see  that  species 
of  the  same  genus  are  phenetically  similar,  and  larvae  of  mem- 
bers of  a  family  also  share  common  features.  Even  larvae  of 
suborders  and  orders  share  some  larval  characters.  This  would 
be  expected  since  evolution  operates  on  all  stages  in  the  life 
cycle,  not  just  the  adult.  Evolutionary  pressures  on  the  larval 
stage  seem  to  be  particularly  intense  in  those  groups  that  rely 
on  the  larvae  for  widespread  dispersal  in  the  ocean.  Here  the 
larvae  appear  well  adapted  for  life  in  the  planktonic  realm,  and 
it  can  truly  be  said  that  the  larva  and  the  adult  perform  in  "two 
quite  separate  evolutionary  theaters"  (Moser  and  Ahlstrom, 
1974).  In  this  volume  we  are  focusing  on  what  we  know  to  date 
about  larval  evolution  within  various  groups  of  fishes  (Table  4). 

Northwest  and  Alaska  Fisheries  Center,  2725  Montlake 
Blvd.  E.,  Seattle,  Washington  98112  and  Southwest 
Fisheries  Center,  P.O.  Box  271,  La  Jolla,  California 
92038. 


TECHNIQUES  AND  APPROACHES 


Early  Life  History  Descriptions 
E.  M.  Sandknop,  B.  Y.  Sumida  and  H.  G.  Moser 


FISHERIES  studies  require  accurate  identification  of  subject 
species.  Identification  of  the  developmental  stages  of  fishes 
is  complicated  by  the  small  size  of  the  specimens,  their  fragility, 
and  the  relatively  great  changes  in  their  structure  and  pigmen- 
tation. Experience  has  shown  that  major  changes  can  occur  over 
very  small  growth  increments  and  these  can  only  be  documented 
by  a  continuous  growth  series.  Published  descriptions  of  de- 
velopmental series  vary  in  quality,  perhaps  more  than  do  species 
descriptions  of  adults.  Prior  to  Bertelsen  (1951)  and  Ahlstrom 
and  Ball  (1954),  most  published  descriptions  were  based  on 
relatively  few  specimens,  which  were  described  individually.  In 
their  study  of  the  early  life  history  stages  of  the  jack  mackerel 
(Trachunts  syinmetricus),  Ahlstrom  and  Ball  (1954)  used  over 
500  eggs  and  a  series  of  about  250  larvae,  transforming  speci- 
mens, and  juveniles  to  describe  development.  Changes  in  struc- 
ture and  pigmentation  were  thus  described  as  a  dynamic  con- 
tinuum, with  emphasis  on  variation,  in  contrast  to  the  approach 
of  most  previous  workers.  Developmental  osteology  was  con- 
sidered an  integral  part  of  the  description  as  were  seasonal  and 
geographic  distributions  of  eggs  and  larvae.  This  paper  was  fol- 
lowed by  several  others  (Ahlstrom  and  Counts,  1955,  1958; 
Uchida  et  al.,  1958;  Kramer,  1960)  and  these  became  models 
for  subsequent  descriptive  papers,  including  some  which  treated 
several  species  in  various  taxonomic  groups  (Moser  and  Ahl- 
strom, 1970;  Ahlstrom,  1974;  Ahlstrom  et  al.,  1976;  Moser  et 
al.,  1977;  Kendall,  1979;  Brownell,  1979;  Richardson  and 
Washington,  1980;  Fahay,  1983;  Leis  and  Rennis,  1983).  The 
following  is  a  brief  account  of  the  elements  involved  in  preparing 
early  life  history  accounts  of  teleosts. 

Sources 

The  major  source  of  material  is  plankton  collections.  Typical 
survey  tows  strain  a  column  of  water  200  m  to  the  surface  and 
sample  eggs  and  subsequent  larval  stages  of  a  major  portion  of 
the  fish  fauna  (Smith  and  Richardson,  1 977).  Fishes  which  have 
highly  stratified  vertical  distributions  are  undersampled  by 
oblique  tows  and  require  special  gear  or  tow  strategies.  For 
example,  surface  dwellers  can  be  sampled  by  neuston  nets  (Zait- 
sev,  1970;  Nellen  and  Hempel,  1970;  Hempel  and  Weikert, 
1972;  Nellen,  1973a;  Ahlstrom  and  Stevens,  1976)  and  those 
species  residing  near  the  bottom  may  be  sampled  by  epi-benthic 
plankton  nets  (Schlotterbeck  and  Connally,  1 982).  Larger  larvae 
and  transforming  stages  are  poorly  sampled  by  typical  survey 
tows  principally  because  of  accumulated  mortality,  increased 
avoidance  capacity,  and  migration  out  of  the  sampling  zone. 
These  stages  are  more  effectively  sampled  by  trawls  (Tranter, 
1968),  dip-netting  with  attractor  lights  (Klawe,  1 960),  light  traps 
(Faber,  1982),  and  fish  predators  (Haedrich  and  Nielsen,  1966). 
Recently,  scuba  divers  have  collected  oceanic  larvae  with  their 
delicate  structures  intact  (Harbison  et  al.,  1978;  Govoni  et  al., 
1 984).  Developmental  series  may  also  be  obtained  by  rearing 


larvae  from  eggs  collected  at  sea  or  from  captive  brood  stock 
(Houdeetal.,  1970,  1974;  Houde  and  Swanson,  1975;  Richards 
etal.,  1974;  Houde  and  Potthoff,  1976;  Moser  and  Butler,  1981). 
This  method  becomes  essential  when  working  with  speciose 
faunas  (e.g.,  Sebastes,  warm  water  shorefishes),  if  only  to  de- 
termine which  species  cannot  be  identified. 

Use  of  Specimens 

The  characters  and  techniques  used  in  identifying  develop- 
mental stages  are  discussed  elsewhere  in  this  volume  (see  Ken- 
dall et  al.;  Matarese  and  Sandknop;  Powles  and  MarkJe).  From 
the  continuous  developmental  series  two  subseries  are  assem- 
bled and  these  form  the  basis  for  the  description.  The  first  series 
is  used  to  describe  morphology  and  pigmentation.  Specimens 
in  the  second  series  are  cleared  and  stained  by  a  variety  of 
techniques  to  describe  the  development  of  cartilaginous  and 
osseus  features  (Potthoff,  this  volume). 

The  number  of  specimens  used  to  construct  these  series  is 
dependent  on  several  factors:  1)  specimen  availability,  2)  length 
(duration)  of  the  development  period,  and  3)  complexity  of 
developmental  change.  A  guideline  is  that  there  should  be  enough 
specimens  to  demonstrate  the  beginning,  progression  and  com- 
pletion of  significant  developmental  changes  in  morphology  and 
pigmentation.  Usually  more  specimens  are  required  for  species 
which  have  extended  larval  periods;  however,  many  fishes  which 
transform  at  small  sizes  undergo  great  change  over  small  length 
intervals.  For  example,  lined  sole  {Achirus  lineatus)  hatch  at  1 .6 
mm,  transform  at  about  4.0  mm,  and  complete  a  large  suite  of 
developmental  changes  over  a  2.5  mm  length  interval  (Houde 
et  al.,  1 970).  The  majority  of  marine  teleosts  transform  between 
10  and  30  mm  and,  for  these,  major  developmental  events  can 
be  documented  by  specimen  length  increments  of  0.5-1.0  mm. 
Multiple  samples  representing  1  mm-intervals  are  required  to 
study  fine-scale  character  variation;  however,  such  studies  have 
rarely  been  done  (Ahlstrom  and  Moser,  1981). 

A  table  of  morphometric  measurements  constructed  from  the 
unstained  series  provides  data  on  the  size  at  important  devel- 
opmental milestones  (e.g.,  hatching,  notochord  flexion,  fin  for- 
mation, transformation)  and  provides  a  basis  for  analyzing 
structural  change  and  allometric  growth.  These  specimens  can 
be  used  to  construct  character  matrices  of  complex  or  diagnostic 
pigment  changes.  Illustration  specimens  chosen  from  the  series 
provide  an  integrated  view  of  major  characters  and  also,  if  ac- 
curately executed,  are  themselves  morphometric  and  meristic 
documents  (Sumida  et  al.,  this  volume). 

The  stained  series  is  used  to  construct  a  meristic  table  that 
forms  the  basis  for  following  the  development  of  fin  rays  and 
supporting  elements,  the  axial  skeleton  and  cranial  bones  (Dunn, 
this  volume).  Fine  bony  structures,  such  as  cranial  spines  are 
also  apparent  in  these  preparations. 

Published  descriptions  employing  these  basic  elements  are 


23 


24 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


the  basis  for  ontogenetic  studies  of  fishes.  These  are  essential 
for  the  identification  of  ichthyoplankton  collections,  and  also 
present  characters  for  systematic  analysis.  Data  provided  in 


these  descriptions  have  proved  useful  in  studies  of  the  physi- 
ology, behavior  and  ecology  of  the  early  stages  of  fishes. 

National  Marine  Fisheries  Service,  Southwest  Fisheries 
Center,  P.O.  Box  271,  La  Jolla,  California  92038. 


Synopsis  of  Culture  Methods  for  Marine  Fish  Larvae 
J.  R.  Hunter 


THE  objective  of  this  paper  is  to  provide  a  synopsis  of  present 
technology  for  small-scale  laboratory  culture  of  marine  fish 
larvae.  The  technology  of  marine  fish  culture  is  relevant  to  this 
book  because  it  is  one  of  the  best  ways  to  obtain  a  taxonomic 
series.  "Ahlie"  Ahlstrom  was  a  strong  proponent  of  this  ap- 
proach and  I  lectured  on  the  subject  at  his  request  for  his  courses 
on  larval  fish  systematics.  Marine  fish  culture  has  often  been 
reviewed  (May,  1970-,  Houde,  1972a;  Houde  and  Taniguchi, 
1979;  Shelboume,  1964;  Kinne,  1977)  and  many  additional 
references  may  be  found  in  the  previous  reviews.  The  key  feature 
of  my  review  is  that  it  is  a  condensed  practical  guide  and  key 
to  the  literature  for  beginners  interested  in  small-scale  laboratory 
culture  of  marine  fish  larvae;  culture  of  freshwater  fishes  is  not 
considered. 

Eggs 

Sources.  —  Pelagic  fish  eggs  can  be  obtained  from  plankton  tows, 
by  catching  ripe  fish  and  fertilizing  the  eggs,  and  by  induction 
of  spawning  of  laboratory  brood  stock. 

Let  eggs  taken  in  plankton  tows  stand  in  quart  bottles  for  0.5 
h,  then  remove  plankton  from  bottom  of  jar  and  add  fresh  sea 
water  (a  second  decanting  may  be  required).  Jars  are  stored  on 
their  sides  in  an  insulated  ice  box  with  a  refrigerant  for  24  h  or 
longer  with  the  temperature  kept  within  spawning  range. 

Virtually  all  marine  clupeoid  fishes  (Blaxter  and  Hunter,  1982) 
and  probably  most  other  pelagic  marine  fishes  spawn  at  night, 
hence  running  ripe  fish  are  more  common  at  night  or  just  before 
sunset  (final  egg  maturation  or  hydration  occurs  just  before 
spawning).  After  an  egg  is  spawned  in  sea  water  its  fertility 
decreases  but  the  maximum  time  for  it  to  become  infertile  is 
highly  variable  among  species,  varying  from  6  minutes  to  over 
3  hours  (Ginzburg,  1972).  Sperm  in  sea  water  may  remain  fertile 
for  days  (Ginzburg,  1972)  although  fertility  periods  as  short  as 
30  seconds  have  been  observed  (Haydock,  1971).  Owing  to  the 
great  variation  in  the  time  eggs  and  sperm  remain  fertile  it  is 
preferable  that  sperm  and  eggs  be  mixed  immediately  after  they 
are  obtained. 

Storage  of  gametes  may  be  helpful  since  mature  males  and 
females  are  not  always  available  simultaneously  and  crosses 
between  subpopulations  may  be  desired.  It  is  well  known  that 
sperm  can  be  stored  for  extended  periods  ( 10  or  more  hours)  if 
kept  cool  and  maintained  in  the  concentrated  form  and  not 
activated  by  sea  water  (Ginzburg,  1972;  Erdahl  and  Graham, 
1980).  Fertilization  of  Clupea  harengus  eggs  may  be  obtained 


after  6-7  days  dry  storage  at  4°  C  but  a  high  hatching  rate  is 
expected  only  after  periods  less  than  36  h  (Blaxter  and  Holli- 
day,  1963).  It  is  now  possible  to  extend  the  life  of  fish  sperm 
for  much  longer  periods  using  cryopreservation  techniques 
(-  196°C)  (Erdahl  and  Graham,  1980).  Various  cryoprotective 
agents  have  been  used  to  freeze  sperm  of  marine  fishes  including 
glycerol  (Blaxter  and  Holliday,  1963),  glucose,  NaCI,  Ringer's 
solution  and  fish  serum  (Hara  et  al.,  1982). 

The  stress  of  capture  causes  female  Katsiiwonus  pelamis  to 
ovulate  and  spawn  within  24  h  after  capture  but  eggs  are  often 
not  viable  (Kaya  et  al.,  1982),  Maturing  marine  fish  in  the  lab- 
oratory and  spawning  them  by  hormone  injections  has  become 
routine  in  recent  years  and  is  preferable  to  stress  techniques. 
Examples  include  Engraulis  mordax  (Leong,  1971),  Scomber 
japonicus  (Leong,  1977),  Chanos  chanos  (Liao  et  al.,  1979), 
Bairdiella  icistia  (Haydock,  1971),  Paralichthys  denial  us  and 
Pseudopleuronectes  americanus  (Smigielski,  1975a,  b)  and  oth- 
ers (see  review  of  Lam,  1982).  Induction  of  spawning  in  the 
laboratory  may  require  an  open  sea  water  system,  large  holding 
tanks  (e.g.,  -3  m  dia.  or  larger),  temperature  and  light  control. 

Handling  and  stocking.— To  count  eggs  without  damaging  them 
we  recommend  a  polished  wide  bore  (~3  mm)  pipette;  count 
30-50  late  stage  eggs  at  a  time  in  a  depression  slide  under  a 
dissection  microscope,  and  wash  eggs  off  the  slide  by  immersion 
of  the  entire  slide  in  sea  water.  Counting  eggs  is  critical  because 
higher  mortalities  and  slower  growth  result  from  excess  stocking 
densities  (Houde,  1975  and  1977).  As  a  rule  stocking  densities 
in  rearing  tanks  of  8  eggs/I  or  less  seems  preferable  and  most 
rearing  successes  have  occurred  when  stocking  did  not  exceed 
20  eggs/1  (Houde,  1975).  Similarly,  the  mortality  of  Mugil  ceph- 
a/(«  larvae  seems  to  remain  constant  (2-3%  loss/day)  at  stocking 
densities  of  1-30  larvae/1  (Kraul,  1983). 

Apparatus 

Containers  and  lighting.  —  Larvae  appear  to  grow  faster  and  show 
fewer  signs  of  starvation  when  reared  in  large  containers  (100 
1)  rather  than  in  smaller  ones  (10  1)  (Theilacker,  1980b).  Opti- 
mum container  size  doubtless  varies  with  species  but  40  1  con- 
tainers are  probably  the  minimum  size  that  should  be  used  and 
I  prefer  100-400  1  containers.  We  use  cylindrical  black  fiberglass 
containers  although  excellent  results  are  obtained  using  ordinary 
rectangular  glass  aquaria  (Houde,  1975). 

It  is  traditional  to  provide  a  daily  cycle  of  illumination  to 


HUNTER:  CULTURE  METHODS 


25 


larvae  in  rearing  containers  although  constant  illumination  is 
occasionally  used.  Typically  fluorescent  lamps  are  used  which 
provide  2,000-3,000  lux  at  the  water  surface  (Houde,  1978; 
Hunter,  1976).  Night  light  levels  vary;  we  provide  no  light  at 
night  whereas  Houde  (1978)  provides  a  dim  light  of  40-90  lux 
at  night,  which  is  substantially  above  the  visual  threshold  for 
feeding  for  larval  E.  morda.x  (6  mm  larvae  50%  feeding  thresh- 
old =  6  lux,  and  10-15  mm  larvae  50%  threshold  =  0.6  lux, 
Bagarinao  and  Hunter,  1983).  Clearly,  longer  periods  for  visual 
feeding  will  probably  enhance  growth  if  food  is  limited.  Rearing 
at  high  light  intensities  such  as  natural  sunlight  may  greatly 
increase  production  of  algae  and  zooplankton  in  the  culture  tank 
and  thereby  increase  larval  survival  (Kraul,  1983).  On  the  other 
hand,  solar  UV  radiation  is  clearly  lethal  to  younger  larvae 
(Hunter  etal.,  1 982)  and  use  of  deep  tanks,  or  shaded  or  covered 
tanks  (screen  cloth,  acrylic  plastic,  glass  or  mylar  film)  is  rec- 
ommended for  the  first  1-2  weeks  of  larval  life  if  tanks  are  to 
be  exposed  to  solar  radiation. 

Water  qualily.—C\osed,  non-circulating  systems  are  typically 
used  to  rear  marine  fish  larvae  at  least  during  the  younger  stages, 
because  in  an  open  system  planktonic  larvae  and  their  foods 
are  easily  lost.  Older  (nektonic)  larvae  are  able  to  resist  the 
current  and  to  consume  a  daily  ration  in  a  short  period  so  a 
partially  open  system  can  be  used.  We  fill  our  rearing  containers 
with  UV  treated  sea  water  that  is  passed  through  three,  in  line, 
cartridge  filters  (5,  3  and  1  ^m  pore).'  Although  not  a  common 
practice  in  small  scale  rearing  work,  the  addition  to  rearing  tanks 
of  antibiotics  (sodium  penicillin  G  at  50  i.u./ml  plus  strepto- 
mycin sulphate  at  0.05  g/ml)  slightly  improved  survival  of  Pleu- 
ronectes  platessa  eggs  through  hatching,  but  surprisingly  this 
single  treatment  greatly  improved  survival  of  larvae  through 
metamorphosis  (Shelboume,  1975). 

Use  of  a  closed  system  requires  attention  to  water  quality,  a 
problem  which  may  be  intensified  at  higher  rearing  tempera- 
tures. In  the  most  complete  study  of  water  quality  in  rearing 
tanks  for  marine  fish  larvae,  Brownell  (1980a,  b)  considered 
seven  variables  (pH,  dissolved  oxygen,  carbon  dioxide,  am- 
monia, nitrite  and  nitrate),  but  only  high  pH,  low  dissolved 
oxygen  and  un-ionized  ammonia  had  effects  at  levels  likely  to 
be  encountered  in  rearing  tanks.  First  feeding  incidence  declined 
by  50%  in  all  species  he  studied  when  dissolved  oxygen  con- 
centrations were  between  4  and  4.75  mg/1  (49-58%  saturation). 
Dissolved  oxygen  in  our  rearing  containers  usually  is  not  sat- 
urated after  planktonic  foods  are  added,  and  typically  it  is  about 
80%  saturation  even  with  aeration.  Clearly  water  quality  is  im- 
proved by  aeration  and  frequent  water  changes  and  lank  clean- 
ing. Werner  and  Blaxler  (1980)  exchanged  20%  of  the  water  in 
Clupea  harengus  cultures  (9°  C)  3  times  per  week  but  at  high 
temperatures  greater  replacement  rates  are  required.  For  ex- 
ample Houde  (1977)  replaced  20%  of  the  tank  sea  water  on 
alternate  days  while  culturing  Anchoa  mitchilli  and  Achirus  lin- 
eatus  at  26-28°  C.  Frequent  tank  cleaning  is  important  as  heavy 
mortalities  may  result  from  toxins  produced  by  debris  on  the 
container  bottom  (Kraul,  1983).  Aeration,  unless  very  gentle, 
can  cause  heavy  mortalities  among  delicate  eggs  and  newly 
hatched  larvae.  In  fact,  Shelboume  (1964)  recommends  no  aer- 


'  Aqua-Pure  model  APIO.  AMP  Cuno  Division,  Inc.,  Meriden.  Con- 
necticut USA. 


ation  for  Pleuronectes  platessa  larvae.  I  recommend  very  gentle 
aeration  but  not  until  a  week  or  so  beyond  the  first  feeding  stage. 
The  mortality  of  cultured  fish  larvae  often  increases  during 
the  period  of  initial  swim  bladder  inflation  in  physoclistous 
fishes  (Doroshev  et  al.,  1981;  Kuhlmann  et  al.,  1981)  and  this 
could  be  related  to  water  quality.  Symptoms  include  delay  or 
complete  failure  of  inflation  or  excessive  inflation;  in  either  case 
normal  swimming  patterns  are  disrupted  and  death  frequently 
results.  The  causes  of  abnormal  inflation  are  not  clear;  preven- 
tion of  larvae  from  reaching  the  water  surface  prevented  excess 
inflation  in  M.  cephalus  larvae  (Nash  et  al.,  1977),  whereas  the 
same  treatment  in  Atractoscion  nobilis  larvae  had  no  effect.  In 
A.  nobilis  excess  inflation  was  associated  with  abnormal  devel- 
opment of  gas  secretory  tissue  suggesting  a  more  complex  etiol- 
ogy (SWFC.  unpubl.  data).  Failure  to  inflate  the  swim  bladder 
is  a  common  problem  in  Morone saxatilus  culture  and  turbulent 
aeration  may  reduce  the  incidence  of  this  disease  (Doroshev  and 
Comacchia,  1979)  but  it  now  appears  that  reduction  in  salinity 
from  17  ppt  to  4  ppt  has  a  much  greater  eflect  in  reducing  the 
incidence  of  swim  bladder  malfunction  (S.  Doroshev  and  J. 
Merritt,  U.  Cal.  Davis,  pers.  comm.). 

Food 

The  most  critical  aspect  of  rearing  marine  larvae  is  manage- 
ment of  their  food.  Food  must  be  the  correct  density,  size, 
nutritionally  adequate  and  must  remain  suspended  in  the  water 
column  which  usually  requires  the  use  of  living  pelagic  organ- 
isms. 

Food  size.— Typ\c&\  pelagic  fish  larvae  are  2.5-4.0  mm  when 
they  begin  feeding  and  acceptable  prey  are  20-1 50 /um  in  breadth 
(Houde  and  Taniguchi,  1979).  Some  large  larvae,  e.g..  larval  C. 
harengiis  (B\di\\.QT.  1965).  Pleuronectes  platessa  {Riley.  1966)  or 
small  larvae  with  large  mouths,  e.g.,  Merluccius productus  {Sum- 
ida  and  Moser,  1980),  can  begin  feeding  on  prey  300  Mm  or 
larger  in  breadth.  The  optimal  food  size  increases  as  larvae  grow 
(Hunter,  1981),  so  any  culture  technique  should  provide  a  stead- 
ily increasing  range  of  food  sizes,  because  if  the  food  is  too  small 
growth  slows  and  mortality  occurs  (Hunter,  1981).  Food  size 
requirements  can  be  expressed  in  terms  of  the  ratio  of  prey  width 
to  mouth  width.  The  50%  threshold  for  feeding  on  a  prey  of  a 
particular  width  occurs  when  this  ratio  is  about  0.75,  although 
occasionally  larvae  consume  prey  as  wide  as  the  width  of  their 
mouth  (ratio  =  1)  (Hunter,  1981).  At  the  onset  of  first  feeding 
a  small  prey  of  about  'A  the  mouth  width  seems  to  be  preferable 
as  capture  success  is  low  at  this  time  but  within  a  few  days  larvae 
are  able  to  consume  food  of  about  V2  the  mouth  width. 

Wild  zooplankton— V/i\d  zooplankton,  primarily  the  naupliar 
and  copepodite  stages  of  marine  copepods  but  also  mollusc 
veligers,  tintinnids,  cladocera,  and  appendicularia  larvae,  are 
the  natural  foods  of  most  marine  fish  larvae  and  probably  also 
the  best  source  of  food  for  rearing  a  larval  taxonomic  series. 
Wild  zooplankton  provide  a  wide  range  of  sizes  and  types  and 
are  probably  nutritionally  superior  to  cultured  rotifers  and  Ar- 
lemia  nauplii  (Kuhlmann  et  al.,  1981).  Collection  of  wild  zoo- 
plankton may  require  less  effort  than  production  of  cultured 
food  except  for  brine  shrimp  nauplii  (see  below).  Zooplankton 
is  collected  in  nets  of  about  50  ^m,  and  is  graded  by  size  in  the 
laboratory  using  various  nylon  nets  (Houde,  1977,  1978),  This 
eliminates  the  larger  zooplankton  which  larvae  would  be  unable 


26 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


to  consume  and  which  may  be  larval  predators.  Fish  larvae, 
particularly  yolk-sac  stages,  are  vulnerable  to  various  carnivo- 
rous copepods,  amphipods,  euphausiids  and  chaetognaths 
(Hunter,  1981). 

Cultured  foods.— T-wo  cultured  foods,  the  rotifer  Brachiomts 
plicatilis,  and  nauplii  of  the  brine  shrimp,  Arteinia.  should  be 
considered  as  potential  foods  for  rearing  marine  fish  larvae  as 
many  fish  larvae  can  be  reared  on  a  combination  of  these  two 
foods.  These  two  foods  may  also  be  used  as  a  supplement  to 
diets  of  wild  plankton.  Groups  of  fishes  that  have  been  reared 
to  metamorphosis  on  a  combination  oi Brachionus  and  Anemia 
or  on  Artemia  alone  include  C.  harengns,  species  of  serranids, 
scombrids,  atherinids,  various  flatfishes,  sciaenids,  and  saganids 
(May,  1970;  May  etal.,  1974;  and  unpubl.  SWFC  data). /lr?ew;a 
nauplii  are  recommended  only  for  larvae  with  differentiated  guts 
as  they  are  quite  resistant  to  digestion  whereas  copepods  are  not 
(Rosenthal,  1969). 

Methods  for  culturing  rotifers  using  algae  are  given  by  Thei- 
lacker  and  McMaster  (1971);  culture  methods  employing  for- 
mulated artificial  diets  or  freeze  dried  algae  (Gatesoupe  and 
Robin,  1981;  Gatesoupe  and  Luquet,  1981)  and  ones  using 
brewers  yeast  also  exist.  Many  of  the  essential  facts  given  in 
these  original  papers  will  not  be  repeated  here  but  I  will  point 
out  a  few  practical  points  regarding  rotifer  culture  using  algae. 
Suitable  algae  species  for  rotifer  culture  include  Dunaliella, 
Nannochloris,  Tetraselmis,  and  Chlorella  which  may  be  grown 
using  standard  culture  media  (Guillard,  1975)  or  using  liquid 
commercial  plant  fertilizers  (dosage  for  fertilizer  containing  8% 
total  nitrogen  =  0. 1  ml  of  fertilizer/1;  dosage  among  brands  is 
adjusted  depending  on  total  N  content).  We  prefer  commercial 
plant  fertilizers  that  have  an  organic  base  such  as  liquid  fish 
fertilizers  and  avoid  those  that  have  soil  penetrants.  A  daily 
doubling  rate  can  be  expected  in  healthy  rotifer  cultures,  and 
cultures  can  be  maintained  for  weeks  or  even  months  by  adding 
fresh  algae  or  nutrients  and  sea  water,  although  single  batch 
harvesting  after  about  2  weeks  gives  more  dependable  results. 
Rotifers  are  harvested  using  gravity  flow  through  a  nylon  filter 
(20-40  ^m  mesh)  as  pumps  may  kill  rotifers. 

Production  ofArlemia  nauplii  is  simple  since  all  that  is  needed 
is  to  hatch  the  cysts  ("Anemia  eggs").  Cysts  from  a  variety  of 
strains  of  Anemia  are  commercially  available.  The  strains  differ 
considerably  in  average  naupliar  size  (423-775  ^m  length),  in 
pesticide  content  (DDT,  PCB,  and  chlordane)  and  in  certain 
fatty  acids  (Klein-MacPhee  et  al.,  1982).  These  authors  show 
that  very  low  survival  (15%)  of  P.  amehcanus  larvae  occurred 
when  they  were  fed  San  Pablo  Bay  (San  Francisco)  nauplii 
whereas  survival  of  larvae  fed  other  strains  varied  from  60- 
80%.  Beck  et  al.  ( 1 980)  gave  similar  results  for  Menidia  menidia 
larvae.  Of  all  the  strains  tested  in  these  papers  the  Australian 
and  Brazilian  strains  seem  the  most  suitable  for  rearing  larvae 
and  the  San  Pablo  Bay  (USA)  the  least. - 

Anemia  hatcheries  vary  from  a  jar  to  complex  automated 
systems.  The  J.  D.  Riley  Anemia  hatching  box  has  been  used 
with  slight  modification  in  many  laboratories  for  over  20  years. 
It  is  a  sea  water  filled  box  separated  in  half  by  a  sliding  partition; 
Anemia  cysts  are  added  to  one  side  (I  g/l)  and  they  hatch  1-2 


^  Exotic  Anemia  cysts  are  available  from:  Artemia  Inc.,  P.O.  Box 
2891,  Castro  Valley,  California  94546  USA  and  Biomarine  Research. 
4643  W.  Rosecrans,  Hawthorne,  California  90250  USA. 


days  later  depending  on  the  temperature  selected  (23-30°  C). 
The  tank  is  then  illuminated,  the  partition  raised  slightly  off  the 
bottom,  and  the  nauplii,  attracted  by  the  light,  swim  beneath 
the  partition  leaving  behind  the  hatching  debris  and  unhatched 
cysts  (Shelboume,  1964).  A  semiautomatic  version  of  this  sys- 
tem is  described  by  Nash  (1973),  and  various  other  improve- 
ments in  aeration,  illumination,  temperature,  and  other  factors 
have  increased  yields  to  lO'  nauplii  per  4.8  g  of  cysts  (San 
Francisco  Bay  Brand)  (Dye,  1 980).  In  recent  years  decapsulation 
of  Anemia  cysts  using  hypochlorite  bleach  has  become  popular 
because  it  increases  yields,  increases  the  dry  weight  of  the  nau- 
plius  (Bruggeman  et  al.,  1 980)  and  eliminates  contamination  of 
larval  fish  rearing  tanks  with  unhatched  cysts. 

It  should  also  be  noted  that  freshly  hatched  Anemia  nauplii 
are  clearly  more  nutritious  than  older  starving  individuals  and 
consequently  new  batches  should  be  frequently  produced.  In 
general,  prey  with  full  stomachs  are  probably  nutritionally  pref- 
erable to  ones  with  empty  stomachs.  Similarly,  more  Dicen- 
trarchits  labrax  larvae  seem  to  survive  when  rotifers  are  nutri- 
tionally enhanced  by  30  min  immersion  in  a  solution  containing 
vitamins  and  soluble  proteins  (Gatesoupe  and  Luquet,  1981). 

Mass  culture  of  marine  copepods  is  difficult  and  laborious 
and  therefore  not  recommended  when  a  taxonomic  series  is  the 
sole  objective.  Nevertheless,  culture  of  marine  copepods  may 
be  the  only  way  some  fish  larvae  can  be  reared  if  wild  zooplank- 
ton  is  not  readily  available  and  larvae  die  when  fed  Anemia 
nauplii  (rarely  are  more  than  a  single  strain  of  Anemia  tested, 
however).  Harpacticoid  copepods  (Tignopus  sp.,  Tishe  sp.,  and 
Euterpina  sp.)  are  the  most  frequently  used  copepods  because 
of  ease  of  culture;  for  culture  techniques  see  Kahan  et  al.  (1982) 
and  Hunter  (1976).  Euterpina  may  be  preferable  to  Tignopus 
or  Tishe  because  the  nauplii  and  copepodites  of  Euterpina  are 
pelagic  and  therefore  available  to  the  larvae  whereas  nauplii  and 
copepodites  of  Tigriopus  and  Tishe  tend  to  remain  on  surfaces 
and  are  therefore  less  available  (Kraul,  1983).  See  Nassogne 
(1970)  and  Zurlini  et  al.  (1978)  for  laboratory  culture  of  Euter- 
pina. 

Eood density.  —The  optimal  food  density  for  fish  larvae  depends 
upon  the  size  of  the  food  organism  and  size  or  age  of  the  larvae. 
Densities  of  1-3  organisms/ml  have  been  routinely  used  for 
larvae  fed  wild  zooplankton  (largely  copepod  nauplii)  during 
the  first  1-2  weeks  of  feeding  (Houde  and  Taniguchi,  1979). 
The  same  density  range  is  used  when  cultured  .Anemia  nauplii 
are  the  food.  A  higher  density  range  (IO-20/ml)  is  used  for 
cultured  B.  plicatilis  which  are  about  1/10  of  the  weight  of  an 
.irtemia  nauplius  (Theilacker  and  McMaster,  1971).  A  very 
small  food  particle,  the  dinoflagellate  Gymnodinium  splendens 
(40  nm  dia),  is  used  for  the  first  2  days  of  feeding  in  northern 
anchovy  larvae  (Lasker  et  al.,  1970;  Hunter,  1976)  at  a  high 
density  of  about  lOO/ml.  In  very  active  species  such  as  S.  ja- 
ponicus  or  the  siganid  Siganus  canaliculatus  high  food  densities 
can  cause  heavy  mortality  because  of  overfeeding  since  most 
larval  fishes  seem  to  lack  a  satiation  mechanism  (May  et  al., 
1974;  Hunter,  1981).  Overfeeding  seems  to  occur  only  when 
such  easily  captured  prey  as  .irtemia  nauplii  are  used  as  food. 

Piscivorous  fish  /arvac  — Piscivorous  fish  larvae  such  as  the 
scombroids,  Sphyraena  and  others  pose  special  problems  in 
culture.  Fish  larvae  are  an  ideal  food  for  such  larvae;  in  fact, 
our  only  success  in  rearing  Katsuwonus  pelamis  larvae  to  meta- 
morphosis was  probably  related  to  an  abundant  supply  of  yolk- 


HUNTER:  CULTURE  METHODS 


27 


sac  fish  larvae  as  food.  Zooplankton  is  the  initial  food  until 
piscivorous  feeding  habits  develop  (Houde,  1972b;  Mayo,  1973; 
Hunter  and  Kimbrell,  1980).  Piscivorous  larvae  manipulate  their 
larval  prey  and  consequently  are  less  dependent  on  mouth  size 
when  consuming  larval  fish.  Sibling  cannibalism  is  common 
under  reanng  conditions  in  such  fishes.  Increasing  the  food  den- 
sity may  increase  survival  as  may  elevating  the  temperature, 
thereby  accelerating  growth  through  the  most  cannibalistic  sizes; 
at  least  in  scombroids  sibling  cannibalism  declines  at  meta- 
morphosis (Mayo.  1973;  Hunter  and  Kimbrell,  1980).  Sorting 
by  size  and  isolating  the  larger  larvae  is  probably  the  only  certain 
method  for  controlling  losses  due  to  cannibalism,  however. 

Phytoplankton 

Phytoplankton  blooms  are  often  maintained  in  larval  culture 
tanks  to  reduce  the  detrimental  effects  of  metabolic  by-products 
which  accumulate  in  static  rearing  tanks  (Houde,  1974)  and  to 
provide  food  for  larval  food  organisms.  In  many  cases  dense 
blooms  of  phytoplankton  enhance  larval  growth  and  survival 
and  I  recommend  the  practice  but  the  mechanism  is  obscure. 
The  phytoplankters  used  are  various,  easily  grown,  small  species 
such  as  Chlorella.  Anacystis,  Nannochloris,  Tetraselmis.  Dun- 
aliella.  Isochrysis.  Phaeodactylum  and  others.'  They  are  main- 
tained at  high  densities  (10,000  or  more  cells/ml)  in  the  rearing 
tanks.  At  high  cell  densities  larvae  ingest  these  small  phyto- 
plankters, perhaps  inadvertently  (Moffatt,  1981)  but  they  appear 
not  to  be  able  to  exist  on  them  as  a  sole  food  source  (Houde, 
1974;  Scura  and  Jerde,  1977).  They  may  supplement  the  food 


'  For  a  nominal  fee  starter  cultures  of  manne  phytoplankton  can  be 
obtained  from  R.  R.  L.  Guiliard.  Bigelow  Laboratory  for  Ocean  Sciences. 
McKown  Point,  West  Boothbay  Harbor,  Maine  04575  USA;  culture 
methods  are  discussed  by  Guiliard  (1975). 


ration  either  directly  or  indirectly  through  the  ingestion  of  prey 
having  guts  full  of  algal  cells  (Moffatt,  1981).  Evidence  now 
exists  that  enhancement  of  growth  and  survival  of  larval  Scoph- 
ihalmus  maximiis  by  blooms  of  Isochrysis  and  Phaeodactylum 
is  due  to  the  inclusion  in  the  diet  of  certain  polyunsaturated 
fatty  acids  not  occurring  in  the  normal  laboratory  rotifer  diet 
(Scott  and  Middleton,  1979).  It  is  interesting  in  this  regard  that 
Dunaliella  which  lacks  the  fatty  acids  did  not  enhance  S.  max- 
imiis larval  growth  or  survival. 

Effects  of  Culture 

Extrapolation  from  cultured  larvae  to  natural  populations  must 
be  done  with  caution  because  culture  may  affect  the  morphology, 
behavior  and  biochemistry  of  larvae  (Blaxter,  1976).  The  mor- 
phological characteristics  most  susceptible  to  modification  in 
tanks  are  those  partially  controlled  by  environmental  conditions 
such  as  vertebrae  and  fin  ray  counts.  Reared  larvae  also  may 
be  more  heavily  pigmented  than  sea  caught  specimens  (Watson, 
1982).  This  appears  to  be  related  to  the  expanded  nature  of  the 
melanophores,  not  to  added  numbers  of  pigment  cells.  In  ad- 
dition, pigmentation  events  may  occur  at  smaller  sizes  in  reared 
material  (S.  Richardson,  Gulf  Coast  Research  Laboratory,  Ocean 
Springs,  Mississippi,  pers.  comm.).  Laboratory  reared  larvae  are 
often  heavier  and  have  deeper  bodies  than  their  wild  counter- 
parts, making  some  morphometric  measurements  on  laboratory 
specimens  useless  (Blaxter,  1975).  The  differences  in  preserva- 
tion and  handling  between  laboratory  and  sea-caught  larvae  also 
make  direct  size-specific  comparisons  difficult.  Shrinkage  in 
length  may  vary  greatly  depending  on  the  duration  larvae  re- 
main in  plankton  nets  and  shrinkage  differences  between  reared 
and  wild  specimens  can  be  misinterpreted  as  morphological 
differences  (Theilacker,  1980a). 

National  Marine  Fisheries  Service,  Southwest  Fisheries 
Center,  P.O.  Box  271,  La  Jolla,  California  92038. 


Identification  of  Fish  Eggs 
A.  C.  Matarese  and  E.  M.  Sandknop 


A  wide  variety  of  egg  types  exists  among  teleost  fishes  in  both 
freshwater  and  marine  environments.  Eggs  may  be  pelagic 
and  nonadhesive  or  demersal  and  either  adhesive  or  not.  They 
may  possess  a  variety  of  specialized  structures  aiding  in  flotation 
or  attachment.  Depending  on  egg  type  and  associated  repro- 
ductive ecology,  many  characters  are  useful  in  identification. 
These  characters  have  been  reviewed  for  pelagic  marine  eggs  by 
Rass(1973),  Robertson  (1975a),  Russell  (1976),  and  Ahlstrom 
and  Moser  ( 1 980);  we  have  liberally  and  extensively  drawn  from 
the  latter.  Important  characters  for  other  egg  types  have  been 
discussed  in  part  by  Balon  (1975a,  1981a),  Hardy  (1978a,  b), 
Jones  et  al.  (1978),  and  Snyder  (1981).  Characters  such  as  size 
and  possession  of  oil  globules  are  important  for  all  types;  how- 
ever, perivitelline  space  and  chorion  sculpturing  are  more  im- 
portant in  pelagic  eggs,  while  in  demersal  eggs  special  coatings. 


chorion  thickness,  or  nature  of  egg  deposition  may  be  more 
useful. 

A  wealth  of  potential  characters  useful  in  egg  identification 
exists;  however,  it  is  still  difficult  to  identify  eggs  of  most  species 
with  certainty.  Except  for  late  stages,  few  may  be  recognized  at 
the  species  level.  Some  characters  are  useful  at  a  family  level, 
but  presently  it  is  not  productive  to  speculate  on  the  systematic 
significance  of  any  characters  (see  Kendall  et  al.,  this  volume). 
Presently,  the  main  goal  of  taxonomy  with  respect  to  fish  eggs 
is  identification. 

Regardless  of  egg  type  or  reproductive  ecology,  a  summary 
of  identification  characters  useful  to  an  egg  taxonomist  is  pre- 
sented. Additionally,  we  recommend  using  available  literature 
for  reference  and  encourage  the  building  of  local  fish  egg  col- 
lections. We  follow  Ahlstrom  and  Ball  (1954)  in  subdividing 


28 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


1.34x0.66 

Engraulis  mordax 


B 


1.0x1.06 

Ophidion   scrippsae 


Unidentified 


0.58-0.74 

Vinciguerria  lucetia 


1.9 


Glyptocephalus  zachirus 


0.80 

Symphurus  atricauda 


H 


Prionotus  stephanophrys 


2.92 

Icosteus  aenigmaticus 


1.35 

Etrumeus  teres 


Fig.  13.  Fish  eggs.  Captions  under  each  illustration  indicate  the  species  and  the  diameter  or  dimensions  of  the  egg  in  millimeters.  A.  Engrauli 
mordax.  original;  B.  Ophidwn  scrippsae.  onginal;  C.  Unidentified,  original;  D.  Vincigiierna  tucetia.  from  Ahlstrom  and  Counts  (1938);  E 
Glyptocephalus  zachirus.  from  Ahlstrom  and  Moser  (1980);  F.  Symphurus  atricauda.  original;  G.  Prionotus  stephanophrys.  onginal;  H.  Icostei. 
aenigmaticus,  original;  and  I.  Etrumeus  teres,  original. 


'is 

E. 

•osleus 


MATARESE  AND  SANDKNOP:  EGG  IDENTIFICATION 


29 


egg  development  as  follows:  Early— from  fertilization  to  closure 
of  blastopore.  Middle— from  closure  of  blastopore  to  tail  bud 
lifting  off  yolk,  and  Late  — from  tail  bud  lifting  off  yolk  to  time 
of  hatching. 

Identification  Characters 
Shape.— The  vast  majority  of  all  egg  types  are  spherical.  Ex- 
ceptions include  ellipsoidal  eggs  as  found  in  anchovies,  En- 
graulis  and  Anchoa.  and  slightly  flattened  or  ovoid  eggs  as  seen 
in  members  of  the  families  Gobiidae,  Scaridae,  and  Ophidiidae 
(Fig.  13A.  B).  A  number  of  demersal  eggs  have  somewhat  ir- 
regular shapes,  especially  those  associated  with  large  egg  masses. 
The  perciform  family  Congrogadidae  has  cruciform  shaped  eggs 
(Herwig  and  Dewey,  1982).  An  unidentified,  star-shaped  egg  is 
encountered  infrequently  in  the  Alaska  region  (Fig.  13C). 

Size.—T\\t  average  marine  and  freshwater  fish  egg  size  is  about 
1.0  mm.  According  to  Ahlstrom  and  Moser  (1980),  pelagic  fish 
eggs  range  from  0.5  mm  [Mncigiicnia  (Fig.  13D)]  to  about  5.5 
mm  (Muraenidae).  Demersal  eggs  may  range  higher  in  size  (up 
to  7.0-8.0  mm),  e.g.,  members  of  the  families  Salmonidae,  An- 
arhichadidae,  and  Zoarcidae.  Mouth  brooders,  e.g.,  in  the  catfish 
family  Ariidae,  have  among  the  largest  eggs  with  sizes  from  1 4 
mm  to  26  mm. 

Oil  globules.— The  oil  globule  provides  useful  characters  in  fish 
egg  identification;  these  include  presence  or  absence,  number, 
size,  position,  color,  and  pigmentation.  Among  both  pelagic  and 
demersal  eggs,  the  most  common  form  contains  a  single  oil 
globule.  Eggs  may  lack  an  oil  globule  as  in  most  gadines  and 
pleuronectids  (Glyplocephaliis).  contain  only  one  (Icosteiis),  or 
have  multiple  oil  globules  as  in  the  cynoglossids  and  triglids 
(Symphums  and  Prionotus)  (Fig.  13E,  F,  G,  and  H).  In  pelagic 
eggs  with  a  single  oil  globule,  the  size  ranges  from  <0.10  mm 
to  >  1.0  mm  (Ahlstrom  and  Moser,  1980).  The  position  of  the 
oil  globule  within  the  yolk  sac  is  usually  posterior,  but  several 
groups  contain  species  that  have  an  anterior  placement  (e.g., 
labrids  and  carangids)  and  others  have  an  intermediate  place- 
ment (argentinids).  In  some  fishes,  oil  globules  migrate  during 
embryonic  development.  Some  members  of  the  family  Bathy- 
lagidae  initially  possess  multiple  oil  globules  that  eventually 
coalesce  into  a  single  globule  (Ahlstrom,  1969).  Although  not  a 
totally  reliable  character,  the  oil  globule  color  can  be  useful, 
especially  in  the  identification  of  freshly  taken  demersal  eggs. 
Lastly,  many  species  have  oil  globules  with  melanistic  pigment, 
Icosteus  (Fig.  13H)  and  Icichthys. 

Yolk.— The  degree  of  yolk  segmentation  is  an  important  iden- 
tification character.  Yolk  is  usually  segmented  in  primitive  forms, 
e.g.,  Etruineus  (Fig.  131),  and  homogeneous  in  higher  forms 
(Rass,  1973;  Ahlstrom  and  Moser,  1980).  The  opaqueness  of 
yolk  found  in  catfishes,  salmonids,  and  gars  can  be  diagnostic' 
Pigment,  which  may  also  be  diagnostic,  can  be  present  dunng 
various  developmental  stages  from  middle  to  late.  Yolk  color 
is  often  important  especially  in  demersal  eggs.  Among  demersal 
eggs  vitelline  circulation  patterns  within  the  yolk  sac  are  useful 
in  identification.' 


'  P.  Douglas  Martin,  Chesapeake  Biological  Laboratory,  P.O.  Box  38, 
Solomons,  Maryland  20688.  Personal  communication,  October  1982. 


Chorion.  — A.  number  of  characteristics  associated  with  the  cho- 
rion or  egg  envelope  can  be  useful  in  identifying  fish  eggs  and 
have  been  shown  to  be  highly  adapted  to  the  environmental 
conditions  under  which  an  embryo  develops  (Ivankov  and  Kur- 
dyayeva,  1973;  Stehr  and  Hawkes,  1979;  Laale,  1980;  Stehr, 
1982).  The  most  important  character  of  the  chorion  is  whether 
it  is  smooth,  as  is  in  most  fishes,  or  sculptured.  Among  fish  eggs 
with  patterns,  the  size  and  texture  (e.g.,  raised  hexagons,  pus- 
tules) of  the  design  are  diagnostic.  Raised  polygonal  surfaces  are 
found  in  several  unrelated  species  (Stehr,  1982),  e.g.,  Synodus 
and  Pleuronichthys  (Sumida  et  al.,  1979),  and  pustules  occur 
among  some  bathylagids  and  argentinids.  Mugil  cephalus  eggs 
(Fig.  14A),  previously  considered  to  have  a  smooth  chorion, 
have  a  raised  patterned  surface  visible  by  scanning  electron 
microscope  (Boehlert,  this  volume).  In  many  groups  of  fishes, 
the  chorion  has  various  degrees  of  ornamentation  consisting  of 
projections,  threads,  filaments,  or  stalks  which  may  aid  in  flo- 
tation (pelagic)  or  attachment  (demersal).  In  some  scombere- 
socids,  e.g.,  Cololahis  (Fig.  14B).  some  exocoetids  and  ather- 
inids,  pelagic  eggs  are  attached  to  each  other  or  to  a  substrate 
by  filaments.  Spines  are  found  in  some  myctophiforms  and 
exocoetids,  and  stalks  occur  in  some  demersal  egg  groups,  e.g., 
blenniids  and  Osmerus  mordax.  In  ostraciid  eggs,  a  patch  of 
pustules  is  present  near  the  micropyle  (Fig.  14C). 

Recently,  thickness  of  the  chorion  has  been  of  diagnostic  value 
(Ivankov  and  Kurdyayeva,  1973;  Boehlert,  this  volume).  Stehr 
and  Hawkes  (1979),  using  scanning  electron  microscopy,  found 
that  most  marine  teleosts  with  pelagic  eggs  have  thin  chorions 
in  relation  to  egg  diameter  whereas  demersal  eggs  tend  to  de- 
velop much  thicker  chorions.  Color  of  the  chorion  is  an  im- 
portant diagnostic  character,  especially  for  freshly  taken  de- 
mersal eggs  in  the  marine  intertidal  environment  (Matarese  and 
Marliave,  1982).  A  number  of  freshwater  demersal  fishes  have 
eggs  that  possess  a  special  coating  associated  with  the  chorion 
which  can  be  either  gelatinous  or  adhesive,  e.g.,  Perca.  Icialurus, 
and  Notropis  (Snyder,  1981). 

Penvilelline  space.  —  Most  fish  eggs  have  a  narrow-  to  medium- 
width  perivitelline  space,  but  wide  spaces  are  common  in  some 
groups,  especially  among  the  more  primitive  fishes  that  have  a 
segmented  yolk,  e.g.,  Clupeiformes  (Sardinops.  Fig.  14D),  An- 
guilliformes,  and  Salmoniformes  (Chauliodus.  Fig.  14E)  (Ahl- 
strom and  Moser,  1980).  Large  perivitelline  spaces  are  also  found 
among  some  unrelated  higher  forms,  such  as  cypnnids  (Nolro- 
pi.s).  percichthyids  (Morone  saxatill.s).  or  pleuronectids  (Hip- 
poglossoides). 

Embryonic  characters.— CharacXers  associated  with  the  devel- 
oping embryo  are  extremely  useful  in  egg  identification,  partic- 
ularly in  the  middle  and  late  stages  of  development.  Many  eggs 
not  identifiable  in  the  early  stages  are  easily  recognizable  using 
embryonic  characters  such  as  pigment  on  embryo  or  finfold  and 
morphology.  In  some  fishes,  embryonic  pigment  in  the  late  stages 
has  already  undergone  sufficient  migration  and  rearrangement 
to  the  point  where  it  resembles  the  yolk-sac  larva;  this  is  com- 
mon in  several  groups  including  gadiformes,  e.g.,  Merluccius 
(Fig.  14F),  Gadus.  and  Theragra.  and  heavily  pigmented  flat- 
fishes like  Pleuronichthys  and  Hypsopsetta.  Characteristic  late- 
stage  pigment  bands  appear  in  Glyptocephalus  (Fig.  13E).  In 
most  freshwater  species,  pigment  is  not  present  prior  to  pigment 
cell  migration  but  appears  sometime  after  the  cells  have  mi- 


30 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


0.76-0.80 

Mugil  cephalus 


B 


1.7x1.9 

Cololabis  saira 


1.54x1.68 

Ostraciidae 


1.35-2.05 

Sardinops   sagax 


2.93 

Chauliodus  macouni 


1.07-1.18 

Merluccius  productus 


H 


2.0 


Eumicrotremus  orbis 


2.65-2.90 

Trachipterus  altivelus 


0.88 

Stomias  atri venter 


Fig.  14.  Fish  eggs.  Captions  under  each  illustration  indicate  the  species  and  the  diameter  or  dimensions  of  the  egg  in  millimeters.  A.  Mugil 
cephalus.  original;  B.  Cololabis  saira.  original;  C.  Ostraciidae,  original;  D.  Sardinops  sagax.  original;  E.  Chauliodus  macouni.  original;  F.  Merluccius 
productus.  from  Ahlstrom  and  Counts  ( 1 955);  G.  Eumicrotremus  orbis.  from  Matarese  and  Borton  unpubl.  MS;  H.  Trachipterus  altivelus.  original; 
and  I.  Stomias  aim-enter,  original. 


grated  lo  their  actual  destinations  (Snyder,  1981).  As  seen  in 
the  cyclopterid,  Eumicrotremus.  most  late-stage  demersal  em- 
bryos resemble  the  newly  hatched  larva  with  respect  to  all  char- 
acters (Fig.  1 4G).  The  morphology  of  the  head,  gut,  and  postanal 


body  as  well  as  the  number  of  myomeres  is  used  for  identifi- 
cation within  all  tish  egg  groups.  A  number  of  specialized  char- 
acters associated  with  the  embryo  are  essential  for  identification 
when  present,  e.g.,  elongated  fin  rays— J'rachiplerus  (Fig.  14H), 


MATARESE  AND  SANDKNOP:  EGG  IDENTIFICATION 


31 


precocious  fin  development  (caudal— exocoetids  and  Tricho- 
don\  pelvic— Trachi mis),  and  pelvic  disc  development  in  some 
cyclopterids  (Eumicrotremus)  (Fig.  14G). 

Miscellaneous  characters.  —The  presence  of  a  secondary  mem- 
brane inside  the  chorion  occurs  in  some  groups,  although  it  is 
lacking  in  most  fishes.  Sloniias  alnvcnter  eggs  have  a  double 
membrane  (Fig.  141).  These  membranes  occur  in  some  of  the 
more  primitive  fishes  including  members  of  the  Anguilliformes, 
Clupeiformes,  and  Salmoniformes.  In  some  species,  like  the 
freshwater  cyprinid  Abbottina  rivularis  (Nakamura,  1969),  the 
secondary  membrane  is  thick  and  gelatinous.  The  presence  and 
size  of  the  micropyle  are  diagnostic  in  other  fishes,  particularly 
freshwater  demersal  eggs  (Laale,  1980;  Riehl,  1980).  Among 
freshwater  fishes,  the  cleavage  pattern  is  important  for  egg  iden- 
tification. In  the  more  primitive  families  (Acipenseridae,  Poly- 
odontidae,  Lepisosteidae,  and  Amiidae),  cleavage  pattern  is  typ- 
ically semiholoblastic  as  opposed  to  the  meroblastic  pattern  seen 
in  the  higher  teleosts.  Genetic  studies  have  shown  differences 
in  LDH  A  zymograms  to  be  a  useful,  diagnostic  tool  for  the 
identification  of  Gadus  morhua  and  Melanogrammus  aeglefinus 
eggs  (Mork  et  al.,  1983). 

Ecological  and  behavioral  considerations.- \  number  of  con- 
siderations related  to  mode  of  reproduction  and  collection  rather 
than  the  characters  of  the  eggs  themselves  are  essential  when 
identifying  any  type  offish  egg.  In  identifying  demersal  eggs  one 
must  consider  where  they  were  collected  — on  rocks,  on  plants, 
in  masses,  and  if  parental  care  is  involved.  Nest  type,  nature  of 
egg  deposition,  and  the  presence  of  guarding  parents  can  all  be 
essential  clues  to  proper  identification.  Also,  for  any  egg  type 


one  must  note  spawning  time  (season),  location  depth,  and  gear 
used  for  collection.  In  addition,  the  rearing  of  unknown  eggs  to 
an  identifiable  larval  stage  is  useful  in  species  determination  as 
shown  by  Stevens  and  Moser  (1982)  for  the  blenny,  Hypso- 
blennius.  Of  course,  a  necessary  prerequisite  to  accurate  iden- 
tification of  eggs  is  a  thorough  knowledge  of  the  species  present 
in  any  given  area  and  their  breeding  seasonality. 

Summary  of  Characters 

Characters  most  useful  in  identification  of  fish  eggs  are  the 
following:  ( I )  egg  shape— spherical,  ellipsoidal,  irregular,  or  oth- 
erwise; (2)  egg  size— fish  eggs  range  in  size  from  0.5  to  26.0  mm; 
(3)  oil  globules— presence  or  absence,  number,  size,  color,  po- 
sition, and  pigmentation;  (4)  yolk  — segmented  or  homogeneous, 
nature  of  segmentation,  color,  pigmentation,  and  circulation 
pattern;  (5)  chorion— smooth  or  ornamented,  type  of  ornamen- 
tation, thickness,  color,  and  coatings;  (6)  perivitelline  space- 
width;  (7)  embryonic  characters— morphological  features,  pig- 
ment patterns,  and  special  structures;  (8)  miscellaneous  char- 
acters—inner or  secondary  membrane  (presence  or  absence,  lo- 
cation), cleavage  pattern,  micropyle  (size),  and  biochemical 
analysis;  and  (9)  ecological  and  behavioral  considerations— col- 
lection (gear,  location,  season,  etc.),  and  mode  of  reproduction 
(nests,  parental  care,  etc.). 

(A. CM.)  National  Marine  Fisheries  Service,  Northwest 
AND  Alaska  Fisheries  Center,  2725  Montlake  Boule- 
vard East,  Seattle,  Washington  98112;  (E.M.S.) 
Southwest  Fisheries  Center,  P.O.  Box  271,  La  Jolla, 
California  92038. 


Identification  of  Larvae 

H.  POWLES  AND  D.  F.  Markle 


MINOR  errors  in  identification  of  larval  fishes  can  lead  to 
major  misinterpretations  of  ecological  and  taxonomic 
phenomena.  Fish  identification  and  taxonomy  are  largely  based 
on  adult  characteristics  and  since  these  develop  during  the  larval 
period,  new  characters  must  be  discovered  and  validated  in 
order  to  identify  larval  fishes.  Usually  larvae  possess  fewer  char- 
acters than  adults  and  are  more  fragile.  Identification  can,  there- 
fore, be  difficult  and,  frequently,  must  be  based  on  a  combi- 
nation of  character  states. 

Since  larval  anatomy  is  by  its  nature  dynamic  (a  given  spec- 
imen being  a  snapshot  of  the  process  linking  embryos  to  adults), 
developmental  series  are  essential  to  identification.  Three  dif- 
ferent approaches  are  used  to  identify  larvae,  the  first  two  of 
which  arc  based  on  developmental  series:  I)  to  raise  eggs  and 
larvae  from  fertilized  eggs  of  known  parents;  2)  to  work  back- 
wards from  the  adult  utilizing  characters  common  to  succes- 
sively earlier  ontogenetic  stages;  and  3)  to  extrapolate  from  pre- 


vious results  obtained  by  (1)  or  (2)  to  synthesize  generic  or 
familial  diagnoses  and  identify  by  process  of  elimination  or 
limited  corroboration  (Ahlstrom  in  Berry  and  Richards,  1973; 
Leiby,  1981). 

There  are  pitfalls  in  all  approaches.  Laboratory-reared  larvae 
are  frequently  more  heavily  pigmented  than  wild-caught  spec- 
imens and  may  show  greater  meristic  variation  (Lau  and  Shaf- 
land,  1982).  Laboratory  rearing  may  be  financially  and  logis- 
tically  difficult  or  impossible  for  fishes  of  interest.  Ontogenetic 
transformations  arc  based  on  associations  of  adult  diagnostic 
characters  with  characters  that  persist  in  progressively  earlier 
ontogenetic  stages.  This  method  requires  careful  attention  to 
methodology,  as  well  as  good  ontogenetic  series  which  are  not 
always  available.  Purely  descriptive  accounts  of  larval  series 
(laboratory-reared  or  reconstructed)  may  not  be  useful  for  iden- 
tification purposes  if  no  diagnostic  characters  that  will  distin- 
guish sympatric  congeners  and/or  similar-looking  forms  are  pre- 


32 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES -AHLSTROM  SYMPOSIUM 


sented.  Novel  sorts  of  characters  or  ways  of  manipulating  data 
are  sometimes  needed  to  identify  larvae  and  the  data  required 
may  not  be  retrievable  from  "standard"  descriptive  accounts. 
Synthesis  and  elimination  is  the  normal  procedure  used  by  tax- 
onomists  to  identify  adult  fishes.  It  has  been  called  the  "look- 
alike"  system  when  applied  to  larval  fishes  (Leiby,  1981).  It  is 
basically  a  simple  procedure  but  the  pitfalls  are  numerous  and 
subtle.  As  with  some  early  adult  fish  taxonomy,  premature  syn- 
thesis may  often  be  based  on  the  wrong  characters  (e.g.  con- 
vergent characters)  and  lead  to  spurious  identifications. 

General  references  on  larval  fish  identification  include  Berry 
and  Richards  (1973),  Ahlstrom  and  Moser  (1976)  and  Moser 
(1981).  Some  recent  works  which  provide  exposure  to  a  wide 
range  of  larval  forms  and  literature  are  Ahlstrom  and  Moser 
(1981)  and  Fahay  (1983)  for  marine  taxa,  and  Auer  (1982)  and 
Balon  (1975a,  1981a)  for  freshwater  taxa. 

The  purpose  of  the  following  is  to  describe  the  tools— pref- 
erably sharpened,  polished  and  comfortable  to  use— which  should 
be  at  hand  when  the  ichthyologist  sits  down  to  identify  larval 
fishes.  Our  emphasis  is  on  three  main  factors:  1 )  the  larval  fish  — 
its  anatomy,  ontogeny,  and  phyletic  relationships;  2)  the  study 
area— its  ecology  and  zoogeography  and  3)  the  investigator— 
his  experience,  knowledge  and  ingenuity. 

Systematics,  Ontogeny  and  Anatomy 

Perhaps  the  most  important  type  of  character  for  identifica- 
tion of  larvae  is  meristic,  as  counts  usually  do  not  increase  or 
decrease  once  established.  All  meristic  characters  can  be  im- 
portant, but  vertebra/myomere  counts  and  fin  element  counts 
are  of  particular  value.  Meristic  variables  are  useful  at  different 
taxonomic  levels,  e.g.,  principal  caudal  fin  ray  and  pelvic  fin 
element  counts  at  the  family  or  order  level,  median  fin  elements 
at  the  genus/species  level,  pectoral  fin  ray  counts  at  the  species 
level.  Frequency  distributions  of  meristic  counts  are  extremely 
important  (particularly  when  it  is  uncertain  whether  develop- 
ment of  a  character  is  complete)  but  often  are  not  given  in 
published  literature.  Some  important  characters  may  not  be 
included  in  published  studies  (e.g.,  pectoral  fin  rays,  procurrent 
caudal  rays).  Differences  in  methodology  and  variable  attention 
to  detail  may  also  affect  the  quality  of  published  meristic  data. 
Thus,  published  studies  must  be  treated  with  caution  and  one 
must  be  prepared  to  collect  and  compile  one's  own  information 
when  opportunities  arise.  Despite  potential  problems  with  pub- 
lished works,  these  are  the  obvious  place  to  start  with  compi- 
lations. Few  "regional"  meristic  publications  as  exemplified  by 
Miller  and  Jorgensen  (1973)  exist,  but  many  publications  on 
larval  fishes  include  extensive  tabulations  of  meristic  infor- 
mation. 

Various  ways  exist  for  facilitating  use  of  meristic  compila- 
tions. A  simple  taxonomic  listing  (e.g..  Miller  and  Jorgensen, 
1973)  can  be  time-consuming  to  use,  while  a  "gazetteer"  format, 
with  species  arrayed  in  order  of  counts  (e.g.,  Fahay,  1983)  may 
be  more  practical.  X-Y  plots  of  two  meristic  variables  (e.g.. 
Berry,  1959b)  can  include  frequency  distributions  and  be  very 
useful  for  separating  closely-related  forms. 

A  second  suite  of  characters  of  broad  use  is  specialized  larval 
characters  which  may  characterize  whole  groups.  These  include 
but  are  not  limited  to:  characteristic  shapes  (e.g.,  Anguilli- 
formes/Elopiformes,  Pleuronectiformes),  spination  (Acanthur- 
idae,  Holocentridae),  fin  development  patterns  (argentinoids), 
fin  element  development  (Pleuronectiformes,  epinepheline  Ser- 


ranidae),  fin  placement  (pelvic  fin  placement  in  Pleuronecti- 
formes), eye  shape  (myctophid  subfamilies,  salmoniform 
groups),  and  phoiophore  development  pattern  (Gonostomati- 
dae).  The  elucidation  of  such  characters  is  a  focus  of  this  volume, 
and  reference  should  be  made  to  specific  chapters  for  further 
detail.  The  important  point  is  that  a  broad  knowledge  of  larval 
fishes  is  frequently  necessary  for  accurate,  efficient  identification 
of  larvae. 

Finally,  identification  of  larvae  depends  on  a  suite  of  dynamic 
characters  (pigmentation,  body  form,  spination,  fin  develop- 
ment pattern,  etc.),  which  may  change  rapidly  and  differentially 
over  a  small  size  range.  Generally,  a  combination  of  such  char- 
acters is  required  for  accurate  identification;  this  is  particularly 
true  in  early  stages.  These  characters  can  vary  extensively,  even 
within  a  species,  due  to  regional  differences;  method,  time  or 
area  of  collection;  preservation  method  or  duration.  Develop- 
mental changes  can  be  extremely  rapid  (e.g.,  changes  in  mela- 
nophore  distribution  from  some  yolk-sac  to  post-yolk-sac  lar- 
vae). Again,  no  extensive  treatment  of  these  characters  is  possible 
here,  but  the  important  point  is  that  detailed,  disciplined  ob- 
servations of  larvae  are  essential  for  accurate  identification. 

The  importance  of  osteological  characters  for  larval  identi- 
fication is  increasingly  recognized  (Dunn,  this  volume).  Use  of 
these  depends  on  clearing  and  staining  techniques  (PotthofT,  this 
volume)  or  X-ray  techniques  (Tucker  and  Laroche,  this  vol- 
ume). As  with  meristics,  osteological  characters  may  be  useful 
at  different  taxonomic  levels.  Caudal  osteology  has  been  widely 
used  because  of  its  early  development  and  relative  simplicity, 
but  cranial  osteology  and  pterygiophore  patterns  are  also  useful. 
Recent  application  of  cartilage-staining  techniques  has  permit- 
ted use  of  cartilaginous  structures  in  identifying  larvae  (e.g., 
Fritzsche  and  Johnson,  1980).  Other  internal  characters  such  as 
gut  shape  (Ahlstrom  and  Moser,  1976;  Govoni,  1980)  may  also 
be  useful. 

Keys  have  not  generally  been  used  in  larval  fish  identification 
because  of  the  dynamic  nature  of  characters  (a  separate  key 
would  be  required  for  each  size  class  or  development  stage)  and 
because  of  "incompleteness"  of  information  (i.e.,  it  has  usually 
been  impossible  to  completely  cover  a  defined  region  or  sys- 
tematic group  with  a  key).  Generally,  much  more  information 
is  required  to  identify  a  larva  than  an  adult,  and  summarizing 
this  in  a  key  has  been  impractical  (the  information-organizing 
capacity  of  computers  may  eventually  help  to  permit  this).  Ex- 
ceptions, such  as  Bertelsen's  (1951)  key  to  larval  Ceratioidea, 
Johnson's  ( 1 974b)  key  to  genera  of  larval  scopelarchids,  and  the 
key  of  Bertelsen  et  al.  (1976)  to  notosudids  do  exist. 

Because  of  the  complexity  of  identification  of  larvae,  a  wide 
ichthyological  background  is  important.  A  good  knowledge  of 
fish  anatomy  is  essential,  particularly  when  (as  often  occurs) 
damaged  specimens  must  be  identified.  Published  descnptions 
exist,  for  example,  which  interpret  broken  branchiostegal  rays 
as  jugular  pelvic  fin  rays.  A  general  knowledge  of  suspected 
phylogenies  and  inter-relationships  (e.g..  Greenwood  et  al.,  1966; 
Nelson,  1976)  is  essential  if  attempting  to  identify  by  synthesis 
or  elimination.  This  should  at  least  cover  those  groups  to  be 
expected  in  a  given  area,  but  wider  knowledge  is  desirable,  par- 
ticularly in  the  marine  environment  where  exotic  larvae  may 
be  transported  great  distances  (e.g.,  Markle  et  al.,  1 980).  Finally, 
thorough  familiarity  with  the  ontogenetic  continuum  is  neces- 
sary to  place  unknown  specimens  in  perspective.  Absorption  of 
the  yolk  sac,  flexion  of  the  notochord  in  the  caudal  region, 
development  of  median  fins,  and  transformation  from  larval  to 


POWLES  AND  MARKLE:  LARVAL  IDENTIFICATION 


33 


juvenile  stages  (as  defined  by  completion  of  fin  element  devel- 
opment, development  of  scales,  etc.)  are  major  events  in  fish 
development  which  have  been  used  by  various  authors  to  define 
stages  (e.g.,  Ahlstrom,  1968;  Snyder,  1976). 

Ecological  Considerations 

There  are  two  basic  ecological  or  zoogeographic  consider- 
ations when  identifying  larvae:  the  expected  composition  of  the 
larval  ichthyofauna  of  the  study  area  and  the  potential  for  influx 
from  "upstream"  areas. 

Thorough  knowledge  of  the  adult  ichthyofauna  of  the  study 
area  is  essential  in  order  to  know  what  larvae  may  occur;  thus, 
the  most  complete  possible  list  of  adult  species  is  required. 
Literature  may  be  incomplete  or  erroneous,  so  this  list  should 
be  based  on  unpublished  or  personal  observations  as  well  as  on 
standard  faunal  works  or  other  literature.  For  ease  of  use,  the 
list  should  be  organized  by  systematic  groups  (e.g..  Greenwood 
et  al.,  1966;  Nelson,  1976). 

In  addition  to  knowledge  of  the  adult  ichthyofauna,  knowl- 
edge of  spawning  seasons  is  central  to  prediction  of  the  larval 
fish  composition.  As  with  meristic  or  anatomical  information, 
published  information  may  be  incomplete  so  that  personal  col- 
lections and  unpublished  information  may  be  important.  Al- 
though capture  location  and  season  can  be  important  in  elim- 
inating some  species  from  consideration,  caution  is  essential 
here  as  with  other  "elimination"  methods. 

Since  most  marine  fishes  have  planktonic  eggs  and/or  larvae 
and  many  have  a  prolonged  planktonic  life  the  basic  hydrog- 
raphy of  a  study  area  must  be  understood.  A  "downstream" 
study  area  is  potentially  vulnerable  to  an  influx  of  larvae  from 
"upstream"  spawning.  In  addition,  the  direction  of  "streams" 
can  differ  at  different  depths  of  the  water  column  so  the  influx 
may  come  from  more  than  one  direction.  On  the  shelf  oR"Nova 
Scotia  the  general  circulation  is  from  the  northeast  but  there  is 
a  strong  influence  from  the  Gulf  Stream,  both  from  eddies  and 
mixing  which  produces  Slope  Water.  Thus,  for  some  species, 
the  "downstream"  effect  comes  from  the  northeast  while  for 
tropical  and  oceanic  species  it  comes  from  the  southeast. 

Knowledge  of  an  area's  fish  communities  may  help  in  inferrmg 
which  larvae  may  occur  together— for  example,  an  unknown 
specimen  taken  together  with  larvae  from  a  coastal  community 


is  probably  not  a  mesopelagic  species.  Again,  however,  such 
inferences  should  be  considered  critically. 

One  sort  of  ecological  observation  may  be  misleading— al- 
though spawnmg  biomass  may  be  calculated  from  egg  and  larval 
abundance  for  some  species,  the  relative  apparent  abundance 
of  adults  is  not  always  in  proportion  to  the  relative  abundance 
of  planktonic  larvae.  Cryptic  species  may  appear  rare  in  collec- 
tions of  adults  but  larvae  may  be  extremely  abundant  (e.g., 
Gobiidae  in  tropical  and  subtropical  waters)  while  species  which 
appear  extremely  abundant  as  adults  may  be  rare  as  planktonic 
larvae  (e.g.,  the  clupeid  Jenkmsia  lamprotaenia  in  the  Carib- 
bean, Powles,  1977). 

Some  General  Considerations 

Like  larval  development,  identification  of  larvae  is  a  dynamic 
process— the  cumulative  knowledge  of  the  student  is  the  key  to 
accurate  identification.  The  complexity  of  larval  identification 
requires  that  a  wealth  of  information  be  applied  to  the  task,  and 
for  this  reason  some  degree  of  specialization  in  identification  of 
larvae  is  required  for  all  but  the  simplest  identification  prob- 
lems. There  are  many  examples  of  superficially  similar  but  sys- 
tematically very  different  larvae,  and  most  students,  including 
the  authors,  have  experienced  embarrassment  at  an  uncritical 
identification.  Identification  of  larvae  is  frequently  comparative, 
by  elimination,  so  that  wide  knowledge  of  larval  fishes  as  well 
as  caution  are  necessary. 

The  student  must  have  information  of  the  kinds  identified 
above.  Organization  and  ingenuity  are  required  in  order  to  keep 
this  information  usable  — card  files,  looseleaf  binders,  drawings 
and  sketches,  and  well-curated  reference  series  should  be  de- 
veloped or  readily  available. 

Finally,  although  many  beginning  students  are  hesitant  to 
draw,  sketching  and  drawing  (freehand,  on  squared  paper,  or 
with  camera  lucida)  is  one  of  the  best  ways  to  "see"  and  un- 
derstand larval  anatomy.  The  process  is  painstaking  and  often 
frustrating  in  the  early  stages,  but  will  pay  off  in  the  long  term 
with  increased  understanding. 

(H.P.)  Fisheries  and  Oceans,  P.O.  Box  15500,  Quebec  GIK 
7Y7,  Canada;  (D.F.M.)  Huntsman  Marine  Laboratory, 
Brandy  Cove,  St.  Andrews,  New  Brunswick,  EGG  2X0 
Canada. 


Illustrating  Fish  Eggs  and  Larvae 
B.  Y.  SuMiDA,  B.  B.  Washington  and  W.  A.  Laroche 


SCIENTIFIC  illustrations  of  fish  eggs  and  larvae  are  an  in- 
dispensible  component  of  any  descriptive  work,  providing 
a  visual  reference  of  form  and  structure  which  is  not  possible 
to  express  by  written  descnptions  and  measurements  alone. 
Illustrations  facilitate  identification  by  emphasizing  distinctive 
but  often  subtle  morphological  characters  and  allow  for  com- 
panson  of  features  at  difl^erent  developmental  stages  and  with 
morphologically  similar  taxa.  These  qualities  make  illustrations 


the  preferred  and  most  frequently  used  aid  for  taxonomic  iden- 
tification of  fish  eggs  and  larvae. 

The  broad  range  of  morphological  diversity  found  among 
larval  fishes  requires  flexibility  in  technique  and  style  to  produce 
eflTective  illustrations,  but  the  criteria  of  accuracy,  clarity,  and 
consistency  of  style  should  be  met.  The  basic  concept  behind 
illustrating  a  fish  larva  involves  accurately  representing  a  three- 
dimensional,  somewhat  transparent  organism  on  a  two-dimen- 


34 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


sional  sheet  while  emphasizing  characters  which  are  most  useful 
in  identifying  the  actual  larva  from  the  drawing.  Such  characters 
include  the  fins,  pigmentation  patterns,  and  details  of  the  head 
such  as  the  jaws,  spines  and  eyes.  Internal  structures  such  as 
myomeres,  the  gut,  cleithrum,  and  posterior  end  of  the  noto- 
chord  may  also  be  emphasized  but  without  masking  important 
external  characters.  Details  of  other  internal  structures  as  well 
as  shading  or  stippling  for  contrast  are  best  excluded  or  de- 
emphasized  to  maintain  clarity.  Pigmentation  is  important  in 
identification  of  most  larvae  and  should  be  depicted  clearly. 
External  melanophores  can  be  drawn  with  a  fine-tipped  pen  as 
realistically  as  possible.  Internal  pigmentation  can  be  effectively 
represented  by  using  light  stippling  with  a  smaller  sized  pen- 
point.  Care  must  be  taken  to  avoid  confusion  of  internal  struc- 
tures with  pigmentation. 

Specimens  selected  for  illustration  should  ideally  be  those  of 
the  best  condition  available  and  representative  of  the  particular 
developmental  stage  in  both  pigmentation  pattern  and  mor- 
phology. The  number  of  specimens  to  be  illustrated  is  deter- 
mined by  the  nature  and  objective  of  the  publication,  the  amount 
of  material  available  in  various  size  groups,  and  the  degree  of 
morphological  and  pigmentation  change  undergone  by  the  par- 
ticular species  during  ontogeny.  Specimens  from  described  series 
should  be  archived  in  a  museum  collection  for  proper  care  and 
future  reference  after  completion  of  the  illustrations,  and  catalog 
numbers  should  be  published. 

The  detailed  drawing  begins  with  an  accurate  body  outline 
showing  the  proper  body  proportions  and  position  of  fins  and 
critical  pigment  spots.  This  is  most  easily  achieved  by  drawing 
in  light  or  blue  pencil  from  a  camera  lucida-equipped  micro- 
scope. Other  methods  include  drawing  from  a  projection  of  a 
slide  transparency  of  the  specimen  or  tracing  a  photograph.  By 
convention  the  lateral  view  of  the  larva  is  drawn,  with  the  head 
to  the  left.  The  exception  to  this  is  made  with  right-eyed  pleu- 
ronectiforms.  In  some  instances  a  dorsal  or  ventral  view  is  also 
necessary  to  clarify  a  pigment  pattern  or  laterally  projecting 
morphological  structures.  If  sketching  through  a  camera  lucida, 
it  is  helpful  to  use  a  magnification  which  allows  the  entire  spec- 
imen to  be  in  the  field  of  vision  as  long  as  important  details 
remain  visible.  Any  resulting  distortions  at  the  periphery  of  the 
field  can  be  compensated  for  by  differentially  focusing  the  mi- 
croscope on  the  particular  region  involved  while  carefully  pen- 
cilling along  the  image,  then  reconstructing  a  smooth  line  where 
disjointed  lines  meet.  Problems  involving  specimens  that  are 
too  large  or  too  small  can  often  be  overcome  by  using  lens 
adapters  or  eyepieces  of  lower  or  higher  magnification.  Large 
specimens  may  require  being  drawn  in  sections  which  are  later 
pieced  together.  This  original  sketch  should  be  made  large  enough 
to  clearly  indicate  fine  details  such  as  the  full  complement  of 
fin  rays,  but  not  excessively  so  with  the  result  of  producing  lines 
which  bleed  in  the  final  reduction  for  publication.  Related  to 
this  is  the  use  of  appropriate  sizes  of  pen  points  which  produce 
lines  fine  enough  to  draw  minute  details  yet  not  be  lost  in  re- 
production. Therefore,  in  determining  the  original  size  of  each 
drawing,  thought  should  be  given  to  the  desired  reduction  ratio 
as  well  as  the  number  of  illustrations  comprising  each  plate.  An 
opaque  projector  is  most  useful  for  obtaining  a  specific  size  for 
the  final  drawing  from  the  initial  sketch,  but  photocopy  reduc- 
tions also  work  well.  With  this  final  pencilled  sketch,  the  illus- 
trator can  work  with  the  larva  under  a  microscope  as  a  reference 
to  complete  details  of  the  drawing  before  attempting  to  ink  it. 
A  light  table  can  be  helpful  when  tracing  or  inking  over  a  rough 


pencilled  sketch.  The  illustrator  should  always  have  a  set  of 
meristics  of  the  specimen  being  drawn  and  an  understanding  of 
the  important  characters  to  be  emphasized.  A  thorough  inspec- 
tion for  accuracy  is  essential  to  insure  agreement  between  il- 
lustrations and  descriptive  text,  especially  concerning  pigmen- 
tation and  meristic  elements  with  size  and  stage  of  development. 
Ideally  exact  counts  and  measurements  can  be  obtained  directly 
from  the  illustration,  allowing  easy  identification  of  the  larva. 

Illustrations  are  often  designed  for  comparison  of  features  at 
different  stages  of  development  or  for  comparison  of  similar 
features  which  occur  among  different  taxa.  Special  care  should 
be  taken  to  represent  similar  features  in  a  consistent  style  from 
illustration  to  illustration.  For  example,  a  partially  ossified  fin 
ray  element,  an  ossified  fin  ray,  and  a  fin  spine  may  each  be 
depicted  in  a  consistent  but  slightly  different  manner  so  that  the 
illustration  not  only  shows  the  number  and  position  of  fin  ele- 
ments but  also  the  type  of  element  and  its  relative  stage  of 
development. 

Literature  dealing  with  larval  fishes  contains  a  broad  array  of 
illustrative  styles,  techniques,  and  quality.  Many  of  these  are  of 
limited  use  since  they  fail  to  meet  the  criteria  discussed  above. 
Photographs  frequently  yield  unsatisfactory  results  due  to  dif- 
ficulties in  focusing  on  small,  transparent  organisms  so  that  all 
body  parts  appear  equally  sharp,  and  they  preclude  emphasizing 
inconspicuous  but  important  features  for  identification.  Color 
illustrations  in  a  variety  of  media,  although  potentially  valuable, 
particularly  for  xanthophores,  are  limited  due  to  prohibitive 
publication  costs,  poor  reproducibility,  and  the  absence  of  a 
long-lasting  color  preservative.  Half-tone  illustrations  (see  Ahl- 
strom,  1965)  are  effective  but  difficult  to  reproduce.  These  latter 
two  techniques  may  become  more  practical  with  advances  in 
photocopy  technology.  The  preferred  technique  in  widespread 
use  consists  of  pen  and  ink  drawings  done  in  black  India  ink. 
Various  styles  of  illustrations  of  diverse  groups  of  larvae  are 
represented  in  Moser  (1981)  and  in  this  volume  which  serves 
as  a  useful  overview.  Poul  Winther,  George  Mattson,  and  other 
artists  (Ahlstrom  and  Ball,  1954;  Ahlstrom  and  Counts,  1955; 
Bertelsen  and  Marshall,  1956;  Ege,  1953,  1957,  and  1958;  Grey, 
1955b;  Moser,  Ahlstrom  and  Sandknop,  1977;  Moser  and  Ahl- 
strom, 1970;  Tuning,  1961;  Richardson  and  Washington,  1980) 
have  been  instrumental  in  establishing  a  fine  style  of  pen  and 
ink  drawings  which  we  emulate  and  have  found  most  effective 
in  its  applicability  to  larval  fish  identification.  We  maintain  a 
degree  of  flexibility  in  technique  and  style  which  varies  with  the 
taxonomic  group  under  consideration  but  falls  within  the  gen- 
eral framework  discussed  above. 

Illustrating  a  fish  egg  poses  a  more  difficult  problem  than 
illustrating  a  fish  larva  and  will  be  limited  to  a  brief  discussion. 
Encapsulation  by  the  chorion  necessitates  representing  the  three- 
dimensional  quality  of  the  egg  in  the  drawing  while  showing 
important  morphological  and  pigmentation  characters  of  inter- 
nal structures  (Ahlstrom  and  Moser,  1980;  Matarese  and  Sand- 
knop, this  volume)  with  as  much  clarity  as  possible.  Difficulties 
arise  due  to  the  superimposing  of  these  characters  from  a  two- 
dimensional  perspective,  particularly  when  the  chorion  is  or- 
namented, when  an  oil  globule(s)  is  present,  and  when  the  de- 
veloping embryo  is  fully  coiled. 

In  spite  of  the  more  complex  structural  representation  re- 
quired, the  same  criteria  of  accuracy,  clarity  and  consistency  of 
style  apply  to  egg  illustrations.  The  relative  proportions  of  the 
egg  size  to  the  size  of  the  embryo,  oil  globule(s),  and  width  of 
perivitelline  space,  the  number  of  myomeres,  and  length  of  gut 


SUMIDA  ET  AL.:  ILLUSTRATING 


35 


need  to  be  accurately  drawn.  An  effective  balance  between  show- 
ing important  characters  for  identification  and  three-dimen- 
sional reahsm  of  the  egg  is  required  to  maintain  clarity.  Several 
illustrations  of  the  egg  at  different  stages  of  development  and 
from  different  perspectives  are  helpful  in  demonstrating  key 
characters  such  as  embryonic  pigmentation,  myomeres,  and  po- 
sition of  the  oil  globule(s)  in  the  yolksac.  Adherence  to  a  con- 
sistent illustrative  style  is  primarily  critical  for  a  developmental 
series  of  eggs.  As  with  fish  larvae,  pen  and  ink  drawings  provide 
the  most  practical  technique  for  illustrating  fish  eggs,  but  the 
specific  style  of  illustrating  and  details  shown  depend  upon  the 
character  of  the  egg  and  its  stage  of  development.  Many  kinds 


of  illustrative  styles  and  techniques  are  found  in  the  literature 
(see  Ahlstrom  and  Moser,  1980  and  references  cited  therein) 
and  examination  of  these  is  most  helpful  in  effectively  illus- 
trating a  particular  type  of  fish  egg. 

(B.Y.S.)  National  Marine  Fisheries  Service,  8604  La  Jolla 
Shores  Drive,  La  Jolla,  California  92038;  (B.W.)  Gulf 
Coast  Research  Laboratory,  East  Beach  Drive,  Ocean 
Springs,  Mississippi  39564;  (W.L.)  Department  of 
Fisheries,  Humboldt  State  University,  Arcata,  Cal- 
ifornia 95521. 


Clearing  and  Staining  Techniques 

T.  POTTHOFF 


THE  clearing  of  tissues  and  the  staining  of  cartilage  and  bone 
are  indispensable  in  the  study  of  larval  and  juvenile  fishes. 
At  the  National  Marine  Fisheries  Service  Miami  Laboratory 
modifications  of  the  clearing  and  differential  cartilage-bone 
staining  technique  proposed  by  Simons  and  Van  Horn  (1971) 
and  Dingerkus  and  Uhler  (1977)  are  used.  The  modifications 
are  in  part  based  upon  an  unpublished  manuscript  by  W.  R. 
Taylor  and  G.  C.  Van  Dyke  from  the  National  Museum  of 
Natural  History,  Washington,  D.C.  A  wide  size  range  of  fish 
from  3  mm  NL  to  larger  than  500  mm  SL  can  be  cleared  and 
stained.  The  technique  works  well  for  all  sizes,  but  adjustments 
in  the  various  solution  soaking  times  are  made  dependent  on 
fish  size  (Table  5). 

Method 

F/.Ya/ZoA!.  —Specimens  are  fixed  in  1 0-15%  marble  chip  buffered 
formalin.  Samples  previously  fixed  in  formalin  of  lower  than 
10-15%  concentration  and  specimens  presently  in  alcohol  or 
fixed  in  alcohol  should  be  refixed  in  10-15%  formalin  for 
best  results.  Eighty  to  90%  of  all  larvae  of  different  perciform 
families  fixed  in  alcohol  totally  disarticulated  during  clearing 
and  staining.  In  juvenile  and  adult  fish  >  100  mm  SL  the  flesh 
is  routinely  removed  from  the  left  side  before  or  after  fixation. 

Dehydration— This  is  an  important  step,  because  even  small 
amounts  of  water  interfere  with  the  staining  of  cartilage.  Place 
specimen  from  the  formalin  into  solution  of  50  parts  of  95% 
cthanol  and  50  parts  distilled  water.  Do  not  wash  or  soak  spec- 
imens with  water  during  transfer  from  formalin  to  alcohol. 
After  one  day  for  larvae  <  20  mm  SL  and  two  days  for  specimens 
20-80  mm  SL  and  three  to  five  days  for  specimens  >80  mm 
SL  transfer  from  50%  ethanol  into  absolute  ( 100%  or  200  prooO 
ethyl  alcohol.  If  absolute  ethanol  is  not  available,  190  proof  or 
95%  ethanol  can  be  substituted  for  the  absolute,  although  stain- 
ing of  cartilage  will  not  be  as  intense.  A  second  change  of  ab- 
solute alcohol  is  desirable  in  larger  than  20  mm  SL  specimens. 
Leave  larvae  <20  mm  SL  for  one  day  in  the  absolute  alcohol 


and  juveniles  20-80  mm  SL  for  2  days.  Adult  and  juvenile  fish 
80-200  mm  SL  should  be  kept  in  absolute  ethanol  for  3  days 
and  fish  >200  mm  SL  should  be  soaked  for  one  week.  An 
intermediate  absolute  alcohol  change  should  be  given  to  all 
specimens  with  longer  than  one  day  soaking  time. 

Cartilage  staining.  — This  is  accomplished  by  placing  specimens 
in  an  acidified  alcohol  solution  of  the  alcian  blue  stain.  For  best 
results  70  parts  of  absolute  alcohol  should  be  mixed  with  30 
parts  of  acetic  acid  99%  glacial.  To  every  100  ml  of  acidified 
alcohol  20  mg  of  alcian  blue  powder  should  be  added.  The  above 
solution  should  be  used  on  larvae  and  juveniles  from  3  mm  NL 
to  80  mm  SL.  For  larger  fish,  a  staining  solution  of  60  parts 
absolute  alcohol  and  40  parts  of  acid  with  30  mg  of  alcian  blue 
for  every  100  ml  of  acidified  alcohol  should  be  used.  Fish  larvae 
and  juveniles  <80  mm  SL  should  be  left  in  the  alcian  staining 
solution  no  longer  than  24  hours.  Larger  juveniles  and  adults 
should  be  stained  no  longer  than  36  hours.  Specimens  >500 
mm  SL  can  remain  48  hours  in  the  alcian  staining  solution. 
After  the  specified  time  in  the  alcian  solution  the  stain  is  per- 
manently fixed  in  the  cartilage  and  cannot  be  removed  with  any 
chemicals  used  in  the  clearing  and  staining  process.  Staining 
solution  can  be  used  twice  for  staining  larvae  but  should  be 
discarded  after  staining  a  juvenile  or  adult  fish. 

Neutralization.  — This  process  raises  the  pH  within  the  specimen 
thus  allowing  proper  subsequent  bleaching.  The  higher  pH  pre- 
vents further  calcium  loss  from  the  bones  for  better  alizarin  red 
stain.  To  neutralize  the  specimen  remove  it  directly  from  the 
alcian  staining  solution  and  place  it  in  a  saturated  sodium  borate 
solution  for  12  hours  for  specimens  <80  mm  SL  and  for  48 
hours  for  larger  specimens.  For  the  juveniles  and  adults  that 
soak  for  48  hours,  change  the  sodium  borate  solution  once. 

Bleaching  (an  optional  .s/cpA  —  Larvae  with  little  pigment  on 
their  body  (e.g.,  Scombridae)  should  not  be  bleached.  Larvae 
covered  with  pigment  (e.g.,  Istiophoridae)  and  all  juveniles  and 
adults  must  be  bleached.  Prepare  bleaching  solution  by  mixing 


36 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  5.     Method  of  Clearing  and  Staining  Cartilage  and  Bone  in  Larvae,  Juvenile  and  Adult  Fish. 


Length  in  mm,  NL  or  SL 

Steps 

10 

20            .10 

40 

50 

60            70            80            90 

100 

200          .100          400          500 

>500 

Fixation: 

10-15%  formalin 
marble  chip  buffered. 

--►h  - 

—  5  days,  flesh  removed— 
on  left  side 

...-► 

Dehydration: 

1.  50%  distilled  H,0, 
50%of95%ethanol. 

2.  Absolute  ethanol 
(95%  ethanol  may 
be  substituted). 


-1  day ►[- 

-1  day ►(- 

h 


2  days • 
2  days 


>h- 


3  days ■ 


►h 3  days 

-one  intermediate  change 


-►h  • 


-5  days- 
-7  days- 


-► 
■-► 


Staining  cartilage: 
100  ml  solution: 

A.  70  ml  absolute  ethanol, 
30  ml  acetic  acid, 

20  mg  alcian  blue. 
100  ml  solution: 

B.  60  ml  absolute  ethanol, 
40  ml  acetic  acid, 

30  mg  alcian  blue. 


—  I  day 

-Solution  A- 


-►h I'/idays ►h"2  days->- 

-►H  Solution  B ► 


Neutralization: 
saturated  sodium 
borate  solution. 


'/2  day  - 


-►I-- 
I- 


2  days 

-one  intermediate  change - 


-► 
-► 


Bleaching: 
pigmented 
specimens  only. 
100  ml  solution: 
15  ml  3%  H,0„ 
85  ml  1%  KOH. 


-20  min. 


-►!-■ 


-40  min. 


-►  h  -  1  hour ►  I 1 1/2  hours  - 


Trypsin  digestion: 
100  ml  solution: 
35  ml  saturated  sodium 
borate.  65  ml  distilled 
H,0.  trypsin  powder. 


-Keep  in  solution  until  60%  clear,  change  to  fresh  solution  every  10  days- 


Staining  bone: 

1%  KOH  solution  with 
alizarin  red  stain. 


I  day  - 


-►[-■ 


2  days • 


-►h 


-4  days 


Destaining: 

100  ml  solution: 
35  ml  saturated  sodium 
borate,  65  ml  distilled 
HjO,  trypsin  powder. 


-2  days — ►!-  ■ 


Change  to  fresh  solution  every  10  days  until  solution  remains - 
unstained  and  specimen  is  clear 


Preservation: 
30%  glycerin  and 
70%  of  1%  KOH. 
60%  of  glycerin 
and  40%  of  I  %  KOH. 
1 00%  glycerin  with  thymol 
as  final  preservative*. 


1  week — ►!- - -2  weeks- — ►!- 4  weeks- 


*  Direct  sunlight  and  100%  glyceiine  help  to  clear  and  destain  difficult  specimens. 


15  parts  of  3%  hydrogen  peroxide  solution  with  85  parts  of 
1%  potassium  hydroxide  solution.  Bleach  larvae  and  small  ju- 
veniles up  to  80  mm  SL  for  20  to  40  minutes  depending  on 
size.  Larger  juvenile  fish  and  adults  may  be  bleached  1  to  1  Vi 
hours. 

Trypsin  digestion  and  alizarin  red  staining.  — The  clearing  and 
alizarin  staining  process  has  been  well  described  by  Taylor  ( 1 967) 
and  need  not  be  repeated  here.  Simply  continue  after  bleaching 


with  the  Trypsin  digestion,  which  are  Taylor's  steps  4  and  5. 
We  saw  no  need  in  modifying  Taylor's  method. 

Removal  of  semitransparent  tissue.  ~^\\ex\  studying  cleared  and 
stained  material  of  large  fish,  the  structures  studied  (caudal  com- 
plex, pectoral  fin  supports,  pterygiophores,  vertebral  column, 
etc.)  may  have  to  be  dissected  out  and  adhering  tissue  removed. 
This  can  be  accomplished  by  time  consuming  picking  with 
tweezers  or  by  placing  the  material  in  a  two-phase  phenol  so- 


POTTHOFF:  CLEARING  AND  STAINING 


37 


lution  with  the  addition  of  heat  (Miller  and  Van  Landingham, 
1969).  With  this  method  the  bones  are  not  disarticulated,  but 
some  bone  distortion  was  experienced. 

Variables  affecting  results.— The  results  of  the  clearing  and 
staining  procedure  are  not  always  satisfactory  because  of  known 
and  unknown  variables.  Results  can  never  be  predicted  with 
certainty.  The  known  variables  are:  ( 1 )  Time  and  ambient  tem- 
perature the  organism  is  subjected  to  between  death  and  fixation. 
The  longer  an  organism  remains  unpreserved  after  death  and 
the  higher  the  temperature,  the  less  the  tissues  will  clear.  For 
best  results,  specimens  should  be  killed  in  the  fixative,  or  if  that 
is  not  possible,  they  should  be  kept  cool  or  frozen  before  fixation. 
(2)  Effect  of  fixative  and  preservative.  Marble  chip  buffered 
formalin  is  a  good  fixative  for  larval  fish  if  specimens  are  re- 
moved from  it  after  24  hours.  Buffered  formalin  as  a  preser- 
vative destroys  first  the  stain  uptake  in  cartilage.  Bone  decalcifies 
as  buffered  formalin  becomes  acid  over  a  longer  time  period 
and  decalcified  bone  will  not  stain.  Therefore,  it  is  best  to  fix 
specimens  in  10%  formalin  and  then  to  preserve  them  in  70- 
95%  ethanol.  Specimens  fixed  and  preserved  in  ethanol  should 
be  re-fixed  in  formalin  before  clearing  and  staining.  (3)  Time  in 
a  preservative.  The  longer  a  specimen  has  been  preserved,  the 
less  predictable  the  clearing  and  staining  outcome  will  be.  Some 
fish  larvae  from  the  Dana  collection  in  the  1920's  were  cleared 
and  stained.  The  results  were  startling  for  both  Formalin  and 
alcohol  preserved  material  because  some  specimens  cleared  and 
stained  well,  but  most  were  unfit  for  study. 

Other  vanables  which  affect  the  results  of  clearing  and  staining 
exist,  but  are  not  understood.  No  matter  how  carefully  one 
adheres  to  the  procedures,  the  clearing  and  staining  results  are 
not  predictable. 

Interpretation  of  results.  —  Frequently  specimens  will  remain 
opaque  and  overstain  with  alcian  or  alizarin  for  unknown  rea- 
sons. This  makes  viewing  of  cartilage  and  bone  structure  diflicult 
or  impossible.  Such  specimens  can  be  used  for  study  of  fin  ray 
development  and  for  fin  ray  counts. 

Cartilage  or  bone  does  not  always  stain  but  can  be  made 


visible  in  cleared  preparations  by  changing  light  conditions  at 
the  microscope  and  manipulating  the  substage  mirror.  Cartilage 
appears  reticulated  in  structure  whereas  bone  is  structurally  clear 
and  hyaline. 

Erroneous  conclusions  can  be  made  if  one  solely  relies  on 
color  to  determine  cartilage  and  bone.  In  general,  cartilage  will 
appear  blue  and  bone  red,  but  often  alcian  blue  is  taken  up  by 
bones  and  rarely  alizarin  red  by  cartilage.  For  instance,  devel- 
oping fin  rays  often  appear  blue. 

Generally  larger  developed  cartilage  structures  will  stain  bet- 
ter than  small  developing  ones.  Thus,  in  the  same  specimen  one 
may  find  brightly  blue  stained  cartilage,  pale  blue  cartilage,  and 
cartilage  with  no  stain  at  all.  Therefore,  special  care  is  indicated 
when  viewing  newly  developed  cartilage. 

The  ossification  onset  in  cartilage  is  difficult  to  determine.  A 
thin  layer  of  bone  forming  all  around  the  cartilage  can  be  de- 
tected by  examining  the  outer  edges  of  the  cartilage  structure: 
a  shiny  hyaline  line  forms  there,  probably  only  a  cell  layer  thick. 

Investigators  are  often  discouraged  by  clearing  and  staining 
results,  particularly  when  their  sample  is  small.  In  a  larval  de- 
velopmental series  I  usually  clear  and  stain  200  to  400  speci- 
mens, and  I  am  able  to  study  each  aspect  and  area  of  devel- 
opment that  I  wish  to  examine  because  of  the  large  sample  size 
at  hand.  For  example,  in  a  specimen  in  which  the  pectoral  fin 
support  area  is  unclear  and  stained  poorly  the  caudal  area  may 
be  clear  and  stained  well.  Thus,  this  specimen  is  utilized  only 
for  caudal  development,  whereas  in  another  specimen  the  pec- 
toral area  may  be  clearer  and  better  stained.  Thus,  with  a  large 
sample  size,  the  uncertainties  and  vagaries  of  the  clearing  and 
staining  procedure  are  overcome. 

Application  of  clearing  and  staining.— Cleanng  and  staining  is 
helpful  in  identification  offish  larvae  when  external  characters 
are  inadequate.  It  also  aids  systematic  and  phylogenetic  studies 
of  larvae  to  adult  fishes.  This  subject  has  been  discussed  in  detail 
by  Dunn  (1983b). 

National  Marine  Fisheries  Service,  Southeast  Fisheries 
Center,  Miami  Laboratory,  75  Virginia  Beach  Drive, 
Miami,  Florida  33149. 


Radiographic  Techniques  in  Studies  of  Young  Fishes 
J.  W.  Tucker,  Jr.  and  J.  L.  Laroche 


RADIOGRAPHY  is  useful  for  obtaining  skeletal  informa- 
tion in  studies  of  fish  taxonomy  and  morphology.  Al- 
though clearing  and  staining  provides  more  detail,  radiography 
has  other  advantages.  It  produces  an  easily  stored,  long-term 
record  of  the  skeleton  and  does  not  permanently  alter  the  con- 
dition of  the  specimen.  In  many  cases,  counts  can  be  obtained 
more  accurately  from  radiographs  than  from  the  specimens 


themselves.  If  an  x-ray  unit  and  darkroom  are  available,  ra- 
diography is  usually  faster  and  easier  than  clearing  and  staining. 
The  time  saved  may  be  of  value  in  studies  of  population  vari- 
ation, in  which  many  specimens  must  be  examined.  Radiog- 
raphy has  also  been  used  to  monitor  decalcification  of  larvae 
stored  in  formalin  (Tucker  and  Chester,  in  press),  and  has  been 
suggested  for  use  in  toxicological  studies  to  check  large  numbers 


38 


ONTOGENY  AND  SYSTEM ATICS  OF  FISHES -AHLSTROM  SYMPOSIUM 


of  larvae  for  skeletal  deformities.  The  consensus  among  ichthy- 
ologists who  have  used  both  techniques  is  that,  although  clearing 
and  staining  methods  provide  the  detail  necessary  for  describing 
developmental  osteology,  radiography  is  a  simple  and  quick  way 
of  obtaining  counts  from  large  numbers  of  specimens. 

Hard  (shortwave)  x-rays  have  been  used  to  form  shadow  pic- 
tures, or  radiographs,  of  large,  well-ossified  fish  for  almost  four 
decades  (Goshne,  1948;  Bartlett  and  Haedrich,  1966),  but  the 
use  of  soft  (longwave)  x-rays  for  small  specimens  is  relatively 
new.  Although  first  suggested  by  Bonham  and  Baylifr(  1953)  and 
used  by  Watson  and  Mather  (1961  unpubl.  manusc),  useful 
techniques  for  larval  radiography  have  only  recently  been  de- 
scribed (Miller  and  Tucker,  1979).  Potential  larval  fish  radiog- 
raphers should  consult  Miller  and  Tucker's  paper  for  method- 
ological details  and  Quinn  and  Sigl  ( 1 980)  for  basic  radiographic 
principles.  Although  specimen  fragility  determines  the  mini- 
mum size  of  larvae  that  can  be  x-rayed,  sensitivity  of  the  tech- 
nique, which  depends  to  a  large  degree  on  spectral  characteristics 
of  the  radiation,  determines  the  amount  of  detail  present  in  the 
finished  radiograph.  This  section,  therefore,  reviews  the  prin- 
ciples and  current  methods  useful  for  maximizing  detail  in  ra- 
diographs of  fish  larvae. 

Radiographic  sensitivity  refers  to  the  clarity  of  details  in  the 
radiographic  image  and  depends  on  a  combination  of  two  fac- 
tors, definition  and  radiographic  contrast.  Definition  is  sharp- 
ness of  the  image.  Radiographic  contrast  refers  to  the  density 
(darkness)  range  of  the  image  and  depends  on  two  factors,  sub- 
ject contrast  and  film  contrast.  Subject  contrast  refers  to  the 
ratio  of  radiation  intensities  that  pass  through  different  parts  of 
the  specimen.  Film  contrast  refers  to  the  ratio  of  densities  in 
parts  of  the  film  that  have  received  different  degrees  of  exposure. 

In  larval  fish  work,  radiographic  sensitivity  can  be  improved 
by  several  means.  Definition  can  be  improved  by  using  the 
longest  possible  radiation  wavelengths,  by  using  the  finest  grained 
film  available,  and  by  minimizing  geometric  production  of  over- 
lapping shadows  at  tissue  discontinuities  in  the  specimen.  Ab- 
sorption by  x-rays  of  a  given  wavelength  depends  mostly  on  the 
atomic  numbers  of  components  in  the  x-rayed  material,  and  to 
a  lesser  degree  on  thickness  and  density  of  the  material.  Larval 
skeletons,  which  are  thin,  poorly  calcified,  and  of  relatively  uni- 
form composition  and  thickness,  do  not  contrast  radiographi- 
cally  with  the  rest  of  the  body  as  much  as  in  older  fish.  High 
contrast  techniques  should,  therefore,  be  employed.  Subject  con- 
trast can  be  increased  by  increasing  wavelengths  and  by  de- 
creasing the  thickness  of  non-skeletal  tissue  by  dehydrating  the 
specimen.  Film  contrast  can  be  increased  by  using  a  high  con- 
trast film  and  by  increasing  development  time;  however,  over- 
development will  also  increase  graininess  and  reduce  definition, 
and  probably  should  be  avoided. 

The  longwave  (soft)  end  of  the  x-ray  spectrum  is  the  portion 
most  useful  for  x-raying  small  fish,  because  this  low  energy 
radiation  does  not  pass  through  materials  as  easily  as  that  at 
the  shortwave  (hard)  end.  Decreasing  the  tube  voltage  (kv)  caus- 
es a  shift  of  the  emitted  spectrum  toward  longer  wavelengths. 
Resultant  elimination  of  some  of  the  hard  radiation  contributes 
to  better  subject  contrast  and  improves  definition  by  reducing 
clumping  of  silver  grains  in  the  film  emulsion  (graininess).  The 
x-ray  unit  should  be  equipped  with  a  thin  beryllium  window, 
which  allows  passage  of  soft  rays.  A  25  mil  (0.63  mm)  window 
allows  work  at  a  kv  of  20;  a  10  mil  (0.25  mm)  window  extends 
capabilities  to  about  8  kv  (Joseph  Fowler,  Hewlett  Packard,  pers. 
comm.).  However,  the  lower  practical  limit  for  fish  larvae  may 


be  governed  by  restrictions  on  exposure  time,  rather  than  kv 
limitations. 

Another  relevant  factor  is  the  source-to-specimen  distance, 
to  which  image  definition  is  directly  related.  Increasing  the  source- 
to-specimen  distance  improves  definition  by  minimizing  en- 
largement and  distortion.  Practical  limits  are  set  by  air  atten- 
uation, loss  of  radiation  intensity  (roughly  as  the  square  of  the 
ratio  of  the  distances),  and  dimensions  of  the  x-ray  unit.  Geo- 
metric unsharpness  is  the  maximum  width  of  the  zone  of  over- 
lapping shadows  that  are  caused  by  a  non-point  source.  This 
factor  can  be  calculated  to  determine  the  minimum  source  to 
specimen  distance  that  can  be  tolerated.  Use  of  the  minimum 
distance  will  permit  the  shortest  possible  exposure  time  and 
reduce  relative  attenuation  of  soft  rays,  thus  contributing  to 
subject  contrast.  The  formula  for  geometric  unsharpness,  Ug 
(Quinn  and  Sigl,  1980)  is: 


U„ 


D, 


in  which  F  is  the  radiation  source  size.  Do  is  the  source-to- 
specimen  distance,  and  t  is  the  specimen  to  film  distance  (max- 
imum specimen  thickness).  For  F  =  0.5  mm,  D,,  =  460  mm, 
and  t  =  1  mm,  U^  is  0.00 1  mm.  This  level  of  unsharpness  would 
not  be  visible  without  magnification  and  could  be  tolerated  at 
moderate  magnification  depending  on  the  requirements  of  the 
investigator.  To  ensure  that  geometric  unsharpness  is  not  large 
enough  to  affect  quality  of  radiographs,  it  should  be  calculated 
for  the  set  of  factors  relevant  to  each  operation,  keeping  in  mind 
the  level  of  magnification  to  be  used.  With  most  modem  x-ray 
units,  a  distance  of  46  cm  or  less  can  be  used. 

Because  air  attenuates  soft  rays  more  than  hard,  elimination 
of  air  between  the  x-ray  source  and  specimen  allows  a  greater 
proportion  of  soft  radiation  to  reach  the  specimen.  Decreasing 
the  source  to  specimen  distance  helps  some,  but  also  increases 
geometric  unsharpness,  unless  the  source  is  very  small.  A  vac- 
uum would  be  ideal  but  is  impractical.  Replacement  of  the  air 
in  a  cabinet  unit  with  helium  allows  the  use  of  lower  kv  with 
reasonably  short  exposure  times  and  provides  an  increase  in 
subject  contrast.  Helium  can  be  conserved  and  reused  if  it  is 
placed  in  a  small  volume  plastic  cylinder  that  has  its  ends  sealed 
with  dry-cleaning  plastic. 

Before  a  specimen  is  x-rayed  it  should  be  dehydrated  as  much 
as  can  be  tolerated  to  increase  the  signal  (skeleton)  to  noise 
(non-skeleton)  ratio.  For  best  results,  the  specimen  should  be 
placed  in  50-75%  ethyl  alcohol  for  a  short  period,  maybe  30- 
60  min,  depending  on  size.  Then  the  specimen  should  be  placed 
on  the  film  holder,  blotted  to  remove  surface  liquid  and  bubbles, 
and  quickly  x-rayed  and  returned  to  a  container  of  liquid  before 
desiccation  damage  occurs. 

The  specimen  should  be  placed  as  close  as  possible  to  the  film 
emulsion.  This  can  be  accomplished  without  wetting  the  film 
by  sandwiching  it  between  two  thin  sheets  of  black  polyethylene. 
Details  for  construction  of  a  convenient  film  holder  (cassette) 
are  presented  in  Miller  and  Tucker  (1 979).  Polyethylene  is  trans- 
parent to  soft  x-rays  and  is  good  cassette  material.  Vinyl,  as  well 
as  wood,  paper,  and  any  metal  are  relatively  opaque  to  soft 
x-rays,  and  vinyl  or  metal  make  good  labels. 

Single  coated  Type  R  (now  Type  XAR)  film  has  provided  the 
best  quality  radiographs  of  larvae.  High  resolution  plates  give 
better  resolution  but  are  too  slow.  Type  R  film  is  slow  relative 
to  other  films  but  within  practical  limits.  It  has  ultra-fine  grain 


TUCKER  AND  LAROCHE:  RADIOGRAPHY 


39 


Fig.  15.  Positive  image  of  radiograph  of  a  southern  flounder  (Paralichthys  tethosligma)  larva,  9.7  mm  SL,  stored  m  7%  borax  buffered  seawater 
formalin  for  seven  years.  Radiographic  exposure  data:  Faxitron  Model  43805N;  Kodak  Type  R  film;  source  to  film  distance.  46  cm;  9  kv;  600 
mAs;  under  helium.  Intemegative  processing  data:  radiograph  was  projected  onto  4  in  x  5  in  professional  copy  film  (Kodak  4125)  with  an  Omega 
(4  in  X  5  in)  Pro  Lab  Enlarger;  exposure  was  1  s  at  f  S'/j;  film  was  developed  in  Kodak  HCl  10  (dilution  E)  for  5  min  at  23  C.  Print  processing 
data:  a  positive  pnnt  was  made  on  Kodak  Polycontrast  Rapid  11  RCF  paper  using  a  polycontrast  no.  3  filter  in  the  Omega  enlarger;  exposure  was 
5  s  at  f  5.6;  print  was  developed  in  Kodak  Ektaflo  diluted  to  simulate  Dektol  1:1,  at  23  C.  (The  intemegative  and  printing  procedure  was  devised 
and  performed  by  Tom  Smoyer  of  Harbor  Branch  Foundation.) 


and  high  contrast.  The  single  emulsion  is  necessary  for  avoiding 
two  images  (on  both  sides  of  the  film).  Coarser  grained  and 
lower  contrast  films  will  produce  inferior  radiographs. 

Exposures  should  not  be  longer  than  about  5  min,  and  for 
many  specimens  5  min  is  too  long.  Larvae  will  quickly  desiccate, 
and  even  if  not  damaged,  may  shrink  and  cause  blurred  images. 
Specimen  damage  or  image  blurring  will  determine  the  mini- 
mum size  of  larvae  that  can  be  x-rayed.  Specimens  can  be  pro- 
tected by  an  overlying  sheet  of  dry-cleaning  plastic  if  care  is 
taken  to  remove  bubbles.  During  exposure,  unneeded  portions 
of  the  film  can  be  protected  for  later  use  with  lead  vinyl  masks. 

The  manufacturers'  instructions  for  mixing  chemicals  and 
processing  films  should  be  followed  as  closely  as  possible.  Fre- 
quent agitation  of  the  film  while  it  is  developing,  rinsing,  and 
fixing  is  important  to  ensure  uniformity  of  chemical  reactions. 
Both  undeveloped  and  developed  films  should  be  stored  away 
from  light,  heat,  humidity,  and  chemical  fumes  (particularly 
formalin,  alcohol,  and  hydrogen  peroxide).  Radiographs  are  best 
observed  directly,  emulsion  side  up,  with  a  dissecting  or  phase 
contrast  microscope.  Printing  of  radiographs  is  best  done  via 
an  intemegative  (Fig.  15).  This  compresses  the  tonal  range  so 
that  finer  detail  can  be  preserved  in  the  print. 


The  major  limitation  of  the  technique  is  probably  inadequate 
radiation  intensity  at  low  kv.  This  limit  may  have  been  reached 
with  x-ray  units  equipped  with  10  mil  beryllium  windows.  Sat- 
isfactory radiographs  of  4-1  5  mm  larvae  have  been  made  at  8- 
10  kv  and  300-800  mAs  (milliamperes  x  seconds).  Some  im- 
provement can  be  expected  if  the  air  is  replaced  with  helium; 
however,  exposure  time  will  eventually  become  prohibitively 
long. 

Because  machine  and  specimen  characteristics  vary,  a  stan- 
dard formula  for  producing  high-quality  radiographs  cannot  be 
provided.  At  least  initially,  the  larval  fish  radiographer  must 
proceed  by  trial  and  error  with  the  machine  and  specimens  at 
hand.  As  familiarity  develops,  the  results  will  improve  signifi- 
cantly. We  stress  that  an  accurate  and  detailed  logbook  con- 
taining specimen  and  exposure  data  should  be  kept,  and  that 
procedures  should  be  standardized. 

(J.W.T.)  Harbor  Branch  In.stitiition,  Inc.,  RR  l,Box  196-A, 
Fort  Pierce,  Florida  33450;  (J.L.L.)  Gulf  Coast  Re- 
search Laboratory,  East  Beach  Drive,  Ocean  Springs, 
Mississippi  39564. 


Histology 

J.  J.  GOVONI 


WHILE  contemporary  systematists  rely  upon  a  broad  scope 
of  biological  features  to  infer  relationships  among  taxa, 
the  definition  and  comparison  of  morphological  characters  re- 
mains one  of  their  most  useful  tools.  The  small  size  and  often 
altricial  development  of  fish  larvae,  however,  make  it  difficult 
to  resolve  the  morphology  of  structures  other  than  skeletal  ele- 
ments. By  clarifying  tissue  composition  and  by  enhancing  mor- 
phological resolution,  histological  techniques  may  aid  the  sys- 
tematist  in  defining  characters  at  the  tissue  as  well  as  at  the 
microanatomical  level,  thereby  providing  additional  character 
states  to  be  examined  for  synapomorphies  and  perhaps  onto- 
genetic precedence.  Because  of  their  small  size,  sections  of  whole 
larvae  can  be  prepared  (Fig.  16)  and  structural  relationships  of 
organ  systems  examined.  Insofar  as  there  is  no  clear  separation 
between  gross  and  micro-anatomy  beyond  the  limits  of  human 
visual  resolution,  histological  techniques  may  otfer  yet  another 
tool  useful  in  phylogenetic  analysis. 

Techniques 

Flvi2;/o«.  — Inasmuch  as  autolysis  is  rapid  in  larval  tissue  (Thei- 
lacker,  1978),  fixation  is  difficult  (Richards  and  Dove,  1971). 


Specimens  reared  in  the  laboratory  or  specimens  taken  from 
brief  plankton  tows  (O'Connell.  1980)  are  the  most  suitable  for 
histological  preparation  and  study;  specimens  sorted  from  field 
collections  fixed  in  formalin  and  seawater  will  usually  yield  poor 
quality  preparations.  Neutral  buffered  (phosphate  buffi;rs)  for- 
malin (see  Humason,  1979)  enhanced  with  <4%  acrolein  (van 
der  Veer,  1 982)  is  recommended  for  rapid  and  thorough  fixation. 
Glutaraldehyde  (2.5%)  is  also  a  useful  fixative  (Hulet,  1978). 

Difference  in  the  osmolality  of  tissues  and  ambient  water  may 
distort  cells  and  tissues,  especially  of  marine  larvae.  Such  arti- 
facts have  not  been  observed  in  preparations  of  clupeiform  and 
perciform  larvae,  but  may  be  of  concern  in  the  preparation  of 
anguilliform  leptocephali  (Hulet,  1978).  Forsterand  Hong  (1958) 
and  Hulet  (1978)  provided  applicable  saline  solutions  that  may 
eliminate  distortion  and  enhance  staining. 

Sectioning  and  staining.  — Sxandsivd  animal  tissue  techniques 
(e.g.,  Humason,  1979)— dehydration,  paraffin  embedding,  and 
sectioning— have  been  used  to  trace  the  development  of  organ 
systems  (O'Connell,  1981a),  as  well  as  to  assess  the  pathology 
of  starvation  in  fish  larvae  (Umeda  and  Ochiai,  1975;  O'Con- 


Fig.  16.     Sagiual  section  ot  a  Leiostomus  xanlhurus  larva,  4.4  mm  notochord  length  (glycol  methacrylate  section  stained  with  alkali  blue  6B- 
neutral  red). 


Fig.  17.  Example  comparisons  of  larval  fish  tissue  and  microanatomy.  Abbreviations:  AM,  axial  musculature;  CS,  collagenous  supporting 
shafts;  EP,  epidermal  cells;  M,  midgut;  MC,  mucous  cell;  NF,  nerve  fiber.  (A)  The  integumentary  epithelium  of  a  Brevoortia  patronus  larva 
showing  hyaline  plates  (arrow),  a  tissue  characteristic  of  some  clupeiform  larvae.  Note  that  erosion  of  the  outer  layer  of  epithelium  is  evident. 
(Scale  bar  =  20  /jm;  glycol  methacrylate  section  stained  with  acid  fuchsin  — toluidine  blue.)  (B)  The  integumentary  epithelium  of  a  Leiostomus 
xanthurus  larva  showing  lack  of  hyaline  plates  in  epithelial  cells.  (Scale  bar  =  10  iim;  glycol  methacrylate  section  stained  with  alkali  blue  6B  — 
neutral  red.)  (C)  Axial  musculature  of  a  Brevoortia  patronus  larva  showing  two  opposing  layers  of  muscle  fibers,  a  tissue  characteristic  of  clupeiform 
larvae.  (Scale  bar  =  50  livn,  glycol  methacrylate  section  stained  with  acid  fuchsin  — loluidine  blue.)  (D)  Axial  musculature  of  a  Leiostomus  xanthurus 
larva  showing  muscle  fiber  layers  in  parallel  alignment,  a  tissue  characteristic  of  perciform  larvae.  (Scale  bar  =  50  iim\  glycol  methacrylate  section 
stained  with  alkali  blue  6B  — neutral  red.)  (E)  Cross  section  of  the  elongate  dorsal  ray  of  an  Echiodon  dawsoni  larva.  (Scale  bar  =  20  ixm:  glycol 
methacrylate  section  from  Govoni  et  al.,  1984.)  (F)  Cross  section  of  the  elongate  dorsal  ray  of  a  Bregmaceros  atianticus  larva.  (Scale  bar  =  15 
Min;  glycol  methacrylate  section  stained  with  acid  fuchsin  — toluidine  blue.) 


40 


GOVONI:  HISTOLOGY 


41 


.:•»'  /■ 


B 


.f \ 


MC 


tiSNHBC^- 


EP 


AM 


'^  .-    i^. 


t 


w 


•tr\ 


►'i'^   . 


,^ 


"     f  >i  ••' 


D 


V- 


*,'" 

AM 

^ 

"« 

'-« 

.   \ 

~<^ 

.» 


■^, 


•-  *k- 


\. 


cs 


42 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


nell,  1976;  Theilacker,  1978).  These  techniques  will  suffice  for 
the  examination  of  soft  tissue  morphology  given  adequately 
fixed  specimens.  To  avoid  their  loss,  small  specimens  may  be 
prestained  with  borax-carmine  before  embedding  and  section- 
ing; this  stain  can  be  washed  out  before  subsequent  histological 
staining  (Engen,  1968). 

Plastic  embedding  (Bennett  et  al.,  1976)  is  advantageous  for 
examination  of  small  delicate  structures,  for  precise  records  of 
specimen  orientation  and  section  plane,  and  for  the  resolution 
of  fine  cellular  detail.  Glycol  methacrylate  (Bennett  etal.,  1976), 
epoxy  resins  (Humason,  1979),  and  other  low  viscosity  plastics 
(Hulet,  1978;  L.  R.  White  resin,  London  Resin  Company  Lim- 
ited) are  useful  embedding  media.  Small  specimens  that  can 
become  indistinguishable  or  even  lost  in  paraffin  blocks  can  be 
easily  observed  in  the  plastic  block  during  sectioning.  As  whole 
mounts,  specimens  can  be  examined,  measured,  and  meristic 
characters  counted  before  sectioning  (Hulet,  1978).  Techniques 
developed  by  Ruddell  (in  press)  reduce  swelling  of  tissues,  an 
artifact  sometimes  encountered  with  glycol  methacrylate 
embedding.  While  the  spectrum  of  histological  and  histochem- 
ical  stains  applicable  to  plastic  sections  is  somewhat  limited, 
toluidine  blue  counter  stained  with  acid  fuchsin  has  staining 
reactions  analogous  to  the  more  commonly  used  hematoxylin 
and  eosin.  Other  stain  combinations  also  are  applicable  to  larval 
tissue  embedded  in  glycol  methacrylate  (for  examples  see  Go- 
voni,  1980;  Govoni  et  al.,  1984):  alkali  blue  68  counter  stained 
with  neutral  red  reveals  fine  cellular  structure;  VanGiesen's 
picric  acid  counter  stained  with  acid  fuchsin  reveals  collagenous 
fibers,  the  anlagen  of  actinotrichia;  periodic  acid-Schiff  reagent 
reacts  strongly  with  acid  mucopolysaccharides,  including  chon- 
dromucin,  and  can  be  used  to  reveal  cartilaginous  precursors  of 
cartilage  (endochondral)  bone;  alizarin  red  S  reacts  with  Ca  +  + 
ions  and  can  reveal  both  calcified  cartilage  and  bone. 

Examples  of  Application 

Histological  preparations  may  serve  the  systematist  in  two 
ways:  by  clarifying  tissue  composition  and  by  resolving  struc- 
ture, thereby  allowing  for  the  determination  of  ontogenetic  pres- 
ence or  absence  of  tissues  and  by  offering  comparisons  of  tissue 
organization  among  taxa. 

An  example  of  the  first  use  is  in  the  identification  of  cartilage 
and  bone.  The  literature  is  replete  with  errors  that  result  from 
the  naive  interpretation  of  alcian  blue  and  alizarin  red  S  reac- 
tions with  cartilage  and  bone  tissue  in  whole  mounts.  Alcian 
blue  reacts  histochemically  with  the  sulfate  and  carboxyl  groups 
of  mucopolysaccharides  (Pearse,  1968)  including  chondromu- 
cin,  the  ground  substance  of  cartilage,  but  it  may  also  react  with 
developing  bone  matrices,  which  are  rich  in  mucopolysaccha- 
rides as  well  (Belanger,  1973).  An  alcian  blue  reaction,  therefore, 
may  indicate  cartilage  when  developing  membrane  (dermal)  bone 
is  present.  The  reaction  of  alizarin  red  S  with  calcium  ions 
(Pearse,  1968)  may  indicate  calcified  cartilage  as  well  as  true 
bone.  While  the  clearing  and  staining  of  skeletal  elements  re- 
mains a  powerful  tool  (Potthoff,  this  volume),  histological  prep- 
arations can  clarify  the  identity  of  cartilage  and  bone  tissue  in 
extremely  small  specimens  wherein  their  identity  may  not  be 
clear  in  whole  mounts. 

To  date,  comparisons  of  larval  fish  characters  revealed  by 
histological  techniques  have  not  been  extensive  and  examples 


of  application  are  few.  Comparative  histological  sections  of  elo- 
pomorph  and  clupeomorph  larvae  illustrate  the  unique  char- 
acter of  the  elopomorph  leptocephalus  (Smith,  this  volume). 
The  unique  configuration  of  organs  and  tissues  is  apparently 
inclusive  of  anguilliform,  elopiform,  and  notocanthiform  lep- 
tocephali.  Inasmuch  as  Hulet  (1978)  also  found  peculiarities  in 
the  kidney  structure  of  the  eel  leptocephalus  that  may  be  unique 
among  vertebrates,  the  kidney  structure  of  anguilliform  lepto- 
cephali  should  be  compared  with  that  of  other  elopomorph 
leptocephali.  Transient,  hyaline  plates  occur  in  the  basal  end  of 
the  outer  integumentary  epithelium  of  some  clupeiform  larvae 
(Jones  et  al.,  1966;  Lasker  and  Threadgold,  1968;  O'Connell, 
1981a;  Fig.  17 A),  but  this  feature  was  not  mentioned  in  the 
integumentary  descriptions  of  anguilliforms  (Hulet,  1978)  and 
pleuronectiforms  (Wellings  and  Brown,  1969;  Roberts  et  al., 
1973),  nor  is  it  apparent  in  the  perciform  Leiostomus  xanthunis 
(Fig.  1 7B).  These  plates  presumably  function  as  osmotic  barriers 
(O'Connell,  1981a),  but  their  systematic  presence  or  absence  is 
not  completely  established  and  remains  unexplained.  The  or- 
ganization of  axial  musculature  is  another  histological  difference 
among  higher  taxa.  The  two-layered  musculature  of  clupeiform 
larvae  is  aligned  in  opposing  directions  within  myotomal  seg- 
ments (Blaxter,  1969b;  O'Connell,  1981a;  Fig.  17C),  whereas 
in  perciform  larvae  the  orientation  of  axial  muscle  fibers  is 
closely  parallel  (O'Connell,  1981a;  Fig.  1 7D);  this  difference  may 
have  a  functional  basis  related  to  gross  body  form  and  swimming 
postures  (O'Connell,  1981a). 

An  example  of  the  use  of  histological  preparations  to  compare 
microanatomical  characters  is  the  differences  exhibited  in  elon- 
gate dorsal  fin  rays.  Elongate  dorsal  fin  rays  are  features  of  many 
unrelated  taxa  offish  larvae  (Moser,  1981),  but  the  microana- 
tomical structure  of  these  homologous  derivatives  differs  among 
taxa  (Govoni  et  al.,  1984).  A  major  difference  is  the  bilateral, 
paired,  collagenous  supporting  elements  of  the  carapid  elongate 
ray,  as  in  Echiodon  dawsoni  (Fig.  1 7E),  and  the  singular  supports 
of  elongate  rays  of  the  bregmacerotid  Bregmaceros  atlanticus 
(Fig.  17F)  and  the  serranid  Liopropoma  (Kotthaus,  1970). 
Monophyly  in  carapids  has  been  inferred,  in  part,  from  the 
distinctiveness  of  this  synapomorphy,  the  elongate  first  dorsal 
ray  of  their  highly  specialized  larvae  (OIney  and  Markle,  1979; 
Markle  and  OIney,  1980;  Gordon  et  al.,  this  volume). 

The  often  remarkable  similiarity  of  cells  and  tissues,  even 
among  phyla  (Andrew,  1959),  and  the  development  of  tissues 
from  the  undifferentiated  to  the  complex,  may  limit  the  use  of 
a  histological  approach  to  systematics.  Yet,  the  unusual  diversity 
that  characterizes  ontogenetic  patterns  of  fishes  (Wourms  and 
Whitt,  1981),  and  some  apparent  contrasts  in  tissue  organiza- 
tion and  composition  that  correlate  with  current  supraordinal 
classification,  make  histological  comparisons  tenable.  The  pre- 
ceding examples  of  tissue  and  microanatomical  dissimilarities 
may  serve  to  illustrate  the  kinds  of  comparisons  that  may  prove 
useful  in  inferring  relationships  as  more  information  becomes 
available.  Histological  techniques  may  provide  a  potentially 
useful  tool  to  the  systematist;  more  comparative  work  is  clearly 
warranted. 

National  Marine  Fisheries  Service,  Southeast  Fisheries 
Center,  Beaufort  Laboratory,  Beaufort.  North 
Carolina  28516. 


Scanning  Electron  Microscopy 

G.  W.  BOEHLERT 


SCANNING  electron  microscopy  is  an  ideal  tool  for  descrip- 
tion of  microstructure  in  taxonomic  studies.  The  scanning 
electron  microscope  (SEM)  provides  a  surface  image  character- 
ized by  high  resolution  and  depth  of  field  and  a  three-dimen- 
sional quality  unavailable  with  other  techniques.  In  many  cases 
this  allows  one  to  objectively  describe  microstructure  where  only 
subjective  descriptions  were  available  in  the  past.  It  is  the  pur- 
pose of  this  contribution  to  describe  the  techniques  and  use  of 
scanning  electron  microscopy  and  its  application  to  systematic 
investigations  of  fish  eggs  and  larvae. 

The  SEM  has  been  used  in  a  wide  variety  of  systematic  and 
evolutionary  investigations.  With  available  magnifications  from 
10  to  greater  than  100,000  times,  the  SEM  covers  the  range 
from  dissecting  and  compound  light  microscopy  to  transmission 
electron  microscopes.  It  has  thus  been  immensely  important  to 
progress  in  classification  in  the  study  of  micropaleontology,  bot- 
any, insects  and  mites,  and  a  wide  variety  of  microorganisms, 
among  other  taxa  (Heywood,  1971;  Kormandy,  1975).  Taxo- 
nomic applications  of  the  SEM  to  fishes  have  been  more  limited. 
Several  studies  have  used  the  SEM  for  studies  of  morphology, 
including  epidermis,  gill  tissue,  optic  capsules,  eggs,  sperm,  and 
embryosof  fishes  (Dobbs,  1974,  1975). 

Microstructural  analysis  of  otoliths  of  fishes  with  the  SEM  is 
now  common  (Pannella,  1 980).  For  early  life  history  stages,  the 
most  frequent  use  in  identification  and  classification  has  been 
with  the  egg  stage.  The  chorion,  or  external  membrane,  of  many 
species  is  variously  ornamented  with  filaments,  spines,  patterns 
of  ridges,  loops,  blebs,  and  pustules  ( Ahlstrom  and  Moser,  1 980; 
Robertson,  1981;  Matarese  and  Sandknop,  this  volume).  These 
ornamentations  and  the  ultrastructure  of  the  chorion  are  species- 
specific  (I  vankov  and  Kurdyayeva,  l973;Lonning,  1972).  While 
many  of  these  structures  may  be  easily  visualized  with  light 
microscopy  (Hubbs  and  Kampa,  1946;  Kovalevskaya,  1982), 
the  SEM  often  provides  the  best  means  of  adequately  describing 
structures  which  are  very  small  or  transparent  under  the  light 
microscope.  The  egg  chorion  of  Maurolicus  muelleri,  for  ex- 
ample, was  described  as  "drawn  up  into  hexagonally  arranged 
points,"  by  Robertson  (1976)  based  upon  light  microscopy  but 
as  "drawn  up  into  hexagonal  ridges  .  .  .  and  slightly  raised  at 
the  point  of  intersection"  under  the  SEM  (Robertson,  1981). 
Similarly,  Boyd  and  Simmonds  ( 1 974),  among  others,  suggested 
that  the  chorion  of  southern  populations  of  Fundulus  fietero- 
clitus  lacked  fibrils  using  light  microscopy,  whereas  the  SEM 
showed  the  presence  of  numerous  short  and  thin  fibrils  (Brum- 
mett  and  Dumont,  1981).  Thus  for  purposes  of  classification, 
the  SEM  allows  visualization  of  surface  structures  that  are  dif- 
ficult to  describe  with  light  microscopy. 

Methodology 

Preparation  of  biological  material  for  examination  under  the 
SEM  is  concerned  with  preservation,  dehydration,  and  coating 
with  a  conductive  material.  Fixation  of  labile  biological  speci- 
mens is  necessary  because  removal  of  water  during  the  stages 


of  dehydration  may  result  in  collapse  of  cells  and  other  artifacts. 
Depending  upon  the  method  of  fixation  and  dehydration,  the 
artifacts  can  range  from  shrinkage  to  collapse  or  fracture  of  the 
structures  to  be  observed.  It  is  preferable  to  begin  with  fresh, 
live  material.  For  eggs  this  requires  either  laboratory  spawning 
or  abundant  eggs  from  the  field  which  can  be  reliably  collected. 
For  larvae  at  different  stages,  it  is  diflicult  without  laboratory 
rearing  facilities.  Results  with  formalin-fixed  material  from 
plankton  collections  will  generally  be  satisfactory  for  lower  mag- 
nification analysis  of  surface  morphology,  but  may  not  reflect 
the  quality  of  freshly  prepared  material. 

Fresh  material  should  be  fixed  for  electron  microscopy.  Larval 
stages  may  first  be  relaxed  in  anesthetant  solution  (such  as  MS- 
222).  Initial  fixatives  for  both  eggs  and  larvae  are  generally  based 
upon  glutaraldehyde,  with  concentrations  ranging  from  0.5  to 
4.0%;  lower  concentrations  are  typically  followed  by  post-fix- 
ation. A  fixative  which  I  have  found  acceptable  is  that  from 
Dobbs  (1974)  as  follows:  70%  glutaraldehyde-2.0  ml,  flounder 
saline— 34  ml,  and  distilled  water— 34  ml.  The  flounder  saline 
follows  Forster  and  Hong  (1958)  and  contains  the  following  (in 
grams  per  liter):  NaCl,  7.890;  KCl,  0.186;  CaCK,  0.167;  MgCK- 
6H,0,  0.203;  NaH,FO,H_,0,  0.069;  NaHCO,,  0.84.  The  fix- 
ative has  a  final  osmolarity  of  380  mOsm/l.  Fixation  should  be 
for  24  hours.  Other  authors  provide  several  other  fixatives.  One 
suggested  by  Stehr  and  Hawkes  (1979),  while  more  difficult  to 
prepare,  is  also  useful  should  transmission  electron  microscopy 
be  desired  for  the  same  material.  Post-fixation  in  osmium  te- 
troxide  is  recommended  by  several  authors  as  a  means  of  hard- 
ening particularly  soft  tissues.  Generally,  1-2%  osmium  tetrox- 
ide  in  buffered  saline  is  used.  I  have  found  this  unnecessary  with 
fish  eggs  and  larvae,  as  suggested  by  Dobbs  (1974)  and  Stehr 
and  Hawkes  (1979).  It  may  be  considered,  however,  if  collapse 
is  a  problem.  Lonning  and  Hagstrom  (1975)  suggested  that  egg 
chorions  not  post-fixed  would  rupture  under  the  electron  beam; 
I  have  not  noticed  this. 

It  is  the  process  of  dehydration  where  the  greatest  artifacts 
are  likely  to  occur.  With  larvae,  shrinkage  of  tissue  may  occur, 
while  eggs  may  suffer  complete  collapse.  On  larger  eggs,  punc- 
turing the  chorion  with  a  sharpened  dissecting  needle  may  fa- 
cilitate transfer  of  fluids  and  prevent  this  collapse  (Stehr  and 
Hawkes,  1979). 

Removal  of  water  from  the  tissues  is  prerequisite  to  coating 
and  observation,  which  are  both  conducted  under  high  vacuum. 
Two  methods  are  available,  freeze  drying  and  critical  point 
drying.  For  freeze  drying,  unfixed  fresh  material  may  be  used. 
Fixed  material  should  first  be  rinsed  with  distilled  water  to 
remove  salts,  and  then  plunged  with  little  adhering  water  into 
liquid  nitrogen.  Damage  here  may  result  from  formation  of  ice 
crystals  if  freezing  rate  is  too  slow,  but  this  is  typically  not  a 
problem  with  small  eggs  and  larvae  in  liquid  nitrogen.  Boyde 
and  Wood  (1969)  recommend  using  20  ml  chloroform  per  liter 
of  distilled  water  to  increase  nucleation  rates  and  decrease  ice 
crystal  formation.  After  freezing,  the  material  is  immediately 


43 


44 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


introduced  into  the  freeze  dryer,  where  water  subUmes,  leaving 
the  specimen  dry  and  intact.  Critical  point  drying,  on  the  other 
hand,  requires  dehydration  through  a  graded  series  of  alcohols 
(20%  for  24  h.  then  10-20  min  each  in  50%,  70%,  80%,  90%, 
95%,  and  two  changes  of  absolute  ethanol).  The  ethanol  is  then 
replaced  with  either  freon  or  acetone  depending  on  whether 
freon  or  carbon  dioxide  critical  point  dryers  are  used.  The  steps 
of  dehydration  and  transfer  can  be  done  in  small  specimen 
holders  to  minimize  handling  and  possible  surface  damage.  Af- 
ter dehydration,  specimens  must  be  mounted  on  SEM  studs 
with  any  of  several  available  adhesives  and  tapes.  The  dried 
specimens  are  particularly  delicate  and  should  be  handled  with 
a  small  camel-hair  brush  to  avoid  damage  to  the  surface.  They 
are  then  oriented  onto  the  stud  under  a  dissecting  microscope. 
Before  coating,  no  further  preparation  is  necessary  with  larvae, 
but  eggs  have  only  a  small  area  of  electrical  contact  with  the 
stud.  It  is  therefore  advisable  to  use  a  conductive  adhesive  (such 
as  silver  paint)  to  make  a  more  complete  electrical  connection 
and  prevent  charging,  which  decreases  image  quality.  This  paint 
should  be  allowed  to  become  tacky  prior  to  positioning  the  eggs, 
or  it  may  cover  portions  of  the  egg  itself  Finally,  specimens  are 
coated  with  a  thin  conductive  layer,  typically  of  gold  or  gold- 
palladium,  by  either  vacuum  evaporation  or  ion  sputtering,  prior 
to  viewing  on  the  SEM.  At  most  facilities,  trained  SEM  tech- 
nicians are  available;  their  advice  and  assistance  are  invaluable 
and  should  be  sought. 


Results  and  Discussion 

Shrinkage  and  other  artifacts  will  vary  depending  upon  the 
type  of  material,  preservation,  and  method  of  dehydration.  For 
fresh  material  preserved  in  a  mixture  of  formalin,  glutaralde- 
hyde,  and  acrolein,  Stehr  and  Hawkes  ( 1979)  observed  a  shrink- 
age of  approximately  10%  in  the  eggs  of  Platichthys  stellatus 
and  Oncorhynchus  gorbuscha;  the  latter  had  been  punctured 
prior  to  dehydration.  In  the  present  study,  eggs  of  Maurolicus 
muellen  initially  preserved  in  5%  buffered  formalin  showed 
varying  degrees  of  shrinkage  and  collapse  depending  upon  sub- 
sequent treatment.  The  least  shrinkage  (12%,  Fig.  18B)  was 
noted  in  material  which  was  freeze  dried,  whereas  post-fixation 
and  dehydration  through  freon  1 1 3  associated  with  critical  point 
drying  resulted  in  shrinkage  of  up  to  67%  of  the  original  diameter 
(Fig.  18D).  Eggs  of  this  species  show  a  hexagonal  sculpturing; 
under  the  light  microscope  the  sculpturing  is  hyaline  and  difficult 
to  interpret  (Fig.  18A).  Eggs  prepared  by  freeze  drying  clearly 
show  the  surface  sculpturing;  note  particularly  the  ridges,  which 
are  more  clearly  defined  (Fig.  188).  For  comparison,  an  egg 
which  had  partially  collapsed  during  dehydration  is  shown  (Fig. 
18D).  The  obvious  differences  in  shrinkage  point  out  the  im- 
portance of  specifying  method,  initial  size,  and  shrinkage  values, 
particularly  for  comparative  or  taxonomic  studies. 

Eggs  from  other  species  are  shown  to  give  an  idea  of  the  range 
of  chorion  structures  which  may  be  observed.  The  hexagonal 
pattern  on  M.  muellen  overlies  a  highly  porous  surface  structure 


Fig.  18.  (A)  Egg  of  Maurolicus  muellen  from  off  South  Africa  taken  under  the  compound  light  microscope  with  transmitted,  polarized  light. 
Note  the  emphasis  of  the  points  on  the  hyaUne  chorion,  which  represent  the  intersections  of  ridges.  Bar  =  100  ^m.  (B)  Egg  of  A/,  muellen  under 
the  scanning  electron  microscope.  Note  the  areas  between  what  one  would  interpret  as  points  on  Figure  18A.  which  are  now  seen  as  polygonal 
facets  or  ridges.  Bar  =  500  nm.  (C)  Individual  facet  of  the  egg  of  At.  muellen.  Note  the  porous  and  diaphanous  nature  of  the  egg  surface.  Bar  = 
50  Mm.  (D)  Egg  of  A/,  muelleri  post-fixed  in  osmium  tetroxide  and  critical  point  dried.  The  shrinkage  of  this  specimen  is  approximately  65%. 
Note  the  differences  in  morphology  of  the  ridges  and  surface  of  the  egg.  Bar  =  100  /jm.  (E)  E^of  Pleuronichlhys  coenosus.  The  facets  are  relatively 
small  by  comparison  with  M.  muellen  and  the  pattern  units  are  more  regularly  hexagonal.  Bar  =  100  Mm.  (F)  Detail  of  two  hexagons  from  the 
egg  of  P.  coenosus.  Note  the  morphological  differences  between  both  the  ridges  and  chorion  surface  as  compared  to  M.  muellen.  Bar  =  10  Mm. 

Fig.  19.  (A)  Egg  of  Alherinopsis  californiensis.  The  filaments  are  single,  terminate  in  loose  ends,  and  are  distributed  over  the  entire  egg  surface. 
Bar  =  1 ,000  Mm.  (B)  Egg  of  .-itherinops  affiiUs.  The  egg  of  this  species  is  characterized  by  filaments  which  are  looped,  with  no  free  ends  (Curless, 
1979).  This  differentiates  it  from  the  egg  of  ,-1.  californiensis,  as  do  filament  length,  abundance,  and  basal  morphology.  Closed-loop  filaments  have 
also  been  noted  in  .Aniennanus  caudimaculatus  eggs  by  Pietsch  and  Grobecker  ( 1 980).  Bar  =  1 ,000  Mm.  (C)  Chorion  of  Paracaltionymus  costatus 
collected  off  South  Africa.  The  surface  features  are  irregular  and  cover  the  entire  egg  surface.  This  differs  from  species  of  Callionymus.  which 
have  hexagonal  patterns.  Bar  =  10  Mm.  (D)  Chorion  surface  of  Mugil  cephalus.  These  structures  are  irregular  and  cover  the  entire  egg  surface. 
Note  the  superficial  similarity  to  Paracallionymus.  Bar  =  10  Mm.  (E)  Chorion  surface  of  an  advanced  ovarian  egg  of  Coryphaenoides  filifer.  Note 
that  the  surface  "blebs"  are  arranged  in  hexagonal  patterns  and  may  be  the  precursors  of  a  hexagonal  pattern  typical  on  eggs  in  this  family.  The 
pelagic  egg  of  this  species  has  not  been  described.  Bar  =  10  Mm.  (F)  Chorion  surface  of  an  advanced  ovarian  egg  of  Coryphaenoides  acrolepis. 
The  hexagonal  ridges  are  better  developed  than  in  Fig.  I9E.  There  are  holes  under  the  ndges  between  the  intersections,  which  might  indicate  that 
this  species,  whose  egg  is  also  undescribed,  may  have  the  hexagonal  network  supported  on  "stills"  as  described  for  eggs  of  Coelorhynchus  spp. 
(Robertson.  1981;  Sanzo,  1933a).  Bar  =  10  Mm. 


Fig.  20.  (A)  Spines  on  the  chorion  surface  o(  Oxyporhamphus  microplerus.  These  are  distributed  over  the  entire  surface  of  the  egg.  Bar  =  100 
Mm.  (B)  Chorion  surface  from  Scomhereso.x  saurus  collected  off  South  Africa.  The  tufts  are  characterized  by  a  relatively  complex  basal  morphology 
and  depending  upon  method  of  fixation,  may  resemble  small  bundles  of  hairs  or,  as  here,  simply  coalesced  tufts.  Bar  =  10  Mm.  (C)  Micropyle 
and  associated  pores  of  the  egg  of  Laclona  diaphana  from  the  Eastern  Tropical  Pacific.  The  pores  shown  here  are  restricted  to  this  region  around 
the  micropyle  and  appear  to  penetrate  the  outer  layer  of  the  chorion.  Bar  =  50  Mm.  (D)  Secondary,  smaller  pit  structures  on  the  remainder  of  the 
egg  of  Laclona  diaphana.  I  refer  to  these  depressions  as  "pits"  because  closer  examination  does  not  reveal  penetration  through  any  layer  of  the 
chorion,  as  opposed  to  the  pores  surrounding  the  micropyle  in  20C.  Bar  =  1  Mm.  (E)  Head  region  of  a  larval  Sebasles  melanops  shortly  after 
parturition.  Polygonal  epidermal  cells  may  be  noted  on  some  parts  of  the  body.  Bar  =  100  Mm.  (F)  Epidermis  on  the  dorsal  surface,  just  posterior 
to  the  head,  on  an  embryonic  S.  melanops  approximately  28  days  post  fertilization.  Note  the  distinct  microndges  and  cell  borders  characteristic 
of  developing  teleost  epidermis.  Bar  =  10  Mm. 


BOEHLERT:  SCANNING  ELECTRON  MICROSCOPY 


45 


•/N  .\!^^V-U 


46 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES -AHLSTROM  SYMPOSIUM 


K 


BOEHLERT:  SCANNING  ELECTRON  MICROSCOPY 


47 


48 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES -AHLSTROM  SYMPOSIUM 


(Fig.  18C)  as  compared  to  that  oi Pleuronichthys  coenosus  (Fig. 
18E,  F).  Here,  the  hexagons  are  not  only  smaller,  but  the  area 
within  the  facets  does  not  appear  porous.  SEM  was  used  for  this 
species  and  its  congeners  for  egg  description  by  Sumida  et  al. 
(1979).  It  is  interesting  to  note  that  these  authors  discussed  the 
similarity  in  chorion  structure  of  Plenronichthys  spp.  with  that 
oi Synodus  lucioceps.  While  there  were  slight  differences  in  sizes 
of  the  polygons,  the  superficial  similarity  of  chorion  structure 
on  these  phylogenetically  distant  genera  supports  a  functional 
role  (Robertson,  1981)  and  independent  derivation.  In  this  in- 
stance, however,  SEM  was  valuable  for  understanding  and  in- 
terpreting the  differences  between  species  and  genera  subse- 
quently observed  under  the  light  microscope  (Sumida  et  al., 
1979).  Similarly,  Keevin  et  al.  (1980)  used  chorion  ornamen- 
tation to  distinguish  among  genera  of  killifishes. 

Other  ornamentations  include  more  random  ridges  (Para- 
callionymus  costatus.  Fig.  19C,  and  Mugil  cephalus.  Fig.  19D), 
filaments  of  varied  length,  diameter,  and  base  morphology  (Ath- 
erinopsis  califormensis  and  Athehnops  affinis.  Fig.  19A,  B;  see 
also  Hubbs  and  Kampa,  1946),  tufts  (Scomberesox  saurus.  Fig. 
20B),  spines  (Oxyporhamphus  microptents.  Fig.  20A),  and  pits 
and  pores  (Lactoria  diaphana.  Fig.  20C,  D).  In  thecallionymids, 
the  small  eggs  of  species  of  Callionynms  have  hexagonal  sculp- 
turing similar  to  that  oi  Pleuronichthys  (Fig.  18E).  In  Paracal- 
lionymus  costatus  (Fig.  19C),  however,  random  ridges  similar 
to  those  in  Mugil  cephalus  are  apparent. 

Since  chorion  microstruclure  is  formed  by  follicle  cells  during 
oogenesis  (Sponaugle  and  Wourms,  1979;  Stehr,  1979),  patterns 
may  also  be  discerned  in  ovarian  eggs.  The  pelagic  eggs  of  mac- 


rourids  are  poorly  known  but  have  been  described  for  selected 
species  by  Sanzo  ( 1 933a),  Robertson  ( 1 98 1 ),  and  Grigor'ev  and 
Serebryakov  (1981).  For  Pacific  species  of  Coryphaenoides.  pe- 
lagic eggs  remain  poorly  known  but  apparently  have  hexagonal 
patterns  as  in  other  members  of  the  genus;  this,  is  clearly  shown 
in  ovarian  eggs  near  the  maximum  size  observed  by  Stein  and 
Pearcy  (1982;  Fig.  19E,  F).  Thus  SEM  of  developing  ovarian 
eggs  may  be  used  to  discern  differences  which  then  aid  in  iden- 
tification of  eggs  from  plankton  samples. 

For  larval  stages,  SEM  has  been  used  for  the  description  of 
development  of  several  surface  structures,  such  as  the  olfactory 
organ  (Elston  et  al.,  1981)  and  lateral  line  neuromasts  (Dobbs, 
1974).  For  taxonomic  studies,  differentiation  of  fine-scale  mor- 
phological differences,  such  as  dentition  or  fine-scale  spine  ser- 
ration, may  be  useful.  Its  most  valuable  use  may  therefore  be 
for  later  larval  development,  since  pigmentation  and  other  char- 
acteristics in  early  larvae  are  better  seen  with  conventional 
methods  (Fig.  20E,  F). 

To  conclude,  SEM  may  serve  as  an  adj  uct  to  traditional  meth- 
ods in  the  description  of  fine  structure  in  fish  eggs  and  larvae. 
For  high  magnification,  high  resolution  visualization  of  surface 
morphology,  it  remains  the  most  effective  tool  available.  Under 
lower  magnifications,  it  may  allow  one  to  clearly  visualize  struc- 
tures which  are  difficult  to  interpret  using  standard  microscop- 
ical methods  (Fig.  1 8A,  B). 

Oregon  State  University  ,  Marine  Science  Center,  Newport, 
Oregon  97365. 


Developmental  Osteology 
J.  R.  Dunn 


ONE  legacy  left  by  Elbert  H.  Ahlstrom  was  an  appreciation 
of  the  value  of  developmental  osteology  of  teleosts  as  a 
taxonomic  aid  and  as  an  indicator  of  phylogenetic  affinities. 
Although  numerous  studies  have  been  made  on  the  growth  of 
various  bones  in  teleosts,  such  descriptions  have  not  been  widely 
used  in  assessing  relationships  of  fishes.  I  have  recently  re- 
viewed, in  some  depth,  the  application  of  developmental  os- 
teology in  taxonomic  and  systematic  studies  of  teleosl  larvae 
(Dunn,  1983b).  Here  I  present  a  brief  overview  of  some  skeletal 
structures  in  teleosts  whose  ontogeny  offers  potential  utility  in 
inferring  phylogenetic  affinities.  It  is  hoped  that  this  precis  will 
encourage  ichthyologists  to  examine  the  development  of  bones 
in  the  course  of  their  systematic  studies. 

Ontogenetic  Changes  in  Skeletal 
Structures 

Cranial  and  associated  bones— CTaniaX  osteology  has,  of  course, 
been  the  foundation  of  systematic  studies  of  adult  fishes,  but 
the  development  of  cranial  bones  has  been  little  used  in  phy- 
logenetic studies.  Numerous  descriptions  of  the  ontogeny  of 


cranial  bones  exist  in  the  literature  (e.g.,  Bhargava,  1958;  Bert- 
mar,  1959;  Kadam,  1961;  Weisel,  1967;  Moser  and  Ahlstrom, 
l970;Mook,  l977;Leiby,  1979b;  Yuschak,  1982).  Additionally, 
the  sequence  of  ossification  of  head  bones  has  been  described 
for  a  variety  of  taxa  (e.g.,  Moser,  1972;  Aprieto,  1974;  Leiby, 
1979a;  Dunn,  1983a;  Kendall  and  Vinter,  1984).  The  devel- 
opment of  certain  cranial  structures  has  also  been  shown  to  be 
of  taxonomic  value  (Fritzsche  and  Johnson,  1980),  yet  com- 
parative studies  of  the  developmental  osteology  of  the  skull  of 
related  groups  of  teleosts  seem  rare  (e.g.,  Norman.  1926b;  De 
Beer,  1937). 

Available  evidence  suggests  that  the  sequence  of  ossification 
of  the  skull  of  teleosts  is  a  conservative  (i.e.,  relatively  constant 
among  different  phyletic  groups)  process  (De  Beer,  1 937;  Mook, 
1977).  Among  the  bones  which  ossify  first  are  those  in  areas  of 
high  stress,  such  as  feeding  (jaw  bones)  and  respiration  (bran- 
chial region),  as  noted  by  De  Beer  ( 1 937),  Weisel  ( 1 967),  Moser 
and  Ahlstrom  (1970),  Mook  (1977),  Yuschak  (1982). 

Examples  of  ontogenetic  changes  in  skull  bones  which  suggest 
that  these  structures  might  offer  insight  into  phylogenetic  affin- 


DUNN:  DEVELOPMENTAL  OSTEOLOGY 


49 


ities  include  upper  jaw  bones  (Berry,  1 964a),  head  spines  (Ken- 
dall, 1979;  Washington,  1981;  Yuschak,  1982;  Washington  and 
Richardson,  MS),  gill  arches  (Leiby,  1979b;  Yuschak,  1982; 
PotthofTet  al.,  1984),  and  lateral  skull  bones  (Leiby,  1979b). 

Patterns  of  chondrification  may  also  be  of  value  in  inferring 
phylogenetic  relationships.  Washington  and  Richardson  (MS) 
noted  that  while  chondrification  of  skeletal  bones  in  most  scor- 
paeniform  fishes  is  a  relatively  brief  process,  occurring  in  pre- 
flexion  and  early  flexion  larvae,  chondrification  was  prolonged 
(occurring  through  most  larval  development)  in  hexagrammids 
and  in  three  genera  of  cottids.  These  authors  also  considered  a 
unique  pattern  of  ossification  of  cartilaginous  rings  in  the  regions 
of  the  parietal  and  frontal  spines  as  a  synapomorphic  character 
uniting  three  genera  of  cottids. 

Vertebral  column  and  associated  bones.  —  Vertebral  centra,  neural 
and  haemal  spines,  apophyses,  and  ribs  all  undergo  variable 
changes  in  configuration  with  growth.  A  number  of  workers  have 
documented  the  development  of  the  vertebral  column  and  as- 
sociated bones  in  a  variety  of  taxa,  but  attempts  have  not  been 
made  to  analyze  the  phylogenetic  significance  of  the  ontogeny 
of  these  structures.  The  sequence  and  direction  of  ossification 
of  vertebral  centra  is  known  to  vary  among  taxa  (e.g.,  Moser 
and  Ahlstrom,  1970;  Mook,  1977;  Potthoff"  et  al.,  1984),  but 
this  character  has  yet  to  be  analyzed  among  groups  of  fishes. 

Among  those  elements  of  the  vertebral  column  which  have 
been  studied  in  various  taxa,  Potthoff"and  Kelley  (1982)  noted 
that  the  neural  and  haemal  arches  in  Xiphias  first  develop  dis- 
tally  opened,  whereas  in  other  perciforms  studied,  split  arches 
were  observed  in  small  larvae  on  the  anterior  two  centra  only. 
Washington  and  Richardson  (MS),  in  their  study  of  cottid  larvae 
and  scorpaeniform  outgroups,  noted  in  various  taxa  the  reduc- 
tion or  absence  of  the  first  neural  spine,  presence  or  absence  of 
autogenous  neural  arches  on  centrum  one,  shape  of  anterior 
neural  arches,  and  whether  or  not  the  first  neural  arch  was 
distally  fused  or  open.  Potthoff"  and  Kelley  (1982)  cited  the 
unique  position  and  development  of  ribs  in  Xiphias  compared 
to  other  perciforms  studied,  and  Washington  and  Richardson 
(MS)  examined  the  location,  number,  and  position  of  ribs  in 
cottids  and  perciform  outgroups. 

Fins  and  their  supports— Y>OTsaX  and  anal  fins— The  sequence 
of  formation  of  dorsal  and  anal  fins  as  well  as  the  order  of 
development  of  their  constituent  spines  and/or  rays  varies  among 
taxa  (Dunn,  1983b).  This  succession  of  formation  may  be  rel- 
atively constant  among  related  groups  or  it  may  vary,  but  the 
phylogenetic  significance  of  these  events,  if  any,  has  yet  to  be 
analyzed.  Additionally,  numerous  taxa  of  larvae  possess  tran- 
sient, often  bizzare,  structures,  such  as  elongate  dorsal  spines 
or  rays  or  anal  rays  (e.g.,  Kendall,  1979;  Moser,  1981).  These 
structures  are  of  taxonomic  value  and  may  contain  phylogenetic 
information,  but  the  homologies  of  these  structures,  if  any,  are 
not  known  (Govoni,  this  volume). 

PotthoflTet  al.  (1984)  indicated  that  the  second  dorsal  and 
anal  fins  are  the  first  to  develop  in  most  perciform  fishes.  How- 
ever, in  generally  more  advanced  species,  dorsal  fin  rays  (or 
spines)  develop  first  anteriorly  and  second  dorsal  and  anal  fin 
ray  development  starts  after  the  first  dorsal  fin  is  either  partially 
or  fully  developed.  Fahay  and  Markle  (this  volume)  described 
the  sequence  of  fin  formation  in  gadiform  fishes.  Usually  the 
vertical  fins  ossify  at  nearly  the  same  time,  but  two  or  more 
centers  of  ossification  are  present  in  those  genera  (e.g.,  Molva. 


Merluccius)  with  a  single  long  dorsal  fin  (or  a  short  first  dorsal 
fin  preceding  a  longer  second  dorsal  fin). 

The  ontogeny  of  pterygiophores  has  received  considerable 
attention  from  Potthofl"and  colleagues  (e.g.,  PotthofT.  1975,  1980; 
Potthoff'et  al.,  1980,  1984).  The  developmental  pattern  of  fin 
pterygiophores  may  suggest  phylogenetic  relationships.  PotthofT 
and  Kelley  (1982)  noted  that  the  first  dorsal  pterygiophore  in 
Xiphias  arose  from  either  one  or  two  pieces  of  cartilage,  as  is 
the  case  in  Morone  (Fritzsche  and  Johnson,  1 980),  but  not  in 
scombrids.  Washington  and  Richardson  (MS)  observed  the  on- 
togenetic migration  of  dorsal  fin  pterygiophores,  relative  to  neu- 
ral arch  position,  in  three  cottid  genera.  Proximal  and  distal 
radials  may  fuse  during  ontogeny  (Yuschak,  1982)  and  the  pres- 
ence or  absence  of  medial  radials  may  characterize  certain  groups 
of  fishes  (PotthofT  and  Kelley,  1982). 

Pectoral  and  pelvic  fins  and  their  supports.— 'Wilh  some  excep- 
tions, pectoral  fins  develop  rays  later  in  the  larval  period  than 
median  fins  (Dunn,  1983b).  Transient,  elongate  spines  and 
rays  also  develop  in  the  pectoral  fins  of  some  taxa  (Moser  and 
Ahlstrom,  1974;  Moser,  1981);  such  structures  may  be  of  taxo- 
nomic value,  but  their  phylogenetic  significance,  if  any,  and  their 
homologies  are  not  known.  Relatively  few  descriptions  have 
been  published  on  the  development  of  the  pectoral  fin  (e.g., 
Houdeand  PotthofT,  1976;  Potthoff",  1980;  Potthoff"and  Kelley, 
1982;  Yuschak,  1982;  Potthofl["et  al.,  1984),  and  few  systematic 
inferences  have  been  drawn.  PotthofTet  al.  (1984)  noted,  in 
Anisotremus  virginicus.  the  ontogenetic  fusion  of  the  supratem- 
poral-intertemporal,  the  elongation  of  the  anterior  coraco-scap- 
ular  cartilage,  and  the  reduction  in  length  of  the  posterior  pro- 
cess. Washington  and  Richardson  (MS)  examined  the  orientation 
of  the  cleithrum,  as  well  as  its  outer  lip,  the  length  of  the  scapula- 
coracoid  complex,  the  base  of  the  cleithrum,  and  the  cleithral 
extension  over  the  pelvic  bone  (among  other  characters  of  the 
pectoral  girdle)  in  their  analyses  of  cottids  and  their  allies. 

The  ontogeny  of  the  pelvic  fin  and  its  supporting  structures 
also  has  been  little  investigated  (PotthofT,  1980;  PotthoflTet  al., 
1980;  Fritzsche  and  Johnson,  1980)  and  infrequently  used  in 
systematic  studies.  Dunn  and  Matarese  (this  volume)  indicated 
that  in  gadid  larvae  the  length  of  the  posterior-lateral  process 
of  the  basipterygia  differed  among  subfamilies  and  tended  to  be 
reduced  or  wanting  in  those  genera  presently  considered  ad- 
vanced. 

Caudal  fin.— The  development  of  the  caudal  fin  in  teleosts,  a 
subject  Dr.  Ahlstrom  was  extremely  interested  in  (e.g.,  Ahlstrom 
and  Moser,  1976),  seems  to  have  received  more  study  than  other 
bony  structures.  However,  few  workers  have  attempted  to  in- 
terpret the  phylogenetic  significance  of  the  development  of  this 
fin  (Dunn,  1983b). 

The  fusion  of  bones,  reduction  in  size  of  structures,  or'loss 
of  elements  by  absorption  can  frequently  be  observed  in  the 
development  of  the  caudal  fin  in  some  fishes.  Additionally,  based 
on  ontogenetic  evidence,  the  structure  of  this  fin  may  differ  from 
that  commonly  accepted  based  on  adult  specimens  (Dunn, 
1983b). 

Ontogenetic  changes  in  the  caudal  fin  and  associated  bones 
which  have  been  used  to  infer  phylogenetic  relationships  include 
the  reduction  through  fusion  of  ural  centra  (Moser  and  Ahl- 
strom, 1 970;  and  others),  discreet  or  fused  hypural  bones  (Wash- 
ington and  Richardson,  MS;  Dunn  and  Matarese,  this  volume), 
absence  of  the  parhypural  in  certain  taxa  which  normally  possess 
one  (Washington  and  Richardson,  MS),  characteristics  (e.g.. 


50 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES -AHLSTROM  SYMPOSIUM 


shape,  modification,  autogenous  or  fused  to  the  centra)  of  neural 
and  haemal  spines  on  preural  centra  associated  with  the  caudal 
fin  (Washington  and  Richardson,  MS;  Dunn  and  Matarese,  this 
volume),  and  number  of  vertebral  centra  supporting  the  caudal 
fin  (Washington  and  Richardson,  MS;  Fahay  and  Markle,  this 
volume). 

Attention  has  recently  been  directed  toward  the  presence  of 
radial  cartilages  (their  position  and  shape  during  development) 
in  the  caudal  fin  of  certain  teleosts  (Kendall';  PotthofT  et  al., 
1984).  These  structures  may  contain  information  of  value  in 
assessing  phylogenetic  relationships. 

Squamation.—The  development  of  scales  in  teleosts  has  been 
described  for  a  variety  of  taxa  (e.g..  Berry,  1960;  Burdak,  1969; 
Fujita,  1971;  White,  1977;  Potthofl'and  Kelley,  1982).  The  se- 
quence of  development  of  scales  and  their  origin  on  the  fish 
differs  among  taxa,  and  scales  undergo  changes  with  ontogeny 
(e.g..  White,  1977;  Potthoffand  Kelley,  1982).  The  acquisition 


'  Kendall,  A.  W.,  Jr.  1981.  Ventral  caudal  radials— oft  overlooked 
structures.  (Paper  presented  at  annual  meeting  Amer.  Soc.  Ichthyol. 
Herpetol.,  Corvallis,  OR,  June  1981;  Abstract  in  Copeia  1981:935). 


of  scales  on  fish  usually  occurs  during  their  transformation  to 
the  juvenile  stage;  however,  a  number  of  groups  (e.g.,  Zaniolepis. 
serranids,  holocentrids,  and  xiphiids)  acquire  scales  during  the 
larval  period.  Such  developmental  changes  have  apparently  not 
been  analyzed  among  diverse  groups  of  fishes. 

Perspective 

Developmental  osteology  of  teleosts  appears  to  be  an  under- 
exploited  approach  of  potential  value  in  increasing  our  under- 
standing of  the  relationships  of  fishes.  Studies  of  developmental 
osteology  of  teleosts  may  contribute  much  to  our  understanding 
of  homology,  the  central  concept  of  all  biological  comparisons 
(Inglis,  1966;  Bock,  1969;  Wake,  1979)  in  our  search  for  prim- 
itive and  derived  character  states.  A  number  of  investigators 
present  at  this  symposium  are  actively  engaged  in  evaluating 
ontogenetic  changes  in  ossified  structures  in  their  studies  of 
various  taxa  of  larval  fishes.  An  appraisal  of  this  method  may 
well  be  in  the  future,  but  evidence  provided  during  the  course 
of  this  meeting  will  contribute  to  such  an  evaluation. 

Northwest  and  Alaska  Fisheries  Center,  National  Marine 
Fisheries  Service,  2725  Montlake  Boulevard  East, 
Seattle,  Washington  981 12. 


Otolith  Studies 
E.  B.  Brothers 


ALTHOUGH  the  value  of  otolith  studies  in  systematic  ich- 
thyology is  well  established,  essentially  all  studies  to  date 
deal  with  the  otoliths  of  adults,  or  only  incidentally  juveniles, 
and  are  usually  limited  to  the  external  morphology  of  the  typ- 
ically largest  otolith,  the  sagitta  (see  reviews  of  Weiler,  1968; 
Casteel,  1974;  Hecht,  1978;  Huygebaert  and  Nolf,  1979).  Oto- 
liths of  larvae,  which  are  of  recent  interest  in  terms  of  age, 
growth,  mortality,  and  life  history  studies  (Brothers  et  al.,  1976; 
Struhsaker  and  Uchiyama,  1976;  Methot  and  Kramer,  1979; 
Townsend  and  Graham,  1981;  Kendall  and  Gordon,  1981;  La- 
roche  et  al.,  1982;  Lough  et  al.,  1982;  Bailey,  1982;  Brothers  et 
al.,  1983)  have  been  ignored  from  a  taxonomic  point  of  view. 
This  is  perhaps  not  surprising  due  to  their  very  small  size  and 
generally  simpler  form,  with  an  apparent  lack  of  obvious  dis- 
tinguishing external  features.  Although  the  internal  structure  of 
larval  otoliths  appears  to  be  more  variable  than  the  external 
form,  no  comparative  taxonomic  studies  have  been  attempted 
to  date.  In  addition,  relatively  little  has  been  done  on  compar- 
isons of  these  features  of  adult  otoliths,  noting  that  in  a  real 
sense,  the  internal  anatomy  of  the  adult  otolith  is  just  the  cu- 
mulative historical  record  of  ontogenetic  changes  in  external 
structure  and  growth  patterns.  Comparative  studies  on  features 
other  than  external  appearance  have  tended  to  be  at  the  crys- 
tallographic,  mineraiogical  and  chemical  level.  Carlstrom's  ( 1 963) 
research  on  the  crystallographic  structure  of  fish  otoliths  and 
otoconia  was  a  pioneering  attempt  to  apply  structural  and  com- 


positional information  to  understanding  the  broad  outlines  of 
vertebrate  evolution.  A  few  studies  have  followed  this  line  of 
investigation  (Lowenstam,  1980,  1981;  Lowenstam  and  Fitch, 
1978,  1981),  however  the  discrimination  ability  of  crystallo- 
graphic techniques  is  certain  to  be  limited  by  the  relatively  few 
crystalline  varieties  known  to  exist  in  ear  stones.  Analysis  of 
the  amino  acid  composition  of  the  major  organic  fraction  of 
otoliths  (Degens  et  al.,  1969)  offers  another  possibility  for  taxo- 
nomic information,  however  it  is  unlikely  to  be  useful  for  spe- 
cific identification  of  individuals.  Finally,  trace  element  analysis 
of  otoliths  (Gauldie  et  al.,  1980;  Papadopoulou  et  al.,  1978, 
1980)  may  allow  for  stock  and  perhaps  species  discrimination, 
but  again  the  small  sample  sizes  offered  by  larval  otoliths  impose 
severe  or  impossible  methodological  problems  unless  x-ray  mi- 
croprobes  or  ion  microscopes  are  employed.  New  analytic  tools 
for  chemical  studies  could  offer  unique  insights  into  fish  sys- 
tematics. 

Recently  renewed  interest  in  fish  otoliths,  due  primarily  to 
the  recognition  of  daily  growth  increments  (Pannella,  1971, 
1980).  has  resulted  in  an  expanding  effort  toward  collecting, 
examining  and  cataloging  the  otoliths  of  larval  fishes.  As  we 
begin  to  study  the  external  and  internal  structure  of  this  material 
for  systematically  useful  characters,  we  should  begin  to  develop 
a  new  set  of  morphological  criteria  for  species  identification, 
taxonomic  relationships,  and  perhaps  phylogenetic  reconstruc- 
tion. 


BROTHERS:  OTOLITH  STUDIES 


51 


Fig.  2 1  Abrupt  changes  in  external  and  internal  morphology  of  the  sagitta  associated  with  the  end  of  the  larval  stage.  (A)  Scanning  electron 
micrograph  of  the  medial  face  of  the  left  sagitta  (9  mm  SL)  of  a  french  grunt  {Haemuton  flavohtwatum).  (B)  12  mm  SL,  showing  development 
of  "secondary  growth  centers."  (C)  Enlargement  of  area  in  previous  specimen.  (D)  44  mm  SL.  Scale  omitted:  12  mm  =  500  ixm.  (E)  SEM  of 
ground  and  etched  hake  (Merluccius  sp.)  sagitta.  showing  growth  centers  around  the  larval  otolith.  (F)  Photomicrograph  of  ground  sagitta  of  a 
largemouth  bass,  Micropterus  salmoides.  The  larval  portion  of  the  otolith  is  in  the  lower  right  comer. 


52 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  22.     Photomicrographs  of  otoconia  in  teleosts.  (A)  Bonefish,  Albula  vulpes.  free  otoconia.  (B)  Bonefish,  otoconia  embedded  m  the  sagitta. 


General  Methodology 

The  otoliths  (sagittae,  lapilli,  and  asterisci)  of  larval  fish  are 
usually  the  first  calcified  structures  to  appear  in  the  development 
of  an  individual.  At  least  some  of  the  otoliths  are  frequently 
evident  before  hatching.  Over  the  larval  life,  they  vary  in  size 
from  a  few  to  several  hundred  micra  for  different  taxa  and  ages. 
Because  of  their  composition  and  small  size  (high  surface  to 
volume  ratio),  larval  otoliths  are  very  sensitive  to  degradation, 
decalcification,  and  dissolution  in  acidic  solutions  (McMahon 
and  Tash,  1979),  and  great  care  must  be  exercised  in  preserving 
larval  fish  and  otoliths.  Improper  handling  results  in  rapid  and 
irreversible  damage.  Fresh  larvae  are  best  stored  for  later  otolith 
extraction  in  three  ways:  1 )  frozen,  2)  fixed  and  maintained  in 
strong  ethanol  solutions  (preferably  95%),  3)  dried  (e.g.,  on  glass 
slides).  The  last  technique  is  least  preferred  due  to  increased 
difficulties  in  otolith  removal  and  general  damage  to  the  larvae. 
Removal  from  embryos  and  larvae  involves  microscopic  dis- 
section with  fine  needles.  The  use  of  crossed  polarized  filters  is 
sometimes  helpful  in  locating  the  otoliths,  although  they  are 
generally  clearly  visible  in  the  otocysts  or  otic  capsules  with 
standard  transmitted  illumination.  Dissection  is  best  carried  out 
in  water,  and  opaque  larva  can  be  cleared  by  brief  exposure  to 
a  weak  KOH  (1%)  solution.  Air  dried  otoliths  should  be  trans- 
ferred on  the  tips  of  oil  wetted  (immersion)  needles,  and  for 
light  microscopy  may  be  stored  in  oil  on  slides  or  permanently 
mounted  under  coverslips  with  a  neutral  medium  (non-acidic). 
In  the  latter  case,  care  must  be  taken  to  prevent  the  otoliths 
from  being  cracked  or  crushed  as  the  mounting  medium  shrinks 
and  pulls  down  the  coverslip.  In  most  cases  larval  otoliths  are 
small  and  thin  enough  to  preclude  a  need  for  grinding.  Light 
microscopy  is  best  applied  to  studies  of  internal  structures,  al- 
though some  external  features  can  be  viewed  with  either  surface 
microscopy  or  transmitted  light  and  wide  openings  of  the  con- 
denser diaphragm.  Compound  microscopes  should  have  high 
quality  oil  immersion  optics  (preferably  to  at  least  1 ,000  x )  and 
polarizing  filters.  For  the  latter,  a  single,  rotatable  field  polarizer 
helps  in  resolving  internal  structures,  while  an  analyzing  polar- 
izer can  be  employed  to  locate  the  very  small,  but  highly  bire- 
fringent  otoliths  on  slides.  A  moderately  high  resolution  (at  least 
500  lines)  black  and  white  video  system  is  an  additional,  but 
invaluable  accessory.  Such  a  system  reduces  eye  fatigue,  sim- 


plifies group  viewing,  measurement  and  photography,  and  most 
importantly  can  substantially  enhance  image  quality  by  elec- 
tronic adjustment.  It  is  also  a  necessary  component  in  a  variety 
of  automatic  and  semi-automatic  image  analysis  systems. 

Scanning  electron  microscopy  is  most  useful  for  high  reso- 
lution views  of  external  structures,  for  examination  of  fine  (<  1 
fim)  internal  features,  and  for  confirmation  of  suspected  optical 
artifacts.  However  the  technique  is  also  more  expensive  and 
time  consuming  and  may  necessitate  critical  preparation.  Whole, 
cleaned  and  air-dried  otoliths  can  be  mounted  and  coated  by 
standard  techniques.  Internal  views  require  embedding,  grind- 
ing, polishing  and  etching  before  stub  mounting  and  coating. 
The  most  recent  important  development  in  SEM  preparation 
is  the  use  of  etching  solutions  other  than  the  initially  preferred 
HCl.  Haake  et  al.  (in  press)  summarize  a  technique  for  SEM 
preparation  of  larval  otoliths. 

Otolith  Morphology  and  Early  Ontogeny 

There  are  a  number  of  papers  which  deal  with  the  general 
structure  and  composition  (Hickling,  1931;  Degens  et  al.,  1969: 
Blackler,  1974:  Pannella,  1980),  mechanism  of  growth  (Irie, 
1960:  Dunkelburgeretal.,  1980;  Campana,  1983),  and  functions 
of  the  otoliths  and  otolith  organs  (Popper  and  Coombs,  1 980a, b; 
Piatt  and  Popper,  1981).  This  work  has  not  specifically  dealt 
with  larvae,  however  the  gross  morphology  and  processes  should 
be  comparable  with  older  fishes. 

The  otic  capsule  or  otocyst  forms  very  early  in  the  ontogeny 
of  fishes  and  is  an  obvious  landmark  in  the  head  of  newly 
hatched  larvae.  The  earliest  evidence  of  the  otoliths  is  one  to 
several  small  (usually  less  than  10  ixm)  optically  dense  bodies, 
referred  to  here  as  primordia.  From  their  physical  appearance 
and  etching  properties,  the  primordia  are  assumed  to  be  sub- 
stantially composed  of  organic  matrix  (probably  the  fibroprotein 
otolin),  and  are  soon  calcified  and  surrounded  by  an  accreted 
layer  of  calcium  carbonate  and  matrix.  There  are  distinct  dif- 
ferences between  certain  taxa,  usually  at  the  supraspecific  level, 
with  regards  to  the  morphology  of  the  primordia.  Distinctions 
also  exist  between  the  sagitta,  lapillus,  and  asteriscus,  so  com- 
parative studies  must  be  careful  to  properly  identify  the  otoliths 
examined.  Variation  in  primordial  form  involves  the  size,  shape, 
and  number  per  otolith.  Surrounding  the  primordium  (partic- 


BROTHERS:  OTOLITH  STUDIES 


53 


Fig.  23.  Otolith  primordia  and  cores.  (A)  SEM  of  single  primordium  and  core  in  a  french  grunt  (Haemulon  flavolineatum)  lapillus.  (B) 
Photomicrograph  of  single  primordium  and  core  in  a  mimic  blenny  {Labrisomus  guppyi)  sagitta.  (C)  Multiple  primordia  in  the  lapillus  of  a  white 
sucker  {Caloslomus  commersoni).  (D)  Multiple  primordia  in  the  sagitta  of  a  seahorse  (Hippocampus  sp.).  (E)  Multiple  primordia  and  cores  in  the 
lapillus  of  a  banded  killifish  (Fiindulus  diaphamis).  (F)  SEM  of  multiple  primordia  and  cores  in  the  sagitta  of  a  rainbow  trout  {Salmo  gairdneri). 


ularly  in  the  sagitta  and  lapillus)  is  a  discrete,  relatively  ho- 
mogeneous zone  of  calcified  material  usually  delimited  by  a 
distinct,  thin,  optically  dense  (matrix-rich)  layer.  This  layer  de- 
fines the  boundary  of  the  core.  In  some  cases,  careful  exami- 
nation of  the  core  may  reveal  diffuse,  very  faint,  or  extremely 


fine  growth  increments,  however,  they  are  easily  distinguished 
from  the  more  distinct  incremental  growth  pattern  distal  to  the 
core.  Taxonomically  related  differences  in  core  size,  shape  and 
number  generally  parallel  differences  in  the  primordia. 
The  external  morphology  of  larval  fish  otoliths  is  much  less 


54 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  6.    Occurrence  of  Multiple  Primordia  in  Fish  Otoliths  (see 
Text  for  Explanation). 

Order  Mormyriformes 
Mormyridae 

Order  Salmoniformes 
Esocidae 
Umbridae 

Salmonidae  (including  Coregoninae) 
Osmeridae 

Order  Cypriniformes 
Characidae 
Cyprinidae 
Catostomidae 

Order  Siluriformes 
Ictaluridae 
Bagridae 

Order  Atheriniformes 
Exocoetidae 
Oryziatidae 
Cyprinodontidae 
Belonidae 
Anablepidae 
Poeciliidae 
Atherinidae 

Order  Syngnathiformes 
Gasterosteidae 
Syngnathidae 

Order  Scorpaeniformes 
Cyclopteridae  (Cyclopterinae  and  Liparinae) 

Order  Gobiesociformes 
Gobiesocidae 

Order  Perci  formes 
Istiophoridae 
Stichaeidae 
Percichthyidae 

Order  Pleuronectiformes 
Pleuronectidae 


variable  than  seen  for  adults.  Similarity  between  taxa  is  greatest 
in  the  youngest  and  smallest  individuals,  in  which  the  otoliths, 
particularly  the  sagittae  and  lapilli,  tend  to  resemble  flattened 
spheroids  or  hemispheres.  Landmark  features  used  in  char- 
acterizing adult  otoliths  such  as  the  form  of  the  sulcus,  rostral 
projections,  cristae,  colliculi,  ostia  etc.  are  initially  not  evident 
or  weakly  developed  in  most  fishes.  Exceptions  to  this  gener- 
alization may  prove  to  be  useful  taxonomic  characters  (e.g.,  in 
various  istiophorids,  the  sulcus  acousticus  is  clearly  developed 
in  larvae  only  6  mm  SL).  Exaggerated  or  distinctive  morpho- 
logical features  of  adult  otoliths  of  some  taxa  may  also  begin  to 
develop  in  the  early  larval  stages.  For  example,  if  a  species  has 
a  markedly  elongate  sagitta,  such  as  found  in  some  callionymids 
or  fistulariids,  then  the  larval  otolith  may  show  a  tendency  for 
greater  growth  along  the  anterio-posterior  axis.  Unfortunately, 
such  early  evidence  for  adult  otolith  characters  is  often  not 
present,  particularly  for  the  many  species  which  show  an  abrupt 
change  in  otolith  growth  patterns  at  the  end  of  the  larval  phase. 
Nevertheless,  there  are  other  unique  or  distinctive  larval  otolith 
features  in  many  taxa,  and  they  are  potentially  valuable  for 
systematic  studies. 


Aside  from  shape,  there  are  at  least  two  other  "external" 
otolith  characters  which  may  be  used  for  taxonomic  work;  these 
involve  the  relative  sizes  and  times  of  formation  of  the  different 
otoliths;  the  sagittae,  lapilli,  and  asterisci.  In  certain  taxa,  such 
as  the  Ostariophysi,  the  sagitta  is  highly  modified  from  the 
typical  teleost  condition,  being  smaller  and  very  elongate;  and 
the  asteriscus  is  relatively  enlarged.  In  clupeids,  the  lapillus  is 
unusually  small  and  distinctively  shaped.  Differences  of  this  sort 
exist  to  a  lesser  degree  at  lower  taxonomic  levels  and  may  be 
used  in  larvae  for  distinguishing  groups.  The  time  of  appearance 
of  the  otoliths  in  development  is  also  a  variable  feature  offish 
ontogeny.  Many  or  perhaps  most  species  have  sagittae  and  lapilli 
at  hatching,  the  former  usually  noticeably  larger  even  at  this  stage. 
There  is  a  general  positive  relationship  between  egg  size,  time 
to  hatching  and  state  of  otolith  development  at  hatching.  Fishes 
with  very  large  eggs  and  corresponding  hatching  size  may  also 
have  the  asterisci  present  at  this  early  stage,  however  for  the 
majority  of  fishes,  these  otoliths  appear  later,  and  are  sometimes 
not  apparent  until  the  end  of  the  larval  stage.  The  asterisci  are 
distinctive  in  other  respects  as  well;  all  species  I  have  looked  at 
have  a  poorly  defined  core  with  multiple  primordia;  the  calcium 
carbonate  is  deposited  as  vaterite  (Lowenstam  and  Fitch,  1981) 
rather  than  the  aragonite  of  the  sagittae  and  lapilli;  and  there 
are  qualitative  differences  in  the  appearance  of  growth  incre- 
ments. 

Internal  structures  other  than  the  primordium  and  core  may 
also  have  direct  or  indirect  systematic  applications.  It  is  well 
documented  that  otoliths  grow  by  the  addition  of  layers  which 
are  deposited  on  a  diel  cycle  (see  earlier  references  on  larvae, 
plus  review  by  Pannella,  1 980;  also  Barkman,  1 978;  Wilson  and 
Larkin,  1980;  Steffensen,  1980;  Victor,  1982;  Victor  and  Broth- 
ers, 1982).  These  daily  growth  increments  are  usually  simple 
bipartite  structures  composed  of  one  protein-rich  and  one  pro- 
tein-poor calcareous  layer.  In  certain  situations  (especially  fast 
growth  and  large  otoliths)  subdaily  increments  (formed  over 
shorter  time  intervals)  of  similar  structure  may  also  be  present. 
The  timing  of  the  production  of  the  defining  boundary  of  the 
core,  which  also  corresponds  to  the  onset  of  incremental  growth 
around  the  core,  is  another  "internal"  character  that  varies  be- 
tween taxa.  Some  groups  start  incremental  growth  before  hatch- 
ing, others  at  hatching,  and  still  others  at  about  the  time  of  yolk 
absorption  and  the  onset  of  exogenous  feeding  (Brothers  et  al., 
1976;  Radtke  and  Waiwood,  1980;  Radlke  and  Dean,  1982; 
Radtke,  1984).  There  appear  to  be  clear  taxonomic  trends  in 
these  characters  which  are  also  related  to  other  trends  in  egg 
size  and  developmental  rate  and  pattern. 

Some  Examples  of  Taxonomically  Related 

Trends  in  Larval  Otolith  Form: 

External  Morphology 

The  development  of  the  general  form  of  the  adult  sagitta  is  a 
gradual  process  in  many  species,  whereas  in  others  there  may 
be  one  or  more  relatively  abrupt  changes  in  growth  form,  par- 
ticularly around  the  time  of  transformation  from  larva  to  ju- 
venile. This  change  involves  the  development  of  "secondary 
growth  centers"  which  first  appear  externally  as  angular  to 
rounded  protuberances  on  the  sagitta  surface  (Fig.  21;  internal 
structure  is  discussed  below).  The  result  of  the  expanding  growth 
around  these  centers  is  the  eventual  surrounding  of  a  discrete 
larval  otolith  and  the  stronger  development  of  form  and  surface 
characters  of  the  adult  sagitta.  In  examining  the  otoliths  of  over 


BROTHERS:  OTOLITH  STUDIES 


55 


IOh™ 


10h"> 


B 


Fig.  24.     Pnmordia  and  cores  of  goby  otoliths.  (A)  Sagilta  of  adult  sirajo  goby  {Sicydiuni  plumieri).  (B)  Sagitta  from  an  unidentified  goby  larva. 


100  families  of  fishes,  this  soil  of  sagittal  growth  pattern  appears 
to  be  characteristic  in  a  number  of  higher  level  taxa  (e.g.,  many, 
but  not  all,  perciform  families;  some  myctophids;  certain  but 
not  all  anguilloid  families,  pleuronectiform,  gadiform  and  scor- 
paeniform  fishes;  Percopsis,  and  others).  It  is  not  certain  whether 
the  presence  of  this  character  is  consistent  enough  to  be  used 
as  a  diagnostic  feature,  and  it  also  occurs  too  late  in  development 
to  be  of  use  in  larval  identification.  Lapilli  and  asterisci  tend  to 
show  more  gradual  changes  in  shape  and  growth  (Brothers  and 
McFarland,  1981)  and  I  have  not  observed  the  discontinuous 
pattern  described  above.  Lapilli  undergo  transitions  in  incre- 
mental patterns  at  about  the  same  time  that  the  sagitta  changes 
in  growth  form  (Brothers  and  McFarland,  1981;  Brothers,  un- 
published), however  these  are  not  obviously  evidenced  in  ex- 
ternal morphology  of  the  former. 

An  unusual  and  surprising  character  has  been  found  in  a 
preliminary  survey  of  several  of  the  "lower"  teleosts.  This  fea- 
ture, the  presence  of  otoconia  in  the  sacculus  and/or  utriculus 
in  addition  to  the  otoliths,  has  only  been  noted  for  non-teleos- 
tean  bony  fishes,  i.e.,  holosteans,  chondrosteans,  brachiopte- 
rygians,  dipnoans  (Carlstrom,  1963)  and  probably  Latimena 
(Brothers,  unpublished).  Osteichthyan  otoconia  or  statoconia 
are  numerous  (hundreds  to  thousands),  small  (from  a  few  to 
1 00  ^m)  calcareous  bodies  (vateritic,  sometimes  aragonitic)  which 
are  found  in  close  association  with  the  otolith  (Fig.  22).  They 
generally  have  a  very  characteristic  lens  shape,  although  some 
may  tend  towards  an  hexagonal  outline.  Internal  features  are 
variously  developed;  a  primordium-like  body  is  usually  present 
and  incremental  growth  is  seen  in  some.  Unexpectedly,  otoconia 
were  found  in  representatives  of  the  following  teleost  families: 
Albulidae,  Congridae,  Anguillidae,  Muraenidae.  Moringuidae, 
Notopteridae,  Osteoglossidae  and  Pantodontidae.  The  character 
appears  to  be  an  example  of  a  synplesiomorphy  shared  between 
non-teleostean  osteichthyans  and  two  teleostean  superorders, 
and  Osteoglossomorpha  and  the  Elopomorpha.  Not  all  species 
and  possibly  families  in  the  latter  two  groups  show  the  character, 
so  apparently  it  has  been  lost  independently  more  than  once. 
The  presence  of  otoconia  is  usually  not  apparent  until  the  early 
juvenile  stage,  they  are  not  seen  in  the  few  larvae  I've  had 
available,  however,  their  taxonomic  interest  warrants  mention 
here. 


Internal  Morphology 

There  are  a  number  of  taxonomically  related  trends  in  the 
size  and  shape  of  the  primordium  and  core  of  sagittae  and  lapilli. 
Table  6  lists  all  the  families  (of  1 13  sampled)  found  to  have 
representatives  with  multiple  or  clustered  primordia  (inclusion 
in  the  table  does  not  necessarily  indicate  that  all  family  members 
have  the  character).  In  some,  particularly  the  salmonids  and 
related  families,  the  primordia  are  clearly  separated  and  may 
each  be  surrounded  by  discrete  multiple  cores,  whereas  in  others, 
such  as  the  Atheriniformes  and  Gasterosteiformes,  the  multiple 
primordia  are  more  lightly  grouped  and  are  usually  surrounded 
by  a  single  core  (Fig.  23). 

Two  other  primordium  and  core  characters  have  been  found 
to  be  unique  to  certain  taxa.  In  the  gobies  and  related  families 
( 1  5  genera;  Gobiidae,  Microdesmidae,  Eleotridae,  and  Gobioid- 
idae)  all  species  invariably  have  an  elongate  primordium  in  the 
sagittae  and  lapilli  (usually  with  a  slight  central  constnction.  Fig. 
24)  which  has  not  been  seen  in  any  other  group.  Since  this  feature 
is  present  at  hatching,  it  allows  for  rapid  and  certain  identifi- 
cation of  these  speciose  families.  The  parrotfishes  (Scaridae,  4 
genera  examined)  appear  to  have  a  family-specific  early  growth 
pattern  in  the  sagitta  which  also  allows  for  the  identification  of 
very  young  larvae.  The  nearly  spherical  primordium  and  core 
grow  asymmetrically  for  about  the  first  5  days,  adding  new 
increments  in  a  restricted  area  on  the  distal  face  before  the 
growth  pattern  changes  to  one  producing  a  hemispherical  larval 
otolith.  The  result  of  this  pattern  (Fig.  25)  is  that  the  core  is 
clearly  on  a  different  focal  plane  from  a  section  normal  to  the 
majority  of  larval  growth  increments.  The  core  is  therefore 
asymmetrically  placed  nearer  to  the  proximal  or  internal  face 
of  the  sagitta.  This  feature  is  easily  observed  in  whole  larval 
otoliths  and  has  not  been  found  in  related  families  such  as  the 
labrids,  although  these  families  share  other  larval  otolith  char- 
acters. 

A  second  class  of  internal  features  has  obvious  external  man- 
ifestations described  above,  although  they  may  be  distinguished 
externally  for  only  a  discrete  period  in  development.  "Secondary 
growth  centers"  appear  in  optical  sections  or  SEM  views  as  foci 
for  increment  formation  removed  from  the  core  (Fig.  2 1 ).  Sp)ecies 
in  which  otoconia  occur  are  also  found  to  have  these  bodies 


56 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


IOh"" 


lOt"" 


Fig.  25.  Primordia  and  cores  of  parrotfish  sagittae.  (A)  Unidentified  scarid  larva,  medial  face  up,  core  in  focus.  The  dark  crescent  is  a  portion 
of  the  crista  on  the  surface.  (B)  Same  as  previous,  but  with  increments  in  focus.  (C)  Suspected  scarid  larva,  core  in  focus.  (D)  Same  as  (C), 
increments  in  focus. 


incorporated  into  the  otoliths.  The  mechanism  appears  to  be 
that  the  otoconia  adhering  to  the  otolith  surface  are  surrounded 
by  new  material  accreting  on  the  otolith,  and  eventually  these 
"included"  otoconia  are  found  deep  within  the  otoliths  of  larger 
fish.  In  some  species,  such  as  Anguilla  rostrata  otoconia  are 
found  in  dense  bands  corresponding  to  annual  zones.  "Includ- 
ed" otoconia  have  only  been  observed  in  juveniles  or  older 
individuals. 

Transitions  in  otolith  microstructure  involving  changes  in  the 
width  and  optical  density  of  growth  increments  (Fig.  26)  may 
be  related  to  a  variety  of  morphological  and  eco-behavioral 
changes  in  the  early  life  history  offish  (Pannella,  1 980;  Brothers, 
1981;  Brothers  and  McFarland,  1981;  and  numerous  other  pa- 
pers; also  related  works  by  Postuma,  1974,  and  McKem  et  al., 
1974).  Hatching,  yolk  absorption,  changes  in  feeding  and  hab- 
itat, postlarval  transformation,  and  settlement  can  all  poten- 
tially influence  the  deposition  pattern  of  daily  and  subdaily  growth 
increments.  To  the  extent  that  life  history  patterns  consistently 
diflfer  between  taxa,  we  may  expect  to  find  microstructural  evi- 
dence of  events  in  the  early  life  history  which  are  of  systematic 
value.  Difierences  between  taxa  will  then  be  expressed  as  dif- 
ferences in  the  timing  of  marks  (e.g.,  hatching)  and  otolith  tran- 


sitions and  in  their  intensity  and  duration.  Thus  we  may  use 
otoliths  to  record  ecological  information  which  may  then  be 
applied  to  systematic  studies.  An  even  simpler  approach  might 
just  be  a  quantitative  comparison  of  growth  rates  as  determined 
from  daily  growth  records  (once  validated,  and  the  fish  growth- 
otolith  growth  relationships  are  known),  however  care  should 
be  taken  to  avoid  problems  due  to  high  intraspecific  variability 
in  growth  rate  (e.g.,  Methot,  1981;  Bailey,  1982;  Brothers  et  al., 
1984).  Another  possibility  is  the  use  oflarval  life  duration  as  a 
taxonoinic  character.  There  is  evidence  to  both  species  speci- 
ficity and  very  limited  variability  in  some  taxa,  as  well  as  vari- 
ability or  flexibility  in  others  (Brothers  et  al.,  1983;  Thresher 
and  Brothers,  in  press;  Brothers  and  Thresher.  MS.;  Brothers 
and  Erdman,  unpublished),  so  caution  must  be  exercised  in 
using  this  character  as  a  taxonomic  tool. 

A  final  ecologically  related  application  is  the  determination 
of  spawning  time  (and  perhaps  place,  by  correction  for  current 
drift)  by  age  determination  of  larvae,  with  correction  for  the  lag 
between  fertilization  and  increment  initiation  (Townsend  and 
Graham.  1981;  McFarland  et  al.,  unpublished).  When  difl^er- 
ences  in  spawning  times  are  suspected  or  known  to  exist  for 
taxa,  then  larval  age  may  be  used  to  help  in  assigning  identifi- 


BROTHERS:  OTOLITH  STUDIES 


57 


IOh" 


B 

Fig.  26.     Transitions  in  otolith  microstructure  associated  with  settlement  and  transformation  from  the  larval  to  juvenile  stage.  (A)  Striped 
parrotfish  (Scarus  iserti)  sagitta.  (B)  Queen  angelfish  (Holacanthus  ciliaris)  sagitta. 


cation.  Under  the  best  of  circumstances,  when  spawning  is  rel- 
atively discrete  in  time,  differences  of  only  a  few  days  could 
potentially  be  resolved. 

The  last  area  in  which  otolith  studies  might  be  of  value  in 
systematic  studies  is  in  the  presentation  of  descriptive  papers 
on  fish  development.  Until  now  all  illustrations  and  descriptions 
of  development  of  wild  caught  larvae  were  related  to  body  size 
since  we  had  no  information  on  the  age  of  these  specimens.  We 
suspect,  and  in  some  cases  have  direct  knowledge  (cited  earlier) 
that  growth  rates  of  larvae  are  moderately  to  highly  variable, 
yet  we  have  no  data  on  the  relationship  between  age  and  growth 
rate  and  the  appearance  and  form  of  standard  characters  such 
as  pigment,  ossification,  meristics,  and  morphometries.  Perhaps 
some  of  the  variability  seen  in  size  specific  descriptive  accounts 
is  the  result  of  the  effects  of  different  growth  rates  on  the  char- 


acters. I  urge  that  we  should  make  an  extra  effort  to  determine 
the  age  of  wild-caught  larvae,  used  in  descriptive  studies  so  we 
may  be  able  to  establish  age  and/or  growth  rate  specific  accounts 
as  well  as  size  specific  ones.  Of  course  another  problem  with 
size  is  the  highly  variable  shrinkage  rates  caused  by  handling 
and  preservation.  Alternately  we  should  perform  laboratory  ex- 
periments to  examine  the  relationship  between  growth  rate  and 
developmental  rate.  In  this  way  we  may  be  able  to  understand 
some  of  the  underlying  causes  for  intraspecific  variation  in  larval 
fish  characters. 

Section  of  EcoLOCiv  and  Systematics,  Cornell  University, 
Ithaca,  New  York  14853.  Present  Address:  3  Sunset 
West,  Ithaca,  New  York  14850. 


Preservation  and  Curation 
R.  J.  Lavenberg,  G.  E.  McGowen  and  R.  E.  Woodsum 


THOSE  processes  by  which  we  fix  or  kill  living  tissues  without 
significantly  altering  their  gross  anatomy,  and  preserve  or 
maintain  these  tissues  on  a  long-term  basis  have  routinely  re- 
quired the  use  of  formalin  solutions  (Fink  et  al.,  MS;  Markle, 
1984).  This  certainly  is  the  case  for  fish  eggs  and  larvae.  The 
protocols  for  use  of  formalin  as  a  fixative  and  preservative  for 
ichthyoplankton  have  been  reviewed  and  standardized  in  sev- 
eral techniques  manuals  (Ahlstrom,  1976;  Castle,  1976;  Smith 
and  Richardson,  1977).  These  protocols  are  well  established  and 
it  is  not  our  intention  to  repeat  them  here.  Rather  we  wish  to 
elaborate  on  some  of  the  problems  associated  with  preservation 
and  curation,  and  to  propose  recommendations  to  resolve  those 
areas  of  real  or  potential  conflict. 


There  are  two  areas  of  special  concern  to  us  that  dictated  how 
our  investigations  proceeded.  First,  we  wish  to  ensure  that  em- 
bryonic pigment  is  retained  in  both  the  egg  and  larval  stages  in 
both  the  fixation  and  long-term  preservation  procedures.  Sec- 
ond, for  ontogenetic  stages  of  larvae  we  were  guided  by  a  concern 
for  protection  of  mineralized  structures,  guarding  particularly 
against  their  loss. 

Specimens  that  are  well-fixed  and  properly  preserved  are  im- 
portant not  only  to  ichthyoplanktologists  but  to  a  broad  spec- 
trum of  biologists,  fish  systematists,  and  museum  curators. 
Among  fixatives,  bufters  and  preservatives  there  is  no  unani- 
mous agreement  on  the  most  appropriate  ones.  The  problems 
that  plague  our  understanding  of  the  processes  associated  with 


58 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  27.  A  proposed  method  to  archive  the  early  life  history  stages  of  fishes.  In  the  left  foreground  is  a  series  of  three  vials,  the  first  contains 
the  specimens  and  preservation  fluid  and  is  capped  with  a  polyseal  closure.  This  first  vial  is  placed  into  the  second  with  the  documentation.  The 
third  vial  is  a  complete  unit.  As  evaporation  occurs  the  outer  vial  pops  free  of  its  plastic  closure,  indicating  that  the  vial  requires  curatorial 
attention.  The  vials  can  be  placed  together  in  commercially  available  paper  trays,  which  can  be  arranged  in  commercially  available  wooden  trays 
much  like  entomological  collections  are  maintained. 


these  chemicals  and  prevent  us  from  standardizing  a  protocol 
are  not  biological  ones  but  rather  those  of  chemistry. 

Fixatives.  —  Formahn  generally  is  accepted  as  the  most  appro- 
priate fixative.  However,  it  must  be  used  in  a  specific  concen- 
tration, polymerizes  with  age  and  with  contact  with  metals,  and 
is  a  poison.  Tucker  and  Chester  (in  press)  found  that  formalin 
used  with  salt  water  causes  significant  shrinkage,  whereas  an 
unbuffered  4%  solution  of  formalin  mixed  with  freshwater  caused 
the  least  amount  of  shrinkage  and  distortion  during  fixation. 
They  found  that  pigment  preserves  best  in  a  solution  of  un- 
buffered freshwater  formalin.  Although  the  pigment  holds  up 
well  in  this  solution,  the  skeleton  decalcifies  and  reduces  or  may 
even  prevent  staining  for  either  bone  or  cartilage  using  the  meth- 
ods of  Dingerkus  and  Uhler  (1977).  In  the  absence  of  a  suitable, 


inexpensive  substitute  we  recommend  that  formalin  be  used  for 
fixing  zooplankton  samples,  using  the  standard  ichthyoplankton 
protocols  described  by  Smith  and  Richardson  (1977).  This  pro- 
tocol could  be  modified  so  as  to  use  freshwater  rather  than 
seawater  in  preserving  the  sample  (Smith  and  Richardson,  1977: 
16-section  2.1.3.1)  so  as  to  reduce  shrinkage. 

Buffers— The  problems  associated  with  buffers  are  more  diffi- 
cult to  unravel.  Buffers  have  been  used  in  an  attempt  to  control 
fluctuating  pH  during  fixation  and  preservation.  Buffers  are 
needed  to  prevent  a  reduced  pH  in  either  the  fixative  or  pres- 
ervation solution  to  avoid  excessive  acidity  in  formalin  that 
may  decalcify  bone  (Taylor,  1 977).  However,  tissues  clear  when 
the  buffer  makes  the  solution  alkaline.  Taylor's  (1977)  data 
indicate  that  pH  can  fluctuate  only  in  a  narrow  range  without 


LAVENBERG  ET  AL.:  PRESERVATION  AND  CURATION 


59 


causing  some  degree  of  specimen  damage.  A  pH  of  less  than 
6.4  begins  the  process  of  decalcification,  mineral  loss  in  bone, 
whereas  a  pH  in  excess  of  7.0  initiates  the  clearing  process  that 
results  in  translucency. 

Tucker  and  Chester  (in  press)  recommend  that  sodium  borate 
not  be  used  as  a  buffer  on  the  basis  that  it  results  in  high  pH, 
i.e.,  loss  of  pigment  may  occur.  Calcium  carbonate  also  is  not 
recommended  because  it  tends  to  precipitate  out  of  solution  and 
onto  the  larvae.  Hexamine  should  not  be  used  at  all  because  it 
tends  to  clear  specimens  independent  of  pH,  and  to  damage 
them  (Steedman,  1976) 

Markle  (1984)  summarized  five  years  of  data  for  phosphate 
buffered  formalin  solutions  used  as  a  preservative.  He  used  the 
standard  ichthyoplankton  protocol  for  fixation  of  his  samples. 
He  gives  compelling  reasons  for  using  a  phosphate  buffer  to 
control  pH  of  formalin  solutions  used  as  a  preservative  for  fish 
larvae,  on  the  basis  that  the  amount  of  the  buffer  can  be  adjusted 
to  control  pH. 

A  review  of  the  ichthyoplankton  protocols  indicates  that  so- 
dium borate  (borax)  and  calcium  carbonate  (marble  chips)  are 
the  preferred  buffers,  although  Tucker  and  Chester  (in  press) 
recommend  sodium  acetate.  We  wish  to  stress  that  our  knowl- 
edge is  inadequate,  particularly  in  understanding  the  chemistry 
of  these  processes.  Clearly,  a  study  of  the  chemistry  of  fixation 
and  preservation  must  occur  before  a  recommendation  of  an 
acceptable  buffer  can  be  made.  However,  we  agree  with  Markle 
(1984)  that  phosphate  buffers  offer  the  best  alternative  to  borax 
and  marble  chips  for  long-term  preservation  on  the  basis  of 
their  versatility  in  adjusting  pH. 

Presenarives.  —  Afler  the  fixation  process  is  completed,  the  zoo- 
plankton  collections  are  processed  to  obtain  data  on  plankton 
volumes.  Then  the  samples  are  sorted  to  remove  the  ichthyo- 
plankton component,  the  eggs  and  larvae  of  fishes.  After  the 
identification,  enumeration,  and  measurements  offish  eggs  and 
larvae,  they  are  ready  for  long-term  archival  preservation. 
Through  this  process  the  collections  are  usually  maintained  in 
a  buffered  formalin  solution.  However,  Ahlstrom  (1976)  indi- 
cated that  if  an  investigator  was  sensitive  to  formalin  then  eth- 
anol  or  a  similar  preservative  was  acceptable. 

For  final  long-term  archival  preservation  Ahlstrom  (1976) 
indicated  that  fish  eggs  and  larvae  were  separately  vialed,  and 
placed  in  fresh  preservative.  This  fresh  preservative  was  a  one 
percent  buffered  formalin  solution  made  with  freshwater.  Ac- 
cording to  Ahlstrom  (1976)  the  larvae  remained  in  excellent 
condition  for  a  period  of  15-20  years.  Tucker  and  Chester  (in 
press)  recommend  a  long-term  preservative  consisting  of  a  4% 
formalin  solution  made  from  distilled  water  with  sodium  acetate 
used  as  a  buffer.  Whenever  formalin  is  used  as  the  basis  for  a 
long-term  preservation  fluid  for  fish  eggs  and  larvae  there  will 
be  problems  of  pH.  Phosphate  buffers  apparently  control  pH 
best  as  they  are  capable  of  maintaining  pH  within  a  narrow 
range  between  6.4  and  7.0.  Unfortunately  the  use  of  formalin 
as  a  final  preservative  has  the  potential  to  incur  considerable 
curatorial  expenses  just  to  monitor  pH  levels. 

We  recommend  that  70%  ethanol  be  used  as  the  final  pres- 
ervation fluid  on  the  basis  that  it  renders  the  pH  problem  moot, 
eliminates  working  with  the  fumes  of  formalin,  and  eliminates 
problems  associated  with  the  staining  process.  In  recommending 
ethanol  we  wish  to  reduce  or  eliminate  the  bufliering  problems 
and  their  associated  pH  problems  in  formalin  solutions.  After 
fixation,  the  concentration  of  formalin  can  be  reduced  to  a  1% 


solution,  then  this  fluid  can  be  drained  off  during  the  volume 
determination  process  and  replaced  with  ethanol.  It  is  important 
to  transfer  the  collections  directly  from  the  I  %  formalin  solution 
into  ethanol  without  washing  them  through  a  water  bath.  Thus 
a  small  concentration  of  formalin  fixative  will  be  retained  in 
the  ethanol  preservative.  Also,  the  transfer  should  be  a  staged 
one  through  a  series  of  ethanol  solutions,  from  1%  formalin  to 
20%  ethanol  to  45%  ethanol  to  70%  ethanol,  rather  than  a  direct 
transfer.  Zooplankton  collections  should  be  stored  in  the  dark, 
specifically  avoiding  light.  Also,  the  storage  facility  should  be 
as  cold  as  possible,  and  it  should  avoid  fluctuating  temperatures. 
In  summary,  we  recommend  that  formalin  be  the  fixative  of 
record  until  a  suitable  alternative  can  be  established.  Buffers 
should  be  investigated  to  determine  how  they  affect  long-term 
effects  of  fixation  and  preservation.  Phosphate  buflfered  formalin 
is  recommended  as  the  most  suitable  one  to  control  pH  within 
a  narrow  range  to  prevent  melanistic  pigment  loss  and  deminer- 
alization.  We  recommend  that  ethanol  replace  formalin  as  a 
preservative  fluid.  Finally,  the  chemistry  of  fixation  and  pres- 
ervation should  be  addressed  by  a  chemist  to  establish  a  suitable 
protocol  for  processing  zooplankton  samples. 

Curation.— The  chief  problems  with  storage  and  curation  of 
larval  fish  collections  are  to  prevent  fluid  loss,  stabilize  collec- 
tions, and  to  allow  for  retrieval  availability. 

Fluid  losses  through  evaporation  in  small  containers,  such  as 
vials,  can  be  disastrous.  There  are  means  to  reduce  evaporation. 
We  propose  that  a  double  vialing  procedure  be  established  (Fig. 
27).  First,  evaporation  may  be  significantly  reduced,  and  second, 
a  double  vialing  system  provides  a  mechanism  to  eliminate 
abrasion  and  damage  to  fish  eggs  and  larvae.  The  procedure 
calls  for  an  inner  vial  containing  the  specimens  and  preservation 
fluid  sealed  with  a  poly-seal  closure.  This  vial  is  inserted  into 
another  glass  vial,  which  leaves  sufficient  space  for  labels  and 
specimen  documentation.  The  second  vial  is  sealed  with  a  plas- 
tic closure.  The  outer  vial  is  placed  upside  down  over  the  inner 
one.  The  procedure  here  is  to  allow  gravity  to  work  on  vapor 
evaporating  from  the  inner  vial  in  such  a  manner  that  it  must 
be  compressed  before  escaping  from  the  outer  vial.  Essentially 
an  equilibrium  would  be  achieved  that  would  act  to  prevent 
further  evaporation.  In  addition,  a  means  for  specimen  docu- 
mentation can  be  achieved  that  allows  for  maximizing  these 
data  for  curation  without  causing  abrasion  or  damage  to  the 
delicate  specimens. 

Another  important  aspect  of  this  curation  technique  would 
be  its  contribution  to  retrieval  availability.  The  vials  can  be 
integrated  into  an  existing  ichthyological  system  so  as  to  make 
them  immediately  available  to  researchers  while  offering  to 
maximize  long-term  archival  preservation  protection. 

We  would  like  to  thank  all  of  our  colleagues  who  provided 
us  with  information  relative  to  the  fixation,  preservation  and 
curation  of  the  early  life  history  stages  of  fishes. 

On  behalf  of  the  steering  committee  of  the  Ahlstrom  Sym- 
posium we  would  like  to  recommend  that  the  National  Museum 
of  Natural  History  in  Washington,  D.C.,  the  Museum  of  Com- 
parative Zoology  (Harvard  University),  and  the  Natural  History 
Museum  of  Los  Angeles  County  in  Los  Angeles  be  considered 
for  the  deposition  of  the  early  life  history  stages  of  fishes  for 
long-term  archival  care. 

Section  of  Fishes,  Natural  History  Museum  of  Los  Ange- 
les County,  900  Exposition  Boulevard,  Los  Angeles, 
California  90007. 


DEVELOPMENT  AND  RELATIONSHIPS 

Elopiformes:  Development 
W.  J.  Richards 


THE  Elopiformes  comprises  four  genera  of  recent  fishes  and 
each  of  these  genera  is  composed  of  at  least  two  species. 
The  species  are  found  in  tropical  waters  of  the  Atlantic,  Indian 
and  Pacific  oceans.  Elops,  a  cosmopolitan  genus,  is  composed 
of  several  species  and  Megalops  is  composed  of  two  species.  M. 
atlantica  Valenciennes  is  found  in  both  the  eastern  and  western 
Atlantic  and  M.  cyprinoides  (Broussonet)  is  found  in  the  Indian 
and  western  Pacific  Oceans.  Alhula  has  two  recognized  species. 
A.  vulpes  is  cosmopolitan  and  A.  nemoptera  is  found  on  the 
Atlantic  and  Pacific  coasts  of  the  Americas.  Recent  electropho- 
retic  work  indicates  that  there  may  be  additional  species  (Shak- 
lee  and  Tamaru,  1981).  Pterothnssus  has  one  species  along  the 
coast  of  West  Africa,  P.  helloci  Cadenat,  and  one  off  Japan,  P. 
gissu  Hilgendorf 

Larval  stages  of  elopiform  fishes  have  attracted  great  interest 
among  ichthyologists  because  of  their  unusual  leptocephalus 
development,  a  stage  found  in  no  other  group  but  the  Anguil- 
liformes  and  Notacanthiformes.  Consequently  most  recent  clas- 
sifications have  combined  all  fish  with  leptocephalus  larvae 
into  the  Elopomorpha  (Patterson  and  Rosen,  1977).  Forked  tails 
of  the  elopiform  leptocephali  provide  an  easy  means  of  sepa- 
rating them  from  other  leptocephali  which  have  reduced  or  no 
tails  at  all.  The  non-fork  tailed  leptocephali  are  treated  sepa- 
rately in  the  three  subsequent  papers  in  this  volume. 

Recent  classifications  have  altered  our  classical  view  of  elo- 
piform fishes  by  suggesting  a  much  closer  relationship  with  eels. 
Greenwood  et  al.  (1966)  included  all  fishes  with  leptocephalus 
larvae  in  the  superorder  (Elopomorpha).  This  superorder  con- 
tained: Elopiformes  with  two  suborders,  the  Elopoidei  (Elopidae 
and  Megalopidae)  and  the  Albuloidei  ( Albulidae  including  Pter- 
othrissidae);  Anguilliformes  with  two  suborders,  the  Anguil- 
loidei  and  Saccopharyngoidei;  and  Notacanthiformes  with  two 
families  (Notacanthidae  and  Halosauridae).  A  number  of  papers 
have  discussed  this  proposed  classification  and  a  majority  has 
sustained  the  opinion  that  the  Elopomorpha  is  a  monophyletic 
assemblage.  Forey  (1973a)  discussed  the  intragroup  relation- 
ships and  made  some  interesting  observations  on  leptocephali 
in  a  second  paper  (1973b).  Two  significant  classifications  ap- 
peared in  1977,  one  by  Greenwood  and  one  by  Patterson  and 
Rosen.  Both  classifications  concluded  that  Elopomorpha  is  a 
natural,  monophyletic  group  and  that  Albula  and  Pterothrissus 
are  related  to  the  Halosauridae  and  Notacanthidae.  Greenwood 
(1977)  presented  a  concept  of  Elopomorpha  as  a  Cohort  Tae- 
niopaedia  with  two  superorders:  Elopomorpha  comprised  of 
Elops  and  Megalops  in  the  Order  Elopiformes  (Suborder  Elo- 
poidei) and  Anguillomorpha  comprised  of  two  orders,  the  Al- 
buliformes  with  two  suborders  (Albuloidei  and  Halosauroidei) 
and  the  Anguilliformes.  Patterson  and  Rosen  (1977)  defined  a 
cohort  Elopomorpha  of  three  orders:  Elopiformes,  Megalopi- 
formes  and  Anguilliformes,  the  latter  with  two  suborders— the 
Anguilloidei  and  Albuloidei.  Patterson  and  Rosen  (1977)  con- 


cluded that  the  interrelationships  of  the  Elopidae,  Megalopidae 
and  Anguilliformes  are  best  represented  by  an  unresolved  tri- 
chotomy. However,  it  would  seem  that  those  with  forked  tails 
would  be  monophyletic  and  the  reduced  or  tailless  leptocephali 
would  be  derived  from  those  with  tails.  The  trichotomy  scheme 
results  in  paraphyletic  forked  tailed  forms. 

With  the  exception  of  the  species  of  Pterothrissus.  the  species 
of  the  remaining  genera  are  coastal  with  some  stages  entering 
hyposaline  environments.  Pterothrissus  helloci  occurs  benthi- 
cally  from  70  to  500  m,  most  abundantly  from  120  to  250  m, 
off  the  coast  of  West  Africa  from  9°N  latitude  to  20°S  latitude 
(Poll,  1953).  All  elopiforms  are  presumed  to  have  pelagic  eggs 
although  the  eggs  of  all  are  undescribed.  According  to  Smith 
and  Potthoff  (1975)  the  eggs  and  early  larvae  of  Harengula 
jaguana  were  erroneously  attributed  to  Megalops  atlanticus  by 
Breder  (1944),  Mansueti  and  Hardy  (1967),  and  Mercado  and 
Ciardelh  (1972). 

The  larval  stages  have  been  well  described  for  all  genera  and 
are  unique  (Fig.  28).  The  larval  stage  is  represented  by  the  lep- 
tocephalus which  has  been  defined  by  Hulet  (1978)  and  Smith 
(1979).  The  leptocephalus  is  compressed,  transparent  and  leaf- 
like with  a  mucinous  pouch  which  distinguishes  it  from  all  other 
fish  larvae.  It  grows  to  large  size  compared  to  other  fish  larvae, 
it  has  fang-like  teeth  at  the  early  stages  which  are  subsequently 
lost  (possibly  reabsorbed),  its  viscera  is  confined  to  a  narrow 
strand  along  the  ventral  midline,  its  musculature  forms  a  thin 
layer  outside  of  the  mucmous  pouch  and  the  remainder  of  the 
pouch  consists  of  a  mass  of  acellular  material  composed  of 
mucoproteins  and  polysaccharides  enclosed  by  a  continuous 
layer  of  epithelial  cells.  Its  gut  is  in  two  sections,  an  esophagus 
and  an  intestine  which  are  separated  by  a  gastric  region  com- 
posed of  the  stomach,  liver  and  gallbladder.  The  kidney,  of 
various  lengths,  lies  over  the  gut  beginning  near  the  gastric  region 
and  contmuing  posteriorly.  Ventral  blood  vessels  conspicuously 
appear  between  the  aorta  and  the  kidney  and  gut.  In  elopiform 
leptocephali  dorsal,  anal,  pectoral  and  pelvic  fins  are  present 
and  the  caudal  fin  is  large  and  forked. 

Genera  of  elopiform  leptocephali  are  easily  identified  except 
at  small  sizes  prior  to  caudal  development  when  myomeres  are 
difficult  to  count.  The  number  of  myomeres  for  elopiforms  ranges 
from  51  to  92  whereas  most  anguilliform  leptocephali  have 
more  than  95.  Leptocephali  of  the  Cyemidae  have  80  myomeres. 
Smith  ( 1 979)  provides  a  key,  characterizations  and  illustrations 
of  the  genera.  Many  other  workers  have  described  complete 
series  or  individual  stages.  Complete  series  of  Elops  have  been 
described  by  Gehringer  (1959a),  Megalops  by  Wade  (1962), 
Alhula  by  Alexander  (1961),  and  Pterothrissus  by  Matsubara 
(1942).  Among  other  papers  which  describe  and  illustrate  var- 
ious stages  are:  oi  Megalops  by  Delsman  (1926b),  Mercado  and 
Ciardelli  ( 1 972),  Gehringer  ( 1 959b),  Eldred  ( 1 967b,  1 972)  and 
Richards  (1969);  of  Pterothrissus  by  Smith  (1966b)  and  Rich- 


60 


RICHARDS:  ELOPIFORMES 


61 


rv'  ve-  \vN\v  V  V      -^ 


z^-^^^.. 


Fig.  28.  Elopiform  leptocephali.  Top  to  bottom:  Elops  sp.,  33.8  mm  SL,  Luanda,  Angola  (redrawn  from  Richards,  1 969);  Megalops  allanticus. 
22.8  mm  SL.  Luanda,  Angola  (redrawn  from  Richards,  1969):  Plerolhnssus  belloci.  123.9  mm  SL,  off  Angola  (redrawn  from  Richards,  1969); 
and  Albula  vulpes,  64.2  mm  (redrawn  from  Alexander,  1961). 


ards  (1969);  of  Elops  by  Hildebrand  (1963a),  Eldred  and  Lyons 
(1966),  Gomez  Caspar  (1981),  Richards  (1969);  and  of  Albula 
by  Eldred  ( 1 967a),  Poll  (1953),  Gomez  Gaspar  ( 1 98 1 )  and  Hil- 
debrand (1963b).  The  Albula  leptocephali  heads  illustrated  by 
Meyer-Rochow  (1974)  may  be  incorrect. 


The  characters  used  for  distinguishing  the  families  and  genera 
(following  Smith,  1979)  are  as  follows:  Albula  and  Pterothhssus 
leptocephali  have  the  origin  of  the  anal  fin  well  behind  the  dorsal 
fin  by  a  distance  exceeding  the  length  of  the  anal  fin  base  whereas 
Elops  and  Megalops  have  the  origin  under  the  dorsal  fin  or  close 


Table  7.    Meristic  Characters  for  Selected  Elopiform  Leptoc  ephali. 


Taxon 

Source 

Dorsal  rays 

Number  of  anal  rays 

Myomeres 

Elops 
saurus 
spp. 

Gehringer  (1959a) 
Richards  (1969) 

21-26  usually  22-24 
20 

12-15  usually  13-14 

15-17 

78-82  usually  79-80 
70-73 

Megalops 
allantica 
cyprinoides 

Wade  (1962) 
Wade  (1962) 

9-13  usually  12 
10-17  usually  12-17 

16-22  usually  19-21 
18-25  usually  23-25 

51-57 

59-68  usually  62-67 

Alhula 

vulpes 
nemoptera 

Alexander  (1961) 
Rivas(1967) 

16 

7 

65-70  usually  67-68 
69-74 

Pterolhrissus 

belloci 

Richards  (1969) 

51-56 

10-13 

85-92 

62 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


behind  it,  by  a  distance  not  exceeding  the  length  of  the  anal  fin 
base.  Flops  and  Mega/ops  leptocephali  have  lateral  pigment  but 
Albula  and  Pterothrissus  leptocephali  do  not  have  lateral  pig- 
ment. Elops  is  distinguished  from  Mega/ops  by  having  a  de- 
pressed head,  more  dorsal  than  anal  rays  and  the  origin  of  the 
anal  fin  is  under  the  posterior  end  of  the  dorsal  fin  or  slightly 
behind  it.  Megalops  does  not  have  a  depressed  head,  has  fevk'er 
dorsal  rays  than  anal  rays  and  the  origin  of  the  anal  fin  is  under 
the  middle  of  the  dorsal  fin.  Albula  leptocephali  are  separable 
from  Pterothrissus  leptocephali  by  the  distance  between  the  pos- 
terior edge  of  the  dorsal  fin  and  the  origin  of  the  anal  fin.  In 


Albula  this  distance  is  about  2.5  times  the  length  of  the  dorsal 
fin  base  and  in  Pterothrissus  this  distance  is  about  6-7  times 
the  length  of  the  dorsal  fin  base.  Also  the  snout  is  short  in  Albula 
and  prolonged  in  Pterothrissus.  Within  genera,  meristic  char- 
acters are  useful  in  identification  of  the  species  (Table  7). 

The  interrelationships  of  the  elopiform  fishes  are  discussed 
by  Smith  in  a  subsequent  paper  in  this  volume. 


National  Marine  Fisheries  Service,  Southeast  Fisheries 
Center,  75  Virginia  Beach  Drive,  Miami,  Florida  33149. 


Notacanthiformes  and  Anguilliformes:  Development 
P.  H.  J.  Castle 


THE  Notacanthiformes  (spiny  eels)  and  Anguilliformes  (true 
eels)  were  united  with  the  Elopiformes  (tenpounders,  tar- 
pons, bonefishes)  by  Greenwood  et  al.  (1966)  as  the  superorder 
Elopomorpha.  These  authors  noted  that  members  of  the  three 
orders  share  osteological  similarities,  swim  bladder  not  con- 
nected with  ear  (except  for  Megalops),  and  a  distinctive  larval 
phase  (leptocephalus).  More  recent  authors  (Nelson,  1973;  Fo- 
rey,  1973b;  Patterson  and  Rosen,  1977)  recognised  this  rela- 
tionship, though  not  precisely  in  this  form.  There  seems  little 
doubt  that  they  are  indeed  closely  related,  but  in  being  exclu- 
sively elongate  fishes  the  notacanths  and  eels  are  readily  distin- 
guished externally  from  the  short-bodied,  herring-like  Elopi- 
formes. 

NOTACANTI  form  ES 

McDowell  (1973)  reviewed  the  notacanths,  a  morphologically 
discrete  group  of  fishes,  found  on  or  near  the  bottom  on  the 
deeper  continental  slope  into  the  deep  sea,  recognising  2  sub- 
orders, 3  families,  6  genera  and  22  extant  species  (Table  8).  He 
chose  to  give  subordinal  distinction  to  the  Halosauridae  on  the 
one  hand,  and  the  Notacanthidae  and  Lipogenyidae  jointly  on 
the  other,  although  Marshall  (1962)  had  already  demonstrated 
major  structural  similarities  between  these  families. 

The  Notacanthiformes  have  in  common  with  the  Anguilli- 
formes a  leptocephalus  phase,  an  elongate  body  form,  the  as- 
sociated lengthening  of  the  anal  fin,  and  a  reduced  caudal  fin. 
Members  of  the  two  orders  are  otherwise  dissimilar.  Notacanths 
have  well  developed  pelvic  fins;  a  compact,  dorsal  fin  with  spines 
in  some  species;  scales  present  and  prominent  in  some;  and  a 
large  gill  opening  and  opercular  flap.  Eels  lack  pelvic  fins;  the 
dorsal,  unless  secondarily  reduced  or  lost,  is  always  long  and  is 
supported  by  delicate  rays;  scales,  if  present,  are  greatly  reduced; 
and  the  gill  opening  and  its  supporting  structures  are  also  re- 
duced. Furthermore,  notacanth  leptocephali  are  as  distinctive 
from  those  of  the  true  eels  as  are  their  adults  (Fig.  29).  They  are 
greatly  elongate  (up  to  180  cm),  having  a  thin  post-caudal  fil- 
ament in  place  of  a  normal  caudal  fin;  dorsal  and  pelvic  fins  are 
represented  by  compact,  short-based  structures  present  at  some 
stage  of  larval  growth;  there  is  a  minute  pectoral,  straight  gut, 
subterminal  anus  and  the  myomeres  are  V-shaped,  not  W-shaped; 


pigment  occurs  in  a  ventral  series  and  (rarely)  below  the  mid- 
lateral  level. 

Several  quite  different  notacanth  leptocephali  of  this  type  are 
known,  some  almost  certainly  halosaurids  ( Tiluropsis.  Lepto- 
cephalus attcnuatus),  some  possibly  notacanthids  (Tilurus)  and 
others  of  unknown  identity  (Leptocephalus  giganteus).  Eggs  and 
early  larvae  have  not  yet  been  identified  and  information  on 
vertebral  numbers  is  mostly  lacking  for  the  group.  Until  con- 
firmed identifications  have  been  made  and  more  information 
is  forthcoming  from  leptocephali,  ontogeny  is  unlikely  to  con- 
tribute further  to  the  little  that  is  known  of  relationships  in  this 
order. 

Anguilliformes 

The  Anguilliformes  make  up  a  much  larger  and  more  diverse 
assemblage.  I  recognize  21  families.  153  genera  and  720  species 
for  the  group  (Table  9). 

Within  the  Anguilliformes  itself  Bohike  (1966)  reviewed  the 


Table  8.    Composition,  Distribution  and  Habitat  of  the  Nota- 
canthiformes. +  =  All  or  most  species;  (  +  )  =  some  species  only. 


Halo- 

saundae 

Nola- 
canlhidae 

Lipo- 
genyidae 

Taxonomic  components: 

Known  genera  (adults) 
Known  genera  (larvae) 
Known  species  (adults) 

3 
?1 

13 

2 

?l 

8 

1 

0 

1 

Distribution: 

Atlantic:  Genera 
Species 

3 
7 

2 
3 

1 
1 

E.  Pacific:  Genera 
Species 

1 
2 

1 

1 

0 
0 

I.-W.  Pacific:  Genera 
Species 

2 
5 

2 
4 

0 
0 

Habitat  (species): 

Shelf 
Slope 
Abyssal 

(+) 
(+) 

(  +  ) 
(  +  ) 
(  +  ) 

+ 

CASTLE:  NOTACANTHIFORMES,  ANGUILLIFORMES 


63 


Leptocephalus  giganteus   390mm  TL 


'Tilurus" 


"Tiluropsis' 


Fig.  29.     The  three  major  forms  of  notacanth  leptocephali  showing  in  upper  two  the  elongate  snout,  distinct  dorsal  (arrow),  and  ventral 
melanophore  series;  in  lower  left  the  myoseptal  pigment;  and  in  lower  right  the  oval  eye. 


superfamily  Saccopharyngoidea  (gulpers),  a  small  group  of  3 
families,  4  genera  and  8  species  of  highly  modified  mid-water, 
oceanic  eels,  unmistakeable  in  body  form  and  possessing  a  lep- 
tocephalus of  distinctive  type.  Although  they  are  currently  ac- 
cepted to  be  true  eels,  they  are  so  highly  aberrant  in  form  and 
osteology  that  a  case  could  be  made  for  their  retention  in  a 
separate  suborder,  as  indeed  was  proposed  by  Greenwood  et  al. 
(1966).  Other  eel  families  have  been  studied  in  some  detail, 
notably  the  Congridae  (Smith.  1971),  Synaphobranchoidea 
(Robins  and  Robins,  1976),  Ophichthidae  (McCosker,  1977), 
Nemichthyidae  (Nielsen  and  Smith,  1978)  and  others,  but  there 
are  several  major  gaps  and  the  order  has  never  been  compre- 
hensively reviewed. 

With  some  exceptions,  the  families  and  genera  of  eels  occur 
worldwide  (Table  9)  while  eel  species  have  a  more  restricted 
distribution  in  one  or  other  of  the  major  oceans.  Some  meso- 
pelagic,  slope/abyssal  species  and  just  a  few  shelf  species  are 


known  from  both  Indo-west  Pacific  and  Atlantic.  As  for  many 
other  teleosts.  the  Indo-west  Pacific  is  richest  in  genera  and 
species,  despite  relatively  limited  collecting  there,  and  infor- 
mation is  scattered  (Alcock.  1889  e/.yf(7!/.:  Fowler,  1934;Asano, 
1962;  Karrer,  1982).  The  eel  fauna  of  the  Atlantic  is  rather  better 
known  (Blache,  1977;  Bohlke,  1978)  but  by  comparison  the 
group  is  rather  poorly  represented  in  the  East  Pacific. 

Characters.— The  families  and  genera  of  Anguilliformes  are  dis- 
tinguished principally  by  external  characters,  including  mor- 
phometries (Table  10)  but  the  limits  are  not  yet  firmly  estab- 
lished for  all  families  in  the  order.  Osteological  characters,  which 
mostly  reflect  these  external  modifications  are  also  of  value  at 
family  and  generic  levels  (Table  1 1 )  but  are  inadequately  known, 
especially  in  the  Congridae  and  related  families,  and  the  Mu- 
raenidae.  Too  few  genera  have  been  identified  in  their  larval 
form  for  ontogenetic  characters  to  have  been  used  extensively 


64 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  9.    Composition,  Distribution  and  Habitat  of  ihe  Anguilliformes. 


All  or  most  species;  (  +  )  =  some  species  only. 


Synapho-  Ophich-  Netla-  Dench- 

branchi-      Dysom-      Simcn-  thi-  Con-         Muraenc-      stomali-         Colo-  thyi-  Semvo-        Anguil- 

dae         matidae    chelyidae         dae  gndae  socidae  dae  congndae  dae  mendae  lidae 


Helcr- 
Monn-       enchelyi- 
guidae  dae 


Taxonomic  components: 
Known  genera  (adults) 
Known  genera  (larvae) 
Known  species  (adults) 

Distribution: 
Atlantic:  Genera 
Species 
E.  Pacific:  Genera 
Species 
I.-W.  Pacific:  Genera 
Species 

Habitat  (species): 
Freshwater 
Shelf:  Tropical 

Temperate 
Slope/abyssal 
Pelagic 


3 

9 

1 

55 

28 

9 

6 

1 

2 

3 

1 

2 

2 

1 

2 

0 

25 

15 

4 

5 

1 

2 

3 

1 

2 

2 

7 

16 

1 

250 

131 

16 

32 

4 

3 

12 

15 

13 

8 

3 

5 

1 

29 

17 

2 

6 

1 

2 

2 

1 

2 

2 

7 

6 

1 

73 

32 

5 

13 

2 

3 

6 

2 

2 

7 

0 

1 

0 

17 

10 

2 

3 

0 

1 

1 

0 

1 

1 

0 

2 

0 

39 

12 

2 

3 

0 

1 

1 

0 

1 

1 

4 

6 

1 

35 

24 

7 

6 

1 

2 

2 

1 

1 

0 

8 

8 

1 

137 

(  +  ) 

+ 

(  +  ) 

63 

+ 
(  +  ) 

9 

+ 

8 

(  +  ) 
(  +  ) 

2 

3 

6 

13 

+ 

10 

(  +  ) 

+ 

0 

(+) 

+ 

-1- 

-1- 

+ 

(  +  ) 
(  +  ) 

(  +  ) 

(  +  ) 

-1- 

-1- 

(  +  ) 
(  +  ) 

-H 

in  determining  relationships.  Eel  species  are  principally  distin- 
guished externally,  by  teeth  and  cephalic  pore  patterns  and  by 
meristics,  especially  the  number  of  vertebrae.  The  latter  reflects 
the  number  of  myomeres  in  the  leplocephali. 

Many  of  the  adult  characters  by  which  the  families  and  genera 
differ  from  one  another  appear  to  be  correlated  with  the  extent 
to  which  the  rather  sedentary  mode  of  life  associated  with  bur- 
rowing, crevice-dwelling  or  pelagic  habits  has  been  elaborated 
throughout  the  group.  In  most  families  of  eels  there  are  species 
in  which  the  body  is  very  slender,  with  vertebrae  numbering 
180  or  more  (Table  10).  The  pectoral  fins  are  reduced  or  lost 
variously  in  families  (Muraenidae,  Heterenchelyidae),  genera 
(Ophichthidae,  Xenocongridae),  or  even  within  the  life  span  of 


individuals  {Moringita).  The  median  fins  may  also  be  reduced 
to  vestiges  either  in  height  or  in  length  by  a  posteriorwards  shift 
of  their  origin,  or  they  may  be  entirely  lost,  though  pterygio- 
phores  can  be  retained.  Scales  occur  only  in  some  of  the  syna- 
phobranchoids  and  in  the  Anguillidae. 

Other  characters  are  not  so  clearly  associated  with  the  adop- 
tion of  fossorial,  cryptic  or  pelagic  habits.  These  include  the 
ventral  displacement  of  the  gill  openings  (the  extreme  devel- 
opment being  in  some  Synaphobranchidae  and  a  few  Ophich- 
thidae  where  they  are  confluent  ventrally);  the  ventral  displace- 
ment of  the  posterior  nostril  (most  Ophichthidae,  Xenocongri- 
dae, to  some  extent  the  Synaphobranchidae)  so  that  it  may  even 
open  within  the  mouth;  or  its  dorsal  displacement  (Muraenidae), 


Table  10.    Some  Morphological  Characters  of  the  Anguilliformes.  +  =  All  or  most  species;  (  +  )  =  some  species  only;  *  =  presumed 

primitive  condition. 


Synapho- 

Dysom- 

Simen- 

Ophichlhi- 

Con- 

Muraenc- 

Nella- 

Colocon- 

Dcr- 

branchidae 

matidae 

chelyidae 

dat- 

gndae 

socidac 

stomatidae 

gndac 

ichthyidae 

Vertebrae:  Min.* 

126 

107 

121 

110 

105 

120 

186 

148 

126 

Max. 

172 

204 

125 

270 

225 

261 

290 

163 

159 

Scales:  Present* 

-t- 

(  +  ) 

-h 

Absent 

(  +  ) 

-t- 

-1- 

-1- 

-1- 

-1- 

-1- 

+ 

Pectoral:   Present* 

+ 

+ 

+ 

(  +  ) 

-t- 

-1- 

Reduced 

{  +  ) 

{  +  ) 

(+) 

(  +  ) 

(  +  ) 

-1- 

-1- 

Absent 

(  +  ) 

(+) 

(  +  ) 

(  +  ) 

+ 

Caudal:  Present* 

-1- 

+ 

+ 

-1- 

+ 

+ 

+ 

H- 

Reduced 

(+) 

(+) 

Absent 

-1- 

Dorsal  origin: 

Over  pectoral/gill  opening* 

+ 

+ 

H- 

+ 

-1- 

+ 

+ 

+ 

Between  pectoral  and  anus 

(+) 

■f 

Over  or  behind  anus 

(+) 

Gill  openings:   Lateral* 

+ 

+ 

+ 

+ 

-1- 

Displaced  ventrally 

-1- 

+ 

+ 

(+) 

+ 

Posterior  nostril:   Before  eye* 

+ 

+ 

+ 

+ 

+ 

+ 

Displaced  dorsally 

(+) 

(+) 

Displaced  ventrally 

+ 

+ 

-1- 

-1- 

(+) 

(+) 

Lateral  line:  Complete* 

+ 

+ 

+ 

+ 

+ 

-1- 

+ 

+ 

Incomplete 

+ 

CASTLE:  NOTACANTHIFORMES,  ANGUILLIFORMES 


65 


Table  9.    Extended. 


Murae- 
nidae 

Myro- 

congn- 

dae 

Xeno- 

congn- 

dac 

Nem- 

ichlhyi- 

dae 

Cyema- 
Iidae 

Sacco- 

pharyngi- 

dae 

Eury- 

pharyngi- 

dac 

Mono- 

gnathi- 

dae 

13 

1 

8 

3 

2 

1 

1 

2 

7 

0 

6 

2 

1 

1 

1 

1 

170 

1 

15 

8 

2 

3 

1 

6 

8 

1 

6 

6 

2 

1 

1 

2 

30 

1 

7 

7 

2 

3 

1 

2 

6 

0 

2 

3 

1 

0 

1 

1 

12 

0 

2 

6 

1 

0 

1 

2 

13 

0 

7 

3 

1 

0 

1 

2 

120 

0 

8 

7 

1 

0 

1 

3 

(+) 

+ 

+ 

+ 

(  +  ) 

+ 

+ 

+ 

+ 

+ 

or  both  (the  genera  of  Nettastomatidae).  There  may  be  a  re- 
duction of  the  lateral  line  (Muraenidae.  Xenocongridae)  or,  con- 
versely, its  great  elaboration  (the  congrid  Scalanago).  In  some 
eels  there  is  an  enlargement  of  the  mouth  and  teeth-bearing 
surfaces,  either  by  a  forward  prolongation  of  the  premaxillary- 
ethmovomer  and  dentary  (Nemichthyidae,  Cyematidae  and 
others),  or  by  the  turning  backwards  of  the  suspensorium  with 
a  coincident  reduction  or  loss  of  the  palatopterygoid  arch 
(Ophichthidae,  Muraenidae). 

In  all  eels  the  branchial  region  is  elongate,  the  pectoral  girdle 
is  separated  from  the  skull  and  the  posttemporal  is  lost.  This 
lengthening  is  accompanied  by  a  reduction  of  the  opercular 
series,  narrowing  of  the  gill  opening  and  increased  importance 


of  branchial  pump  respiration.  The  branchial  series  is  displaced 
backwards  with  enlargement  of  the  4th  arch  as  pharyngeal  jaws, 
especially  in  the  Muraenidae.  The  long  branchial  wall  is  sup- 
ported by  an  increased  number  of  branchiostegal  rays  which 
curve  up  around  the  branchial  region  and  expand  distally.  In 
the  ophichthids  the  throat  is  further  supported  by  numerous 
accessory  branchiostegal  rays  (Parr's  "jugostegalia")  which  are 
not  attached  to  the  hyoid  arch  and  overlap  in  the  ventral  mid- 
line. 

Overall,  there  is  clearly  a  strong  functional  correlation  be- 
tween the  lengthening,  narrowing  and  smoothing  out  of  the  body 
outline,  the  increase  in  body  flexibility  and  modifications  in 
nostrils,  jaws,  gill  openings  and  lateral  line  with  the  mode  of 
life  which  is  a  feature  of  the  eels  as  a  group. 

Eggs.—T\\e  best  known  stages  in  the  early  life  history  of  the 
Anguilliformes  (less  so  in  the  Notacanthiformes)  are  undoubt- 
edly their  highly  distinctive  leptocephali.  Eggs  and  earliest  larvae 
are  very  poorly  known.  Those  of  the  saccopharyngoids  and  no- 
tacanths  have  not  been  identified.  Grassi  (1913),  Schmidt  (1913), 
D'Ancona(1931b)and  Sparta  (1937  e^^e^M.)  described  eggs  and 
developmental  stages  of  several  Mediterranean  eel  species,  mostly 
from  reared  material.  The  basis  for  identification  of  eel  eggs  was 
thus  reliably  established.  Some  errors  have  been  made:  Eigen- 
mann's  (1902)  eggs  of  Conger  oceanicus  were  apparently  those 
of  Ophichlhus  cruenlifer  {Nap\m  and  Obenchain,  1980);  Fish's 
(1928)  Angiulla  rostrala  eggs  were  those  of  the  muraenid  An- 
archias  yoshiae  (Eldred,  1968).  Little  further  information  has 
been  added  recently,  although  Naplin  and  Obenchain's  (1980) 
detailed  account  of  Ophichlhiis  cruent ifer  demonsUalcs  the  use- 
fulness of  matching  planktonic,  newly  hatched  larvae  with  late 
stage  embryos.  Yamamoto  et  al.  (1975a,  b)  described  live  eggs 
and  early  larvae  of  Angnilla  japonica  spawned  from  a  ripe  fe- 
male that  had  been  artificially  matured,  but  there  have  been  few 
//;  v/vo  studies.  There  is  no  comprehensive  information  available 
for  the  identification  and  comparison  of  eel  eggs,  principally 


Tabi  E  10.    Extended. 


Serrivo- 

Anguil- 

Monn- 

Heteren- 

Muraeni- 

Myrocon- 

Xenocon- 

Nemich- 

Cyemati- 

Sacco- 

Eury- 

Mono- 

meridae 

hdae 

guidae 

chelyidae 

dac 

gndae 

gridae 

ihyidae 

dae 

pharyngidae 

pharyngidae 

gnalhidae 

137 

100 

98 

108 

107 

131 

97 

170 

74 

138 

97 

88 

170 

119 

+ 

180 

227 

216 

131 

156 

400  + 

108 

250 

125 

95 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

-1- 

+ 

+ 

(+) 

+ 

+ 

+ 

(+) 
(+) 

+ 

+ 

(+) 
(+) 

+ 

+ 

+ 

+ 

-1- 

-1- 

-1- 

+ 

+ 

+ 

-1- 

-t- 

+ 

+ 

+ 
(  +  ) 

+ 

+ 

-1- 

-1- 

-1- 

-1- 

(  +  ) 

+ 

(+) 

+ 

(  +  ) 

+ 

+ 

(  +  ) 

-1- 

+ 

+ 

+ 

+ 

+ 

+ 

-1- 

-1- 

+ 

+ 

+ 

-1- 

-1- 

+ 

+ 

+ 

-1- 

-1- 

-1- 

-1- 

+ 

+ 

+ 

-1- 

+ 

+ 

-1- 

-1- 

+ 

+ 

+ 

-1- 

-1- 

+ 

-1- 

+ 

66 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES- AHLSTROM  SYMPOSIUM 


Table  1 1.    Some  Osteological  Characters  of  the  Anguilliformes.  +  =  All  or  most  species;  (  +  )  =  some  species;  *  =  presumed  primitive 

condition. 


Synapho- 

Dysom- 

Simen- 

Ophichthi- 

Muraene- 

Netlasto- 

Colocon- 

Dench- 

branchidae 

matidae 

chelyidae 

dac 

Congridae 

socidae 

matidac 

gndae 

thyidae 

Frontals:  Separate* 

Fused 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

Pterygoid:  Present* 

+ 

+ 

+ 

+ 

Reduced 

+ 

(+) 

+ 

+ 

(  +  ) 

Absent 

(+) 

Hyomandibula:   Forward* 

+ 

+ 

+ 

+ 

+ 

+ 

Vertical 

(  +  ) 

(+) 

(  +  ) 

Backward 

+ 

+ 

+ 

Lateral  line  ossifications: 

Present 

+ 

+ 

+ 

+ 

Absent* 

+ 

+ 

+ 

+ 

+ 

Gill  arches: 

More  or  less  complete* 

+ 

+ 

(  +  ) 

+ 

+ 

+ 

(  +  ) 

Variously  reduced 

+ 

(  +  ) 

(  +  ) 

because  only  a  few  species  have  been  studied  from  just  six 
families.  Major  characters  of  eggs  of  these  families  are  collated 
in  Table  12,  which  also  includes  selected  references.  Eggs  and 
earliest  larvae  of  Ophichthus  cruentifer  are  illustrated  as  an  ex- 
ample in  Fig.  30. 

Eel  eggs  are  large;  the  chorion  is  thin  and  clear,  but  may  have 
minute  chromatophores;  the  perivitelline  space  is  wide;  the  yolk 
makes  up  about  one  half  of  the  egg  diameter  and  is  segmented, 
with  or  without  chromatophores.  Oil  globules  are  usually  pres- 
ent (absent  in  Muraenidae  and  Nettastomatidae)  but  the  number 
and  size  may  vary  during  development.  Development  takes 
around  4  days  at  about  20  C  in  Gnathophis  mystax  (Thomo- 


poulos,  1956)  and  in  O.  tTM£'n//7er(NaplinandObenchain,  1980) 
but  may  be  several  days  longer.  The  yolk  reduces  in  size  and 
the  embryo  reaches  a  hatching  length  of  about  4.5-5.5  mm, 
coiling  once  or  more  around  the  yolk.  While  the  late  embryo 
may  possess  conspicuous  melanophores  and  segmentation,  the 
definitive  number  of  myomeres  and  the  characteristic  pigmen- 
tation of  the  lai~vae,  if  any,  are  not  usually  fully  established  until 
after  hatching. 

Leptocephali.—The  yolk-sac  larva  ("preleptocephalus"  or  en- 
gyodontic  stage)  which  is  liberated  from  the  egg  is  characteris- 
tically elongate,  with  a  tear-drop  shaped  to  elongate  yolk.  It 


Table  12.    Characters  of  Anguilliform  Eggs. 


Family 

1 

2 

Ophichthus 

Ophichthus 

Dalnphi 

. 

ipterichtus 

Ophisurus 

Echehis 

Ophichthid 

Facciolella 

Character 

cruentifer 

remicaudus 

imberbis 

caecus 

.serpens 

mvnis 

(unident  ) 

oxyrhvncha 

Diameter  of  chorion:  Min. 

1.62 

2.10 

2.20 

3.00 

3,04 

3.04 

3.40 

2.96 

Max. 

2.89 

2.40 

2,40 

3.60 

4.00 

3.80 

3.68 

3.24 

Diameter  of  yolk:  Min. 

1.32 

1.32 

1.68 

2.10 

1.60 

1.32 

1.48 

Max. 

1.60 

1.60 

1.60 

1.92 

2.20 

1.85 

1.80 

1.84 

Oil  globule(s):  Absent 

+ 

Present 

-1- 

-1- 

-1- 

+ 

-1- 

+ 

+ 

Number        Min. 

1 

6 

1 

3 

11 

1 

Max. 

1 

22 

4 

40 

28 

1 

11 

Size       Min. 

0.26 

0.08 

0.32 

Max. 

0.65 

0.16 

0.36 

0.36 

Pigment  of  embryo: 

Present  on  caudal 

-t- 

-t- 

-1- 

+ 

-1- 

Present  on  gut 

-1- 

-1- 

+ 

+ 

+ 

Present  on  spmal  cord 

Chorion  smooth: 

-1- 

-1- 

+ 

+ 

+ 

+ 

+ 

+ 

Yolk  segmented: 

-1- 

-1- 

+ 

+ 

+ 

+ 

+ 

+ 

Reference 

a 

b 

b 

c 

d 

e 

f 

g 

Families  represented: 

References: 

a- 

-Naplin 

and  Obencham 

1980 

h  — Sparta, 

1942a 

1     Ophichthidae 

b- 

-Sparta, 

1937 

i— Sparta, 

1939d 

2    Nettastomatidae 

c- 

-Sparta, 

1938a 

j— Sparta, 

1939b 

3    Xenocongridae 

d- 

-Sparta, 

1939c 

k  — Sparta, 

1938b 

4    Congridae 

e- 

-Sparta, 

1940a 

1— Castle  and  Roberison, 

1974 

5    Muraenidae 

f- 

-Sparta, 

1940b 

m  — Mannaro,  1971 

6    Anguillidae 

g- 

-Sanzo, 

1938a 

n-Eldred, 

1969 

0— Yevseyenko,  1974 

CASTLE:  NOTACANTHIFORMES.  ANGUILLIFORMES 


67 


Table  11,    Extended. 


Scmvo- 
mendae 

Anguil- 
hdac 

Monn- 
guidae 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

Heleren-  Myrocon-  Xenocon- 

chelyidae  Muraenidae  gndae  gridae 


Eury- 
Ncmich-  Sacco-  pharyngj-  Mono- 

thyidae  Cyematidae        pharyngidae  dae  gnathidae 


+? 


+? 


+? 


(+) 

+ 


+? 


+ 
+ 


+? 


somewhat  resembles  later  stages  but  the  development  of  larval 
characters  is  progressive.  There  may  be  substantial  differences 
in  pigmentation  between  this  stage  and  the  fully  grown  lepto- 
cephalus  (e.g..  the  congrid  Ariosoma,  Table  17  E,,  Mj,  O  and 
Fig.  37);  typically  the  pigmentation  pattern  is  much  less  com- 
plex. The  engyodontic  stage  has  few,  needle-like  teeth,  lower 
jaws  equal  to,  or  longer  than  upper,  an  unformed  nasal  capsule, 
and  undifferentiated  median  fin-folds  and  hypurals. 

At  about  20  mm  TL  the  leptocephalus  then  enters  the  eury- 
odontic  stage  which  lasts  until  metamorphosis.  It  begins  with 
shedding  of  the  engyodontic  teeth  and  their  replacement  by  3 
series  (usually)  of  shorter,  broad-based  teeth,  the  lower  jaw 


shortens  relative  to  the  upper,  the  head  decreases  in  relative 
length,  and  the  fins  and  hypurals  differentiate.  At  this  stage 
leptocephali  are  highly  distinctive  and  well-known  forms  amongst 
fish  larvae.  At  full  growth  they  are  typically  around  50-80  mm 
but  may  attain  300-400  mm  (Nemichthyidae)  or  1,800  mm 
(Notacanthiformes).  They  are  almost  transparent  except  for  eye 
and  other  pigmentation  and  the  blood  lacks  erythrocytes  and 
haemoglobin.  The  body  is  greatly  compressed  and  leaf-shaped 
or  filamentous,  typically  with  a  small  head,  prominent,  for- 
wardly-directed  larval  teeth  and  a  posteriorly  placed  anus.  The 
electrolyte  make-up  of  their  body  fluids  differs  markedly  from 
that  of  postmetamorphic  forms  (Hulet,  1978). 


Table  12.    Extended. 


Family 

2 

3 

4 

5 

6 

Neltastoma 
nielanurum 

C'h/opsis 

hicolor 

Conger 

conger'' 

Ariosoma 
baleancum 

(jnalhophis 
Gnathophis  sp,               mystax 

Muraena 
helena 

(ivmnothorax 
unicolor 

( i   nigro- 
marginanis 

Angiulla 
iinguiUa? 

2.40 
3.00 
1.44 
1.48 

+ 


2.72 
3.04 
1.40 
1.48 

-I- 

13 

0.04 
0.08 


2.60 
1.7 
-I- 
1 
0.40 


1.80 
1.92 
1.00 
1.04 

-I- 
1 
5 

0.30 


2.93 
3.43 
1.25 
1.50 

-I- 
1 
9 

0.03 
0.10 


2.50 
3.00 
1.50 
1.85 

-t- 


5.0 
5.5 


2.3 
3.4 


3.3 
4.0 
1.5 
2.0 

-I- 


2.3 
2.9 
1.3 
1.6 


1 

2 

0.31 

0.42 


-I- 
+ 
J 


+ 
+ 


+ 
+ 
1 


+ 
+ 
m 


+ 
-I- 
m 


-I- 
-I- 
-I- 
n 


68 


ONTOGENY  AND  SYSTEM ATICS  OF  FISHES -AHLSTROM  SYMPOSIUM 

B  C 


mm 


D 


OPHICHTHUS   CRUENTIFER 


Fig.  30.     Embryonic  and  early  engyodontic  stages  of  Ophichthus  cruenltfer  (adapted  from  Naplin  and  Obenchain,  1980). 


I 


CASTLE:  NOTACANTHIFORMES,  ANGUILLIFORMES 


69 


Metamorphosis  follows  the  euryodontic  stage.  It  is  relatively 
abrupt  and  involves  the  replacement  of  many  of  the  character- 
istic leptocephalus  features  by  those  of  the  juvenile.  The  body 
rounds  up  in  section,  tissue  transparency  is  lost,  the  postorbital 
portion  of  the  head  lengthens,  the  larval  teeth  are  lost  and  the 
definitive  teeth  are  gradually  substituted.  The  anus  and  median 
fin  origins  move  forwards,  though  not  in  all  species.  Pectoral 
and  caudal  fins  are  lost  late  in  metamorphosis  in  those  species 
which  lack  the  fin  in  the  juvenile  and  adult.  There  may  be  a 
substantial  reduction  in  body  length,  extremely  so  in  the  No- 
tacanthiformes.  The  principal  characters  which  are  retained  are 
the  definitive  number  of  myomeres/vertebrae  which  is  estab- 
lished very  early  in  larval  life,  the  number  of  dorsal  and  anal 
fin-rays  which  is  attained  rather  late  in  development,  and  for 
some  species  the  larval  pigment.  The  maintenance  of  larval 
pigment  through  metamorphosis  is  of  prime  importance  in  iden- 
tification at  the  generic  level.  However,  metamorphic  larvae  are 
relatively  rare  in  collections,  possibly  because  they  are  in  any 
case  a  transient  stage;  metamorphics  are  also  benthic  and  hence 
less  accessible  to  collection.  Information  on  these  important 
stages  is  therefore  sparse. 

Identification 

Leptocephali  are  thus  readily  recognisable  amongst  other  fish 
larvae,  apparently  abundant  in  the  warmer  ocean,  and  accessible 
near  the  surface.  Large  collections  of  leptocephali  have  accu- 
mulated, for  some  families  and  genera  there  being  many  more 
specimens  available  than  of  the  adults  (e.g.,  the  moringuid,  Neo- 
conger.  Smith  and  Castle,  1972;  the  Nettastomatidae,  Smith 
and  Castle,  1982).  The  availability  of  such  collections  and  the 
need  for  identification  of  leptocephali  have  resulted  in  the  recent 
rapid  advance  of  larval  studies  (Castle.  1969;  Blache.  1977; 
Smith.  1979;  Fahay.  1 983).  These  studies  have,  understandably, 
emphasized  identification  rather  than  inter-relationships  based 
on  larval  characters. 

Larvae  of  all  but  the  monotypic  families  Simenchelyidae  and 
Myrocongridae  and  those  of  about  half  (82)  of  the  genera  are 
known.  Several  distinctive  larval  forms,  possibly  of  undescribed 
genera  rather  than  families,  are  also  known  (e.g.,  the  congrid- 
like  Leptocephalus  thorianus  Schmidt,  Smith,  1979).  Family 
identification,  largely  by  morphological  and  pigment  characters, 
may  be  arrived  at  from  Table  13,  which  incorporates  infor- 
mation set  out  in  key  form  by  Smith  (1979)  and  Fahay  (1983). 
This  "look-alike"  approach  to  identifying  leptocephali  largely 
suffices  at  the  family  level  but  is  less  satisfactory  in  identifying 
genera,  especially  of  the  Ophichthidae  and  Congridae  (Leiby, 
1981).  More  detailed  information  may  be  necessary,  especially 
for  species  identification,  but  this  will  be  slow  to  accumulate. 
Some  attempt  to  collate  available  data  for  identification  pur- 
poses is  made  in  Tables  14-23,  with  their  complementary  figures 
(Figs.  34  to  43). 

More  than  500  different  leptocephali  have  been  described, 
200  as  nominal  species  of  the  invalid  genus  Leptocephalus  Gron- 
ovius,  1763.  The  procedure  of  formally  naming  eel  larvae  in 
this  way  has  been  both  opposed  (Bohike  and  Smith,  1968)  and 
advocated  (Castle,  1969).  However,  nomenclatural  problems 
associated  with  naming  larval  forms  will  not  be  readily  over- 
come by  ignoring  the  priority  of  larval  names  or  attempting  to 
apply  a  blanket  restriction  on  their  use.  Some  alternative  ref- 
erence scheme,  or  at  least  an  agreed  descriptive  procedure,  does 
seem  appropriate  (Fahay  and  Obenchain,  1978)  to  accommo- 
date the  large  number  of  distinctive  ontogenetic  stages  of  eels. 


Fig.  31.  Anterior  region  of  leptocephalus  of  an  unidentified  ?net- 
tastomatid  (DANA  St.  4181  II,  34<'23'N,  25°53'W,  9  June  1931),  show- 
ing tab-like  extensions  of  the  intestine. 


Few  complete  growth  series  have  been  described  and  illus- 
trated, and  developmental  osteology  is  known  only  for  Anguilla 
anguilla  (Norman,  1926b),  Serrivoiner  spp.  (Bauchot.  1959). 
Ariosorna  baleancum  (Hulet.  1977).  Ophichthus  gomesi  (Leiby, 
1979a),  and  Atyrophispunctatus {Leiby,  1979b).  At  least  in  Oph- 
ichthus gomesi  ossification  of  the  head  skeleton  does  not  occur 
for  most  elements  until  metamorphosis,  although  the  jaws,  sus- 
pensorium  and  branchial  skeleton  are  present  as  cartilage  during 
the  pre-metamorphic  stage.  Leiby's  recent  papers  (1979b,  1981) 
contain  detailed  information  on  the  sequence  of  development 
of  the  skeleton  and  emphasize  the  relevance  of  a  more  thorough 
evaluation  of  developmental  osteology  in  identification  of  lep- 
tocephali. 

In  overall  body  form  leptocephali  range  from  the  greatly  elon- 
gate notacanths  (Castle,  1973,  for  references;  Smith,  1979;  Fig. 
29),  Nemichthys  (Nielsen  and  Smith,  1978;  Smith,  1979;  Table 
19)  and  some  Nettastomatidae  (Smith  and  Castle,  1982)  to  the 
short,  deep  larvae  of  Thalassenchelys  (Castle  and  Raju,  1975; 
Table  22  and  Fig.  42).  the  Xenocongridac  (Smith.  1969;  Table 
22  and  Fig.  42)  and  Cyema  atrum  (Smith,  1979;  Table  23  and 
Fig.  43). 

The  snout  is  typically  rather  sharp,  especially  so  in  some 
Notacanthiformes  (Fig.  29),  Dysommatidae  (Table  14  and  Fig. 


70 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


20  3  mm 


ENGYODONTIC 


37'»  mm 


Gnathophis 


856  mm 


EURYODONTIC 
Fig.  32.     Development  of  teeth-series  in  the  congrid  Gnathophis. 


34),  Nettastomatidae  (Table  19  and  Fig.  39)  and  Cyematidae 
(Table  23  and  Fig.  43),  but  characteristically  short  and  rounded 
in  the  Heterenchelyidae  (Table  18  and  Fig.  38)  and  Muraenidae 
(Table  21  and  Fig.  41),  especially  near  metamorphosis.  In  some 
Dysommatidae  (Table  14  and  Fig.  34)  it  is  produced  forwards 
as  a  conspicuous,  narrow,  ethmoid  rostrum  bearing  at  its  tip  a 
pair  of  "premaxillary"  teeth  and,  in  some  also,  fleshy  tabs  or 
tentacles  along  its  length.  The  rostrum  itself  is  lost  at  meta- 
morphosis so  that  the  snouts  of  post-metamorphic  dysomma- 
tids,  apart  from  their  characteristic  papillae  and  plicae,  are  sim- 
ilar to  those  of  other  eels. 

In  full-grown  leptocephali  the  anus  lies  just  in  advance  of  the 
midpoint  (some  Nettastomatidae,  Table  19  and  Fig.  39;  some 
Muraenidae,  Table  21  and  Fig.  41;  some  Xenocongridae,  Table 
22  and  Fig.  42),  well  behind  the  midpoint  (most  genera),  or  is 
subterminal  (the  congrid  group  Ariosoma-Bathymyrus.  Table 
1  7  and  Fig.  37).  For  those  in  which  it  is  subterminal,  it  advances 
during  metamorphosis,  taking  with  it  the  anal  fin  origin  and  the 
developing  pterygiophores  and  actinotrichia.  Its  position  in  these 
species  is  thus  a  very  rough  measure  of  the  stage  of  metamor- 
phosis. Broadly  speaking,  the  amount  of  forward  movement  of 
the  anus  is  correlated  with  the  length  of  larval  life,  generally 
long  in  Notacanthiformes,  Anguillidae  (1-3  years)  and  Congri- 
dae  (10  months  for  species  of  Gnathophis,  Castle,  1968;  Castle 
and  Robertson,  1974)  but  much  shorter  in  Moringuidae  (3'/2 
months  for  Moringua  edwardsi.  Castle,  1979)  and  probably  also 
for  Muraenidae,  Xenocongridae  and  many  Ophichthidae.  How- 
ever, little  is  known  of  the  duration  of  larval  life  in  most  eels. 

A  special  feature  of  some  Ariosoma-Bathymyrus  larvae  is  an 
exterilium  or  external  intestine  (Mochioka  et  a!.,  1982;  Table 
17Q  and  Fig.  37)  and  in  the  unidentified  larva  illustrated  by 
Weber  (1913)  and  Smith  (1979),  there  are  tab-like  extensions 
of  the  intestine,  of  unknown  significance  (Fig.  31). 

The  olfactory  organ  is  a  round  to  oval  sac  immediately  in 


front  of  the  eye.  As  growth  proceeds  its  single  aperture  pro- 
gressively becomes  vertically  subdivided  by  flaps  growing  from 
the  upper  and  lower  margins.  After  separation  of  the  two  nos- 
trils, the  olfactory  sac  lengthens  in  many  leptocephali,  except 
the  Cyematidae,  Nemichthyidae  and  Serrivomeridae,  so  that 
the  anterior  nostril  moves  forwards  to  near  the  tip  of  the  snout. 
There  it  becomes  subtubular  and  often  turns  downwards;  late 
in  metamorphosis  the  posterior  nostril  may  move  dorsally  or 
ventrally  to  adopt  its  final  position  above  or  behind  the  eye  or 
ventrally  on  or  through  the  upper  lip. 

The  eye  is  usually  round,  but  in  the  notacanthiform  larvae 
referred  to  the  larval  genus  Tiluropsis,  and  in  Leptocephaliis 
attemiatus,  it  is  characteristically  oval,  with  the  long  axis  ver- 
tical. In  all  Synaphobranchoidea,  probably  also  including  the 
Simenchelyidae,  the  eye  assumes  a  so-called  "telescopic"  or 
"tubular"  shape  (Table  14  and  Fig.  34)  and  the  body  of  the  eye 
faces  anterodorsally  and  is  elongate,  with  a  very  deep  retina. 

Teeth  develop  shortly  after  hatching.  These  engyodontic  teeth 
(Fig.  32)  are  few,  needle-like,  forwardly  directed,  each  one  pro- 
gressively shorter  along  the  rami  of  the  jaws;  typically  there  is 
a  pair  of  larger  teeth  anteriorly.  The  engyodontic  teeth  are  shed 
at  the  beginning  of  the  euryodontic  growth  stage  and  are  pro- 
gressively replaced  with  the  3  series  of  shorter,  broad-based  teeth 
in  upper  and  lower  jaws;  the  upper  teeth  are  preceded  by  an 
anteriormost  pair,  slightly  smaller  than  the  first  maxillary  pair, 
which  are  very  large  in  the  supposed  xenocongrid  Thalassenche- 
lys  (Table  22  and  Fig.  42).  As  growth  proceeds  teeth  are  added 
progressively,  to  reach  40-50  at  metamorphosis.  They  are  blade- 
like and  slightly  recurved  in  Paraconger,  bicuspid  in  Coloconger 
(Table  1 8  and  Fig.  38),  or  needle-like  and  distinctly  spaced  in 
the  Heterenchelyidae  (Table  18  and  Fig.  38).  Leiby  (1979b) 
notes  that  the  splanchnocranium  is  so  weakly  developed  in  the 
engyodontic  stage  of  the  ophichthid  Myrophis  pimctatus  that 
the  first  series  of  larval  teeth  cannot  be  used  in  feeding. 


I 


CASTLE:  NOTACANTHIFORMES,  ANGUILLIFORMES 


71 


150 


UO 


130- 


120- 


nO' 


CO 

O 
O 


^   100 

c/) 

O 

I 

CO 

5 


90 


70' 


60 


50 


40 


MYROPHINAE 

1  Myrophis  punctalus 

2  M.  plumbeus 

3  Ahlia  egmontis 

u    Pseudomyrophis  nimius 
OPHICHTHINI 

5  Aplatophis  chauliodus 

6  Ophichthus  rex 

7  O.  ocellalus 

8  O.  gomesi 

9  O.  cruentifer 

10  O.  melanoporus 

11  Echiophis  mordax 

12  Myrichthys  oculatus 

13  M.  acuminatus 


SPHAGEBRANCHINI 

u   Ichthyapus  ophloneus 

15  Apterichtus  ansp 

16  ^.  kendalh 

17  Stictorhinus  potamius 
CALLECHELYINI 

18  Letharchus  velifer 

19  Callechelys  muraena 

20  C,  springer! 

21  C.  perryae 


BASCANICHTHYINI 

22  Carolophia  loxochila 

23  Bascanichthys  scuticans 
2'-  6.  bascanium 

25  Gordiichthys  irrelitus 

26  Phaenomonas  longissimus 


/» 


/• 


25 


-y^' 


i^^ 


A^^ 


20 


.^^^ 


^^,^^' 
-\^^ 


^^ 


'23 


^i'     019 


022 


04 


017 


150     7- 


1= 


/ 


016 


•3 
1 


•^ 


13 


09 


^^ 


oio 


02 


cTu 


"-r~ 

110 


I 

120 


I 

130 


I 

uo 


— I — 

150 


— I — 
180 


100 


I 

160 


1 

170 


190 


— I — 

200 


— I — 
210 


— I — 
220 


MEAN  TOTAL  VERTEBRAEIADULTS)  MYOMERESO-ARVAE) 


Fig.  33.     Position  of  kidney  in  adults  and  larvae  of  26  species  of  Western  Atlantic  Ophichthidae;  black  circles  adults,  open  circles  larvae. 
Adults  of  not  all  species  shown. 


The  gill  opening  is  anteroventral  to  the  pectoral  base  and  any 
movement  to  take  up  an  adult  ventral  position  (Synaphobran- 
choidea,  Ophichthidae)  does  not  occur  until  very  late  in  meta- 
morphosis. 

Pectoral  fins  are  present  as  fleshy  tabs  in  all  very  early  lep- 
tocephali.  If  absent  or  much  reduced  in  the  post-metamorphic 
stage,  the  loss  does  not  occur  until  late  in  larval  life  or  at  meta- 
morphosis (Muraenidae,  Ophichthidae,  the  muraenesocid  Gav- 
laliccps).  Actinotrichia  do  not  develop  until  late  in  the  eury- 
odontic  stage  and  lepidotrichia  not  until  metamorphosis.  The 
range  is  8-22  among  the  species  of  eels. 


Median  fins  are  first  visible  as  undtflferentiated  folds  of  tissue 
and  remain  so  until  the  beginning  of  the  euryodontic  stage.  The 
dorsal  and  anal  fin  skeletons  begin  to  develop  posteriorly  first, 
and  then  progressively  forwards,  the  anal  more  rapidly  than  the 
dorsal.  Pterygiophores  and  associated  muscle  blocks  appear  be- 
fore the  actinotrichia  but  lepidotrichia  do  not  complete  devel- 
opment until  metamorphosis  is  complete.  The  anal  fin  supports 
are  usually  closely  packed  before  the  anus  moves  forwards  dur- 
ing metamorphosis.  The  dorsal  origin  is  less  easy  to  define  until 
late  in  the  euryodontic  stage  and  may  not  take  up  its  final  po- 
sition until  well  into  metamorphosis.  In  the  muraenids  Anar- 


72 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  13.    Major  Morphological  and  Pigment  Characters  of  Anguilliform  Leptocephali  (Families).  +  =  All  or  most  species;  (  +  )  =  some 

species  only. 


Synapho- 

Dysom- 

Simen- 

Ophichthi- 

Muraeneso- 

Nettasto- 

Colocon- 

Derich- 

branchidae 

matidac 

chelyidae 

dac 

Congndae 

cidae 

matidae 

gridae 

thyidae 

Eye:  Tubular 

+ 

+ 

?+ 

Normal 

+ 

+ 

+ 

+ 

+ 

+ 

Hyomandibula:   Backwardly  oblique 

Normal 

+ 

+ 

?  + 

+ 

+ 

+ 

+ 

+ 

+ 

Gut:  A  simple,  straight  tube 

+ 

+ 

+ 

+ 

+ 

with  swellings  or  loops 

1   Swelling 

2  Swellings 

(+) 

3  Or  more 

+ 

+ 

(  +  ) 

(+) 

Body  depth:   a50%TL 

Much  <50%TL 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

Tail  tip:  Broad,  rounded 

Narrow 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

Gut  length:   sHalfTL 

(  +  ) 

(  +  ) 

>HalfTL 

+ 

+ 

+ 

+ 

+ 

(  +  ) 

+ 

+ 

Head:  Elongate 

(  +  ) 

(  +  ) 

Short 

+ 

+ 

+ 

(  +  ) 

+ 

(  +  ) 

+ 

+ 

Snout:  Rounded 

Acute 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

Pigment:  Entirely  absent 

At  least  some  present 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

None  on  gut 

+ 

+ 

Present  on  gut 

+ 

+ 

+ 

+ 

+ 

+ 

Present  dorsally  in  orbit 

(  +  ) 

Absent  from  orbit 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

(  +  ) 

Present  on  spinal  cord 

(  +  ) 

Absent  from  spinal  cord 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

Patch  below  iris 

+ 

Absent  below  iris 

+ 

+ 

+ 

(  +  ) 

+ 

+ 

+ 

+ 

chias.  Uropterygius  and  to  a  lesser  extent  Channotnuraena  the 
dorsal  and  anal  fins  are  much  restricted  and  distinctive  as  such 
early  in  the  euryodontic  stage  (Table  21  and  Fig.  41).  At  least 
in  the  Ophichthidae  (Leiby,  1982),  even  in  those  species  which 
lack  a  dorsal  fin  in  the  adult,  pterygiophores  and  actinotrichia 
develop  in  the  larvae.  There  is  also  a  marked  correlation  between 
position  of  dorsal  fin  origin  in  larvae  and  adults.  In  the  congrid 
Ariosoma  and  related  genera,  the  anus  is  subterminal  and  the 
dorsal  and  anal  are  also  restricted  but  develop  progressively 
forwards  during  late  larval  growth  (Table  17  and  Fig.  37).  Dorsal 
fin-rays  range  in  number  from  1 10  in  Neocyema  erythrosoma 
to  600-700  in  some  ophichthids,  anal  rays  usually  being  some- 
what fewer.  The  large  number  and  apparent  considerable  vari- 
ability of  median  fin  rays  in  most  eels  has  resulted  in  this  meristic 
character  being  neglected,  but  it  may  be  of  considerable  use  in 
larval  identification  (Leiby,  1981). 

The  caudal  fin  develops  at  least  as  early  as  the  anal,  its  sup- 
porting structure  being  3  hypurals,  the  first  two  joined  distally, 
enclosing  a  foramen.  Typically  hypurals  1  and  2  support  4  rays, 
hypural  3  supports  5  rays,  but  the  hypurals  are  much  broader 
in  the  Synaphobranchoidea,  supporting  about  16  rays.  The  fin 
is  resorbed,  the  rays  shorten,  and  finally  become  embedded  in 
the  tail  tip  of  heterocongrin  and  many  ophichthid  larvae  shortly 
before  metamorphosis. 

Myomeres  differentiate  during  embryonic  development  but 
because  of  their  relatively  high  number  and  small  size  it  is  not 
known  for  any  species  whether  the  definitive  number  of  the 
adult  is  established  then,  or  after  hatching.  However,  differen- 
tiation of  the  most  posterior  myomeres,  as  evidenced  visually, 
appears  to  occur  during  the  engyodontic  stage,  even  for  species 


with  very  high  total  numbers  of  myomeres.  Total  counts  for 
species  with  more  than  about  180  are  difficult  to  make  accu- 
rately, even  in  fully  grown  leptocephali.  Myomeres  are  less  readily 
counted  as  body  transparency  is  lost  at  metamorphosis.  The 
range  in  myomere  number  across  the  Anguilliformes  is  74-78 
in  the  short-bodied  Cyema  atrum  to  more  than  400  in  the  greatly 
elongate  Neinichthys  scolopaceus  (Table  10)  with  ranges  for 
species  of  about  10  myomeres  at  the  lower  end  (e.g..  for  Anguilla, 
Jespersen,  1942)  to  about  30  in  the  range  200-300  (e.g.,  for 
Nettastomatidae,  Smith  and  Castle,  1982). 

Vertebrae  first  begin  to  differentiate  posteriorly  just  before 
metamorphosis  with  the  constriction  of  the  terminal  portion  of 
the  notochord  proceeding  anteriorly. 

The  value  of  vertebral  counts  in  defining  eel  species  has  be- 
come firmly  established  in  eel  studies  (Bohlke,  1978).  The  cor- 
relation of  vertebral  number  with  number  of  myomeres  in  larvae 
was  demonstrated  by  Jespersen  (1942)  for  Angidlla  and  taken 
upextensively  in  recent  years  (Blache,  1977;  Smith,  1979;  Smith 
and  Castle,  1982).  In  utilizing  this  agreement  between  larvae 
and  adults,  associated  phenomena  need  to  be  further  explored 
and  assessed,  e.g..  pleomerism  (the  correlation  in  related  species 
of  vertebral  number  and  maximum  body  length  attained:  Lind- 
sey,  1975),  "Jordan's  Rule"  (the  tendency  for  fishes  in  polar  or 
cool  waters  to  have  more  vertebrae  or  other  meristic  parts  than 
have  related  forms  in  tropical  warm  waters,  Jordan.  1892),  and 
sexual  dimorphism  in  vertebral  number  (as  occurs  in  Aforingua 
edwardsi.  Castle  and  Bohlke.  1976). 

The  existence  of  latitudinal  dines  in  vertebral  number  in  eels 
has  been  proposed,  but  not  convincingly  demonstrated,  except 
possibly  for  the  muraenid  Gymnothorax  panamensis  which 


CASTLE:  NOTACANTHIFORMES.  ANGUILLIFORMES 
Table  13.    Extended. 


73 


Scrrivo- 
mcndae 


Anguil- 
lidac 


Monn- 
guidac 


Heicrcn-  Myrocon- 

chelyidae  Muraenidae  gndac 


Xenocon- 
gndae 


Nemich-  Sacco-  Eury-  Mono- 

ihyidae  Cyematidae        pharyngidae       pharyngidae         gnathidae 


+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 


+ 
+ 


+ 
+ 


+ 

+ 


+ 
+ 


+ 

+ 


+ 

+ 

+ 

+ 

+ 
+ 

+ 

+ 

+ 

+ 

{+) 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 
+ 

+ 
+ 

+ 

+ 
+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

(+) 

+ 

+ 

+ 

+ 

+ 

+ 
+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

(+) 
(+) 

+ 
+ 


+ 
+ 
+ 
+ 
+ 
+ 
+ 


+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 
+ 


+ 
+ 


+ 
+ 
+ 
+ 
+ 
+ 

+ 
+ 


+ 
+ 
+ 
+ 
+ 
+ 
+ 
+ 
+ 
+ 


+ 
+ 
+ 
+ 
+ 
+ 
+ 
+ 
+ 
+ 


+ 
+ 
+ 
+ 
+ 
+ 
+ 
+ 
+ 


Randall  and  McCosker  (1975)  show  to  have  a  mean  vertebral 
range  of  1 43  in  Chile  and  1 25  in  the  Gulf  of  California.  Variation 
across  longitude  is  apparently  not  usual  but  may  be  consider- 
able; for  example,  McCosker  (1977,  1979)  demonstrates  that 
the  ophichthid  Myrichlhys  maculatus  has  a  mean  vertebral  count 
of  153  in  the  East  Pacific  to  195  in  the  Red  Sea. 

Two  other  problems  arise  in  using  vertebral/myomere  char- 
acters in  matching  leptocephali  with  their  adult  species.  These 
are  the  prevalence  of  damaged  tails  in  adults  of  some  species, 
especially  those  that  are  slender-tailed  (Nettastomatidae,  some 
Congridae  and  Muraenesocidae)  and  hence  the  unavailability 
of  vertebral  counts;  and  the  overlap  or  near  concordance  of 
vertebral  numbers  within  species  groups.  For  example,  in  the 
western  Indian  Ocean  there  are  15-20  species  of  the  muraenid 
genus  (iymnolhorax  which  have  vertebral  numbers  within  the 
range  130-145.  Unless  other  characters  (e.g.,  fin-ray  numbers) 
can  be  shown  to  differ  significantly  between  these  species,  it  is 
likely  that  their  leptocephali,  all  having  rather  similar  pigmen- 
tation, will  prove  difficult,  if  not  impossible,  to  identify. 

However,  there  is  a  reliable  correlation  between  the  segmental 
position  of  the  larval  kidney  and  that  of  the  adult.  The  larval 
nephros  (opisthonephros)  is  typically  an  elongate  sac  lying  above 
the  gut  approximately  in  the  middle  of  the  body,  i.e.,  near  the 
anus  in  those  larvae  with  a  relatively  short  gut  (Xenocongridae, 
Nettastomatidae,  Ophichthidae)  or  some  distance  in  front  of  it 
in  those  having  a  long  gut  (Congridae).  The  segmental  position 
of  the  kidney  changes  little,  ifat  all,  during  larval  life  and  through 
metamorphosis  into  the  juvenile.  Its  position  then  very  ap- 
proximately agrees  with  the  end  of  the  body  cavity  and  the  first 
caudal  vertebra.  The  correlation  in  the  nephros  position  has 


been  successfully  employed  as  an  identification  character  for  the 
Muraenidae  and  other  families  (Blache,  1977)  and  for  some 
Ophichthidae  (Leiby ,  1981)  but  its  value  has  not  yet  been  com- 
prehensively explored  across  the  Anguilliformes  as  a  whole. 
Further  evidence  for  the  stability  of  nephros  position  from  larva 
to  adult,  at  least  in  the  Ophichthidae,  is  provided  in  Fig.  33. 
The  figure  expresses  the  mean  segmental  positions  of  the  end 
of  the  nephros  in  the  larvae  and  adults  of  various  western  At- 
lantic ophichthids  of  the  subfamily  Myrophinae  and  the  four 
tribes  of  the  subfamily  Ophichthinae.  There  is  close  agreement 
in  position  of  the  nephros  between  larvae  and  adults  of  all 
species.  Furthermore,  the  position  of  the  kidney  (and  first  caudal 
vertebra)  is  conspicuously  further  back  along  the  body  in  the 
tribes  Callechelyini  and  Bascanichthyini.  These  are  readily  re- 
cognisable short-tailed  ophichthids  whose  larvae  can  be  im- 
mediately identified  as  such  by  the  posterior  position  of  the 
nephros.  There  is  considerable  overlap  in  this  character  between 
the  Myrophinae,  Sphagebranchini  and  Ophichthinae  although 
individually  the  species  are  distinct. 

The  larval  nephros  is  typically  supplied  and  drained  by  two 
prominent  blood  vessels  passing  vertically  between  the  lateral 
muscles  to  the  aorta  and  cardinal  veins  below  the  vertebral 
column.  The  segmental  position  of  the  last  of  these  vessels  in 
the  leptocephalus  and  its  correlation  with  the  position  of  the 
first  caudal  vertebra  in  the  adult  has  been  emphasised  in  larval 
identification.  However,  it  seems  simpler  to  use  nephros  posi- 
tion instead. 

In  those  groups  of  larvae  in  which  the  anus  does  not  move 
forwards  during  metamorphosis,  there  is  some  agreement  be- 
tween number  of  preanal  myomeres  and  preanal  vertebrae. 


74 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  14. 


Pigment  and  Morphological  Characters  of  the  Synaphobrachoidea. 

to  Fig.  34. 


+  =  All  or  most  species;  (  +  )  =  some  species  only.  Refer 


atei 

al  pigment 

A. 

A  large  midlateral  patch 

at  about  level  of  anus 

B. 

On  caudal  only 

C. 

A  midlateral  row  of  compact 

or  dendritic  spots 

1 .  Row  complete 

2.  Postanal  row  only 

D. 

A  dorsolateral  row 

E. 

A  ventrolateral  row 

F. 

A  ventral  row 

G. 

A  postanal  row 

Gut  pigment 
H.    Absent 

1.      An  irregular  series  of  dendritic 
melanophores  along  its  length 

Morphological 
J.      Posterior  flexures  of  myomeres 
rounded 

An  opaque  midlateral  area  of 
myomeres  along  length  of  body 
Posterior  flexures  of  myomeres 
angular 

Rostrum  absent 
Rostrum  present 
Gut  straight 

Gut  swollen  or  lightly  arched 
at  points  along  its  length 
Posterior  end  of  gut  markedly 
flexed  downwards 


K. 


M. 
N. 
O. 
P. 


Taxa 

Synapho- 
bronchus 

Nettodarus 

Dvsommma 

Type 

Characters 

A 

B 

C 

D 

(+) 

+ 


+ 
+ 

+ 
+ 


+ 
+ 


+ 
+ 


(+) 


+ 
+ 


+ 
+ 


(+) 
(+) 
(+) 
(+) 
(+) 
(+) 


+ 
+ 


+ 

(+) 


(+) 
(+) 
(+) 
(+) 
(+) 
(+) 


+ 
+ 


+ 
+ 


However,  this  character  is  not  generally  applicable  in  larval 
identification  because  of  forward  movement  of  the  anus  during 
metamorphosis  in  some  species. 

The  gut  is  most  often  a  narrow  straight  tube,  flexed  down- 
wards under  the  pectoral  fin  and  following  the  ventral  margin 
to  the  posteriorly  placed  anus.  The  stomach  is  usually  visible 
as  a  finger-like  sac  at  about  segment  10.  The  most  frequent 
modifications  of  the  gut  tube  are  loops  or  swellings  at  intervals 
along  its  length,  each  usually  accompanied  by  groups  of  mela- 
nophores (Ophichthidae,  Tables  15-16  and  Figs.  35,  36;  Ac- 
romycter.  Table  18  and  Fig.  38;  some  Nettastomatidae,  Table 
19  and  Fig.  39).  The  number  and  state  of  development  (low, 
moderate  or  conspicuous)  of  the  swellings  may  be  diagnostic  at 
family,  genus  or  species  level  but  is  not  always  so  (Leiby,  1981). 

The  liver,  with  associated  gall  bladder,  fills  much  of  the  space 
anteriorly  between  the  gut  and  the  ventral  margin  of  the  lateral 
muscles.  It  has  two  or  three  lobes  in  the  Ophichthidae  (Table 
15  and  Fig.  35),  the  gall  bladder  on  the  second  or  third  lobe, 
and  the  lobes  may  be  distinct  or  connected  by  a  thin  band  of 
liver  tissue. 

Larval  pigment  is  present  in  larvae  of  all  families  except  the 
Anguillidae  and  may  be  highly  elaborated  to  form  complex  and 
distinctive  patterns.  The  pigmentation,  if  present,  is  usually  much 
simpler  in  the  engyodontic  stage  than  later  stages.  Melanophores 
may  begin  to  appear  in  the  embryo  (in  some  Ophichthidae  as 
several  pigment  patches  on  the  gut  similar  to  those  in  the  larvae; 


in  some  Muraenidae  on  the  spinal  cord)  but  typically  do  not  do 
so  until  the  early  engyodontic  stage.  Pigmentation  sometimes 
reaches  its  full  expression  by  the  beginning  of  the  euryodontic 
stage  but  typically  the  complex  patterns  characteristic  of  the 
Ophichthidae  and  other  families  are  not  complete  until  full 
larval  growth.  Subsequently  pigment  may  be  lost  during  meta- 
morphosis (the  congrid  Ahosoma),  but  may  serve  as  a  highly 
important  character  in  matching  larvae  with  adults. 

Individually,  melanophores  may  be  dendritic  (Dysommati- 
dae.  Table  14  Ci-C,  and  Fig.  34).  ocellate  (Congridae,  Table 
18B  and  Fig.  38B),  compact  (Muraenidae,  Table  21 D  and  Fig. 
4 1 )  or  rather  diffuse  (Moringuidae,  Table  23  C,  and  Fig.  43).  They 
may  be  isolated,  grouped  in  clusters  to  form  conspicuous  pig- 
ment patches  (the  congrid  Bathymynis.  Table  1  7G  and  Fig.  37), 
or  they  may  form  well  defined  lines,  series  or  patterns.  In  most 
families  they  occur  on  the  lateral  body  surface,  including  the 
caudal  fin,  on  the  myosepta  (Ariosoma,  Table  17E  and  Fig.  37; 
Bathymyrus.  Table  1 7E  and  Fig.  37;  many  Ophichthidae,  Table 
16  and  Fig.  36),  or  on  the  ventral  body  wall  (Dysommatidae, 
Table  141  and  Fig.  34;  Congridae,  Table  18Land  Fig.  38).  They 
may  occur  deeper  in  the  tissues,  either  on  the  gut,  liver,  kidney, 
suspended  in  the  mucinous  space  between  the  lateral  muscles, 
associated  with  the  spinal  cord  or  vertebral  column  or,  fre- 
quently, on  the  bases  of  the  caudal,  anal  and  dorsal  fin-rays. 

Although  Blache  (1977)  and  Fahay  and  Obenchain  (1978) 
have  attempted  to  summarise  pigment  patterns  in  some  groups 


CASTLE:  NOTACANTH I  FORMES,  ANGUILLIFORMES 


75 


Fig.  34.     Illustrations  accompanying  Table  14. 


76 


ONTOGENY  AND  SYSTEM ATICS  OF  FISHES- AHLSTROM  SYMPOSIUM 


Table  15.    Morphological  Characters  of  Ophichthidae  (Myrophinae  and  Ophichthinae).  +  =  All  or  most  species;  (  +  )  =  some  species 

only.  Refer  to  Fig.  35. 


Myrophinae 


Ophichthinae 


Characters 


Murae-  Neen-  Pseudo-  Ophich-  Sphage-        Bascanich-  Calle- 

4h!ia  mchlhys  Myrophis  chetys  myrophts  thini  branchini  thyini  chelyini 


A.  Body  depth  (euryodontic  stage) 

1.  >10%TL 

2.  <10%TL 

B.  Gut  loops  or  swellings 

1.  Low 

2.  Moderate  to  pronounced 

C.  End  of nephros 

1.  Above  or  just  before  anus 

2.  4-14  myomeres  before  anus 

D.  Liver  lobes  and  oesophageal  swellings 

1.  Two 

2.  Three 

E.  Caudal  fin  at  metamorphosis 

1.  Present,  normal 

2.  Absent  (or  much  reduced) 

F.  Dorsal  pterygiophores  and  rays  before 
metamorphosis 

1 .  Well  developed;  dorsal  origin 
migrates  forwards  4-6  myomeres 

2.  Weakly  developed;  origin  migrates 
forwards  5-50  myomeres  (or  resorbed) 


+  (+) 


+  (+)  + 


(+) 

+ 

(+) 

+ 

(+) 

+ 

(+) 


(+) 

+ 


(+) 
(+) 


of  lai-vae,  the  significance  of  these  has  not  yet  been  comprehen- 
sively reviewed  across  the  Anguiliiformes.  Furthermore,  the  ex- 
tent of  intraspecific  variability  of  pigment  patterns  has  also  not 
been  assessed.  Any  present  discussion  as  to  the  significance  or 
otherwise  of  similarities  and  differences  in  larval  pigmentation 
must  therefore  be  preliminary. 

The  range  of  pigmentation  in  genera  for  which  larvae  have 
been  identified,  and  for  some  other  forms,  is  summarized  in 
Tables  14-23,  family  by  family.  These  tables,  with  their  accom- 
panying figures  and  morphological  information,  may  be  used 
as  a  guide  to  generic  identification,  and  also  as  a  synopsis  of 
pigment  patterns.  Because  these  are  both  complex  and  diverse 
in  some  families,  they  cannot  always  be  simply  displayed  in 
keys.  In  the  Ophichthidae  also,  and  other  families,  further  pig- 
ment patterns  are  known,  probably  representing  other  genera. 
This  is  particularly  so  of  Indo-Pacific  Anguiliiformes  which  have 
not  been  extensively  studied. 

These  tables  and  figures  highlight  common  features  of  pig- 
mentation: (1)  on  the  gut  or  its  adjacent  body  wall,  often  as  a 
regular,  spaced  series  from  throat  to  anus  (Notacanthiformes, 
Congrinae,  Heterocongrinae.  Heterenchelyidae,  Colocongri- 
dae),  or  as  an  interrupted  series  (Nettastomatidae.  Muraene- 
socidae.  Dysommatidae.  Ophichthidae)  or  in  some  other  form 
(Bathymyrinae,  Heterocongrinae,  Muraenidae,  Nemichthyidae, 
Xenocongridae);  (2)  on  the  lateral  body  surface  (Dysommatidae, 
Congrinae,  Nettastomatidae,  Xenocongridae).  often  associated 
in  some  way  with  the  myosepta  (Ophichthidae,  Bathymyrinae, 


Heterocongrinae,  Serrivomeridae.  Derichthyidae);  (3)  on  the 
spinal  cord  (Nemichthyidae.  Muraenidae);  or  (4)  on  the  bases 
of  the  dorsal,  anal  and  caudal  fins. 

The  broad  perspective  on  the  ontogeny  of  the  Anguiliiformes 
and  Notacanthiformes  given  by  the  preceding  deserves  com- 
ment. 

As  adults,  eels  have  adopted  a  somewhat  conformist  body 
plan  notable  for  reduction  and  loss  of  external  features,  though 
the  component  families  of  the  group  are  more  or  less  discrete 
osteologically.  In  contrast,  through  elaboration  of  the  leaflike 
body  form  and  pigment  patterns  their  larvae  display  a  diversity 
which  matches  that  of  any  other  group  of  teleosts.  This  diversity 
involves  some  distinctive  larval  characters  (morphological  and 
pigmentary)  which  allow  leptocephali  to  be  identified  at  the 
family  level.  These  characters  have  not  been  comprehensively 
assessed;  further  definitive  identification  of  larval  forms  will  aid 
any  future  analysis.  Within  families,  larvae  are  generally  similar 
in  body  form  and  pigmentation  but  there  are  several  remarkable 
exceptions.  There  are  some  discernible  character  gradients  in 
larvae  (e.g.,  the  complexity  of  gut  swellings  or  loops  in  Ophich- 
thidae; pigmentation  of  Congridae).  but  these  may  or  may  not 
be  matched  by  adult  character  gradients.  Detailed  meristic  in- 
formation, as  forthcoming  throughout  larval  development,  is 
the  only  satisfactory  medium  for  species  identification,  espe- 
cially in  the  larger  eel  families. 

Zoology  Department,  Victoria  University  of  Wellington, 
Wellington,  New  Zealand. 


CASTLE:  NOTACANTHIFORMES.  ANGUILLIFORMES 


77 


OPHICHTHINAE 


Fig.  35.     Illustrations  accompanying  Table  15. 


78 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  16.    Pigment  Characters  of  Ophkhthidae  (Mvrophinae  and  Ophichthinae). 

to  Fig.  36. 


+  =  All  or  most  species;  (  +  )  =  some  species  only.  Refer 


Characters 


Myrophinae 


Ophichlhinae 


Bas- 

Aturae-         Myr-  Seen-        Pseudo-      Ophich-     Sphagc-      canich-        Calle- 

Ahiia        mchthys       ophis         chelys       myrophis       thini       branchini      Ihyini       chelyini 


B. 


C. 


D. 
E. 
F. 


Lateral  pigment 
A.    Absent 

A  single  spot  mid-laterally  on  nearly 
every  myomere 

An  oblique  row  (or  streak)  of  compact  spots 
below  midlateral  level 

1.  On  all  or  most  myosepta 

2.  On  only  a  few  myosepta,  often  associated 
with  deep  axial  pigment 

Round  groups  of  spots  scattered  over  body 
Extra  spots  on  dorsal  and  ventral  myosepta 
A  group  of  spots  midway  along  body 

Axial  pigment 
G.    Several  deep  postanal  pigment  clusters  below 
vertebral  column  (sometimes  preanal  also; 
may  be  associated  with  myomere  pigment) 

Gut  pigment 
H.    Scattered  spots  along  gut,  usually  prominent 
groups  above  upward  loops,  below  downward 
loops 

Irregular  along  length,  mostly  between  nephric 
duct  and  crest  of  each  gut  loop 
Loop  pigment  associated  with  spots  on  body  wall 
Conspicuous  pigment  patch  at  crest  of  each  gut  loop 


I. 

J. 
K. 


(  +  ) 
(  +  ) 

(+)  + 


(  +  ) 
(  +  ) 


(+) 

(+) 
(  +  ) 
(  +  ) 


(  +  )  + 


(+) 


{+) 


(+) 

+ 

(+) 

(+) 

+ 

Head  pigment 
L.     Spots  along  upper  jaw  near  bases  of  teeth  and 

often  on  lower  jaw 
M.   On  postorbital  region,  pectoral  base  or 
oesophagus 

Other  pigment 
N.    On  bases  of  anal  rays 
O.    On  body  wall  above  anal  base 
P.     On  bases  of  dorsal  rays 
Q.    On  body  wall  below  dorsal  base,  or  before  it 
R.    On  caudal  base 


+ 

(+) 

(+) 

+ 

+ 

+ 

+ 

+ 

(+) 

+ 

(+) 

+ 

+ 

+ 

+ 

(+) 


+ 
+ 


+ 

(+) 


CASTLE:  NOTACANTHIFORMES,  ANGUILLIFORMES 


79 


OPHICHTHIDAE 


Fig.  36.     Illustrations  accompanying  Table  16. 


80  ONTOGENY  AND  SYSTEMATICS  OF  FISHES- AHLSTROM  SYMPOSIUM 

Table  1 7.    Pigment  and  Morphological  Characters  of  Congridae  (Bathvm'i  rinae,  Heterocongrinae)  and  Miiraenesocidae.  +  =  All  or 

most  species;  (  +  )  =  some  species  only.  Refer  to  Fig.  37. 


I^ara- 
Allo-         Ario-       Uniden-     Balhv-        Paru-         Goi-        fivlern-       Con-         (iuvi-       Miirae-     xcnonjy-     .\cnn- 
conger       soma         tified         niyriis       conger       gasta        conger      gresox       aliccps       ncso.x         s/av        mv^lax 


Lateral  pigment 

A.  Absent 

B.  A  midlateral  row  of  single  spots, 
often  with  extra  spots  below 

C.  A  row  of  few  large  spots  between 
midlateral  and  ventral  levels 

D.  A  large  group  of  dendritic  spots 
at  about  myomere  80 

E.  Oblique  rows  of  compact  spots  on 
myosepta  below  midlateral  level 

1 .  Spots  very  close  together  +  +  +  + 

2.  Spots  scattered 

F.  Additional  oblique  rows  present 

1.  Above  midlateral  level  + 

2.  Below  midlateral  level  + 

G.  A  large  midlateral  patch  of  minute 

spots  at  one  third  of  body  length  + 

H.    Scattered  minute  spots  above  and 

below  midlateral  level  + 

Head  pigment 
I.      small  spots  on  throat 
J.      Small  spots  elsewhere  on  head 
Gut  pigment 
K.    Small  spots  ventrally  before  stomach 

and  dorsally  behind  stomach  +  +  + 

L.     Small  spots  ventrally  behind  stomach  (+)         + 

M.    Series  from  throat  to  anus 

1.  Approx.  one  spot  every  1-2  segments 

2.  Spots  widely  spaced  (in  young  only)  + 

3.  6-9  groups  of  spots 
Other  pigment 

N.    Small  spots  on  anal  and  dorsal  bases  +  +  + 

O.    A  series  of  spots  before  dorsal  fin; 

few,  large  (young);  many,  small 

(full  grown)  +  + 

Morphological: 
P.     Posterior  teeth  bladelike 
Q.    An  "exterilium"  intestine  (  +  )  +  + 


CASTLE:  NOTACANTHIFORMES,  ANGUILLIFORMES 


81 


CONG  Rl  DAE 


El 


7^ 


z;! 


MURAENESOCIDAE 


.<;«t<g.ia;.T-^.7rrrfrnrf^i'^-?yi'^ 


K 


M 


T 


Fig.  37.     Illustrations  accompanying  Table  17. 


82  ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 

Table  1 8.    Pigment  and  Morphological  Characters  of  Congridae  (Congrinae),  Colocongridae  and  Heterenchelyidae.  +  =  All  or  most 

species;  (  +  )  =  some  species  only.  Refer  to  Fig.  38. 


Characters  mycler     congrus        ger         Conger 


Bathy- 

Aero- 

Bathv 

urocon 

myaer 

congrus 

ger 

Pseudo- 

Pseiido- 

Pan- 

Gnalho- 

Hilde- 

xenomv- 

phi- 

Scala- 

I'ro- 

Colo- 

Pviho- 

lunch 

phis 

brandia 

siax 

chfhys 

nago 

conger 

conger 

nwhthys 

ihys 

Lateral  pigment 

A.  A  scattered  row  of  spots  below 

midlateral  level  (rarely)  (+) 

B.  A  row  of  spots  above  and  below 

midlateral  level  (  +  )        (  +  ) 

C.  Many,  small  spots  scattered 

below  midlateral  level  (  +  ) 

D.  Single,  small  spots  all  over 

lateral  surface  (  +  ) 

E.  Spots  scattered  in  groups  over 
lateral  surface 


Axial  pigment 

F. 

On  spinal  cord 

G. 

A  single  midlateral  row 

1.  Spots  widely  spaced 

2.   I  spot  every  1-2  segments 

3.   1  spot  every  segment 

4.  2  spots  every  segment 

5.  spots  widely  spaced. 

dendritic 

H. 

An  extra,  scattered  row  below 

I. 

3  deep  spots  postanally 

(+) 

(+)  + 

(+)  +      (+) 

(+)      +  {+)      +  +  {+) 

(+) 


(+) 


Head  pigment 
J.      A  crescentic  patch  below  eye  + 

K.    Spots  on  throat  + 

Gut  pigment 
L.     Spots  in  two  regular  rows 
from  throat  to  anus 

1 .  Close  together  (every 

1-2  segments)  +  +  +  +  + 

2.  Widely  spaced  (at  least 
in  young) 

M.    Spots  m  clumps  on  gut  loops  + 

Other  pigment 

N.    Above  anal  base  +  (  +  ) 

O.    Along  anal  base  +  +  +  + 

P.     Along  dorsal  base  +  + 

Morphological 
Q.    Some  lower  teeth  bicuspid 
R.     Upper  teeth  needle-like 
S.     Lower  teeth  needle-like 


-1- 

+ 

(+) 

-1- 

-1- 

+ 

-1- 

-1- 

+ 

-1- 

+ 

+ 

-1- 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

CASTLE:  NOTACANTHIFORMES,  ANGUILLIFORMES 


83 


Fig.  38.     Illustrations  accompanying  Table  18. 


84 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  19.    Pigment  and  Myomere  Characters  of  Nemichthyidae  and  Nettastomatidae. 

*  =  larva  unidentified.  Refer  to  Fig.  39. 


+  =  All  or  most  species;  (  +  )  some  species  only; 


Characters 


Avocel-         Lahich- 
ttna  ihys* 


Nemu'h- 
thys 


Facci- 
olella 


Neria-  Netten-  ?Netten- 

sloma  chelys  chetys 


Hop-  Sauren- 

lunnis  chelys 


I'ene- 


Axial  pigment 

A.  Deep  on  vertebral  column 

1.  Single  spot,  or  bipartite 

2.  Several  spots  along  body 

B.  Small  spots  on  top  of  spinal 
cord,  at  least  posteriorly 

Head  pigment 

C.  On  snout  and  lower  jaw 

D.  Deep  behind  eye 

Gut  pigment 

E.  A  ventral  row  of  minute 
spots  before  stomach 

F.  A  row  of  minute  spots  above 
gut  along  its  length 

G.  A  patch  of  minute  spots  on 
liver 

H.    A  patch  of  minute  spots 

below  kidney 
I.     Spots  scattered  between 

liver  and  kidney  patches 
J.      A  regular  longitudinal 

series 


(+) 
(+) 


+ 
+ 


+ 
+ 


+ 
+ 


(+) 
(+) 


+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

(+) 
(+) 

Other  pigment 

K.    Several  groups  of  internal 

spots  along  body  subaxially 

1.  One  or  two  in  each  group 

+ 

2.  Each  group  multiple  (4) 

+ 

L.     Spots  on  ventral  body  wall 

postanally 

(+) 

(+) 

M.    Minute  spots  on  dorsal  and 

anal  bases 

+ 

+ 

+ 

Myomeres/vertebrae 

Min. 

177 

174 

ca. 

238 

186 

209 

ca. 

192 

ca. 

Max. 

216 

191 

400  + 

294 

246 

273 

257 

276 

224 

CASTLE:  NOTACANTHIFORMES,  ANGUILLIFORMES 


85 


N  ETTAS  TO  MA  TIDA  E 


NEMICHTHYIDAE 
A2 


Fig.  39.     Illustrations  accompanying  Table  19. 


86 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  20.    Pigment,  Morphological  and  Myomere  Characters  of  Anguillidae,  Derichthyidae  and  Serrivomeridae.  +  =  All  or  most 

species;  (  +  )  =  some  species  only.  Refer  to  Fig.  40. 


Taxa 


Characters 


Anguilla  Dertchlhys  Nessorhamphiis  Platuromdes  Sernvomer  Slemontdium 


Lateral  pigmeitt 

A.  Absent 

B.  Minute  compact  spots  just  below 
midline  on  nearly  every  segment 

C.  Midline  spots  restricted  to  postanal 
region  (a  few  spots  further  forwards) 

D.  A  series  of  minute  spots  on  body  wall 
postanally 

E.  Minute  spots  on  anal  and  dorsal  bases 

Head  pigment 

F.  Absent 

G.  A  cluster  of  minute  spots  in  orbit 
above  eye 

Morphological 
H.    Gut  length 

1.  0.7  total  length 

2.  0.75  total  length 

3.  0.9  total  length 
I.      Dorsal  fin  origin 

1.  Just  anterior  to  anus 

2.  Just  behind  midlength 

3.  At  about  midlength 

J.      Position  of  last  vertical  vessel 

1.  Behind  mid-gut 

2.  Before  mid-gut 

Myomeres/vertebrae 
Min. 
Max. 


(+) 

+ 

(+) 

(+) 

-1- 

100 
119 


126 
134 


135 
139 


153 
170 


147 
169 


137 
141 


CASTLE:  NOTACANTHIFORMES,  ANGUILLIFORMES 


Fig.  40.     Illustrations  accompanying  Table  20. 


88  ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 

Table  21.    Pigment,  Morphological  and  Myomere  Characters  of  Muraenidae.  +  =  All  or  most  species;  (  +  )  =  some  species  only; 

unidentified.  Refer  to  Fig.  41. 


Taxa 

Characters 

Anarchias 

Channo- 
muraena 

Urn- 
prengius 

Enchely- 
core 

Gyninii- 
thorax 

Muraena 

Thvrso- 
idea 

Unidcn- 
Ulied' 

Uniden- 
iihed* 

Lateral  pigment 
A.     Minute  spots  scattered  over  body 
surface 

+ 

Axial  pigment 
B.     Small  compact  spots  ventrally 
on  spinal  cord,  at  least 
posteriorly 


Head 

pigment 

C. 

Few  to  many,  small,  scattered  spots 

often  compact 

D. 

Similar  spots  on  brain 

Gut  pigment 

E. 

Ventral  row  behind  stomach  only 

F. 

Ventral  row  along  whole  of  length 

G. 

Ventral  row  before  stomach,  dorsal 

row  behind  stomach 

H. 

Short  row  before  anus  dorsally 

1. 

In  disjunct  groups  ventrally 

Other  pigment 

J. 

Before  dorsal  base 

K. 

Along  dorsal  base 

L. 

Before  anal  base 

M. 

Along  anal  base 

N. 

Scattered  over  ventral  surface 

anteriorly 

Morphological 

O. 

Dorsal  and  anal  fins  restricted 

to  tip  of  caudal 

P. 

Dorsal  origin  at  myomere  30-40 

Q. 

Dorsal  origin  at  myomere  40-75 

+  + 

+  + 


(+)  (+) 


+ 

(+) 


+  +  + 

+  +  +(+)+  + 

(+)  + 

+  +  +(+)+  + 


CASTLE:  NOTACANTHIFORMES,  ANGUILLIFORMES 


89 


Fig.  4 1 .     Illustrations  accompanying  Table  2 1 . 


90 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  22.    Pigment  and  Myomere  Characters  of  Xenocongridae.  +  =  All  or  most  species;  (+)  some  species  only; 

Refer  to  Fig.  42. 


larva  unidentified. 


Cates-        Chtlo- 
bya  rhinus 


CMop- 


Kau-         Powell- 
pichthys     ichthys* 


Robin- 


Xeno- 
conger* 


Uniden- 
tified* 


I'niden-     Uniden-    Thalassen- 
tifted*         lifted*         chdys 


Lateral  pigment 

A.  Absent  (first  pair  maxillary  and 
mandibular  teeth  very  large) 

Midlateral  pigment  present 

B.  Irregular  double  row  of  minute 
spots  along  body 

C.  One  minute  spot  per  segment 

D.  Round  groups  of  minute  spots 
along  body 

E.  Large  spots,  widely  spaced 

F.  Axial  spots  confined  posteriorly 

Pigment  elsewhere 

G.  W-shaped  rows  of  minute  spots  on 
anterior  margin  of  segments 

H.    Round  groups  of  minute  spots  all 
over  body 

Head  pigment 
I. 


(+) 


Scattered  spots  behind  eye  and  on 

heart 

A  row  of  spots  along  upper  and  lower 

jaws 

A  patch  below  iris 

A  few  spots  on  snout  tip  or  on 

olfactory  organ 

Gut  pigment 
M.    Minute  spots  ventrally  before 

stomach  and  dorsally  behind  stomach 
Minute  scattered  spots  dorsally  along 
gut  to  anus  or  only  postenorly 
Minute  spots  below  gut 
Large  widely  spaced  spots 
Minute  spots  on  liver 
Round  groups  of  spots  on  gut 

Other  pigment 

S.     Spots  on  anal  base  or  rays 
Myomeres/vertebrae 

Min. 

Max. 


J. 

K. 
L. 


N. 

O. 
P. 

Q. 

R. 


+ 
(+)    + 


+     + 


+     + 


136    98    116    97 
141    107     5     125 


130 
136 


ca. 
157 


142 
163 


CASTLE:  NOTACANTHIFORMES,  ANGUILLIFORMES 


91 


Fig,  42.     Illustrations  accompanying  Table  22. 


92 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  23.    Pigment  and  Myomere  Characters  of  Moringuidae,  Cyematidae  and  Saccopharyngoidea.  + 

tt  =  Castle  and  Raju,  1975;  *  =  Larva  unidentified.  Refer  to  Fig.  43. 


All  species;  t  =  Smith,  1979; 


Taxa 

Monn- 

Neo- 

Uniden- 

Neo-            L   holli 

Sacco- 

Eury- 

Mono- 

Uniden- 

Characters 

H"a 

conger 

tifiedt 

Cyema 

cyema*         group* 

pharynx 

pharynx 

gnalhus 

lifiedtt 

Lateral  pigment 

A.    Absent 

+ 

+ 

+ 

+ 

+ 

B.     Scattered  over  lateral  surface 

+ 

C.     Midlateral  series 

1.  Single  spot  behind  anus 

+ 

2.  Multiple  spots  along  body  (5-1  1) 

+ 

+ 

D.    Short  rows  of  spots  on  myotomes 

dorsally  and  ventrally  (juvenile 

pigment) 

+ 

Head  pigment 

E.     On  snout  and  lower  jaw 

1.  Present 

+ 

+ 

2.  Absent 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

Gut  pigment 

F.     One  large  posterior  spot 

+ 

+ 

+ 

G.    One  small  anterior  spot 

+ 

H.    Series  of  spots  along  gut 

+ 

+ 

I.      Minute  spots  scattered  over 

posterior  surface 

+ 

+ 

+ 

+ 

J.      Few  small  spots  on  liver 

+ 

Myomeres/vertebrae 

Min. 

107 

96 

121 

74 

102 

138 

97 

94 

ca. 

Max. 

180 

110 

132 

78 

104 

250 

125 

113 

105 

CASTLE:  NOTACANTHIFORMES,  ANGUILLIFORMES 


93 


Fig.  43.     Illustrations  accompanying  Table  23. 


Elopiformes,  Notacanthifomies  and  Anguilliformes:  Relationships 

D.  G.  Smith 


NOTACANTHIFORMES 

THE  Notacanthiformes  is  composed  of  two  clearly  defined 
families,  the  Halosauridae  and  Notacanthidae.  Overall,  the 
Halosauridae  is  the  more  primitive  family.  McDowell  (1973) 
divided  it  into  two  subfamilies:  the  Halosaurinae,  containing 
only  Halosaurus,  and  the  Halosauropsinae,  containing  Halo- 
sauropsis  and  Aldrovandia.  The  notacanthids  show  a  number 
of  specializations  not  found  in  the  halosaurs,  involving  mainly 
the  mouth  and  dorsal  fin.  The  Notacanthidae  contains  either 
two  or  three  genera,  depending  on  the  placement  of  Lipogenys. 
McDowell  recognized  only  Nolacanlhus  and  Polyacanthonolus 
in  the  Notacanthidae  while  assigning  Lipogenys  to  a  separate 
family.  He  considered  the  Lipogenyidae  and  Notacanthidae  to 
form  a  suborder  of  the  Notacanthiformes,  the  Notacanthoidei, 
which  stood  opposed  to  the  Halosauroidei.  Greenwood  (1977), 
however,  felt  that  Lipogenys  was  closely  related  cladistically  to 
Polyacanthonolus  and  that  those  two  genera  formed  the  sister 
group  of  Notacanlhus.  A  classification  of  the  Notacanthiformes 
based  on  Greenwood's  interpretation  would  be  as  in  Fig.  44. 

Notacanthiform  larvae  cannot  yet  be  identified  confidently 
below  the  ordinal  level  and  hence  can  tell  us  little  about  rela- 
tionships within  the  order.  Smith  (1970)  gave  reasons  to  suspect 
that  the  Tiluropsis  form  (short  head,  vertically  elongate  eye) 
belongs  to  the  Halosauridae.  Circumstantial  evidence  suggests 
that  the  Tiluriis  form  (short  head,  normal  eye)  is  the  larva  of 
the  Notacanthidae.  Tilunis  is  the  only  notacanthiform  larva 
found  in  the  Mediterranean.  Although  adult  notacanthids  of 
both  Notacanlhus  and  Polyacanthonolus  occur  in  the  Mediter- 
ranean, halosaurs  apparently  do  not  (McDowell,  1973).  The 
identity  of  the  third  basic  type  of  notacanthiform  larva,  known 
as  Leptocephalus  giganteus  (long  head,  normal  eye),  cannot  even 
be  guessed  at  this  point. 

Anguilliformes 

The  Anguilliformes,  the  true  eels,  is  the  largest  and  most 
specialized  of  the  elopomorph  orders.  A  definitive  classification 
of  the  Anguilliformes  does  not  yet  exist.  The  scheme  that  follows 
can  be  considered  an  outline  that  will  be  filled  in  and  modified 
as  studies  continue. 

The  eels  can  be  divided  into  two  groups:  those  in  which  the 
frontal  bones  are  fused,  and  those  in  which  they  remain  as 
separate  right  and  left  elements.  This  observation  dates  back  to 
Regan  (1912),  but  its  phylogenetic  significance  has  not  always 
been  agreed  upon.  Regan  himself  said  nothing  about  it  one  way 
or  another;  he  simply  used  it  as  a  key  character.  A  case  can  be 
made  for  the  view  that  the  fusion  of  the  frontals  was  a  single 
event  that  occurred  quite  early  in  the  evolutionary  history  of 
eels  and  therefore  reflects  a  real  phylogenetic  division.  On  the 
whole  the  fused-frontal  group  contains  more  primitive  members 
than  the  divided-frontal  group,  although  the  fused  condition  is 
itself  a  derived  character  state.  Except  for  Anguilla.  all  the  di- 
vided-frontal eels  are  markedly  specialized,  including  pelagic 
and  fossorial  representatives.  Yet  in  none  of  these  lines  has  a 
fusion  of  the  frontals  been  among  the  modifications.  Of  the  more 


specialized  members  of  the  fused-frontal  group,  all  but  the  Ser- 
rivomeridae  can  be  clearly  traced  back  to  more  primitive  mem- 
bers, all  with  perfectly  fused  frontals.  It  is  more  parsimonious 
to  assume  that  fusion  took  place  once  at  a  point  early  in  an- 
guilliform  evolution  than  to  assume  that  it  occurred  several 
times  early  but  not  at  all  later  on. 

The  number  of  families  in  the  fused-frontal  group  is  still 
somewhat  uncertain.  Ten  are  provisionally  recognized  here.  The 
Synaphobranchidae,  Simenchelyidae,  and  Dysommatidae  are 
closely  related  and  could  easily  be  considered  subfamilies  of  the 
Synaphobranchidae  (Robins  and  Robins,  1976).  They  combine 
some  very  primitive  characters  with  some  peculiar  specializa- 
tions and  do  not  seem  to  be  intimately  related  to  any  of  the 
other  families.  The  Nettastomatidae  shares  several  advanced 
characters  with  certain  congrids  and  could  be  considered  a  de- 
rivative of  that  group.  The  interrelationships  of  the  remaining 
families  are  not  clear;  the  resemblances  involve  mainly  primi- 
tive characters.  The  Ophichthidae  is  a  large  and  morphologically 
diverse  family  containing  both  generalized  and  highly  modified 
forms.  It  is  united  by  certain  specialized  characters  such  as  a 
ventrally  displaced  posterior  nostril,  a  reduced  caudal  fin,  and 
numerous  branchiostegal  rays  that  overlap  on  the  ventral  mid- 
line. The  Congridae  (including  Macrocephenchelyidae)  is  also 
a  large  family,  but  without  the  extreme  variety  of  external  mor- 
phology found  in  the  Ophichthidae.  Its  specializations  are  more 
subtle  and  consist  mainly  of  trends  in  several  characters.  The 
Colocongridae  and  Muraenesocidae  have  at  various  times  been 
included  in  the  Congridae,  but  again  the  resemblances  are  main- 
ly in  primitive  characters.  Neither  family  fits  the  pattern  of 
character  modification  found  in  the  Congridae,  and  both  show 
at  least  one  primitive  character  that  is  absent  in  nearly  all  con- 
grids:  separate  hypohyals.  The  Muraenesocidae  is  here  restricted 
to  Muraenesox  itself  and  its  close  relatives  Congresox,  Cyno- 
ponticus  and  Sauromuraenesox.  Of  the  other  genera  previously 
referred  to  this  family,  Hoplimnis  has  been  removed  to  the 
Nettastomatidae  (Smith,  1979;  Smith  and  Castle,  1982),  and 
Xenomystax  (including  Paraxenomyslax)  probably  belongs  in 
the  Congridae.  The  Derichthyidae  and  Serrivomeridae  are  mid- 
water  eels,  the  former  relatively  little  modified,  the  latter  highly 
modified.  The  Serrivomeridae  was  formerly  associated  with  the 
Nemichthyidae,  but  this  seems  unlikely.  The  completely  fused 
frontals  and  massive  palatopterygoid  arcade  of  serrivomerids 
difier  strikingly  from  the  partially  fused  frontals  and  reduced 
pterygoid  found  in  nemichthyids. 

There  are  eleven  families  of  eels  with  divided  frontals:  the 
Anguillidae,  Moringuidae,  Heterenchelyidae,  Myrocongridae, 
Xenocongridae,  Muraenidae,  Nemichthyidae,  Cyematidae,  Sac- 
copharyngidae,  Eurypharyngidae,  and  Monognathidae  (the 
monognathids  actually  have  fused  frontals,  but  they  are  clearly 
related  to  the  saccopharyngids  and  eurypharyngids  and  the  fu- 
sion seems  secondary).  Although  they  are  more  clearly  defined 
than  the  fused-frontal  families,  their  interrelationships  are  still 
uncertain.  Except  for  the  Anguillidae,  they  are  all  distinctly 
specialized,  either  for  burrowing  (Moringuidae,  Heterenchelyi- 
dae), for  midwater  life  (Nemichthyidae,  Cyematidae,  Sacco- 


94 


SMITH:  ELOPIFORMES,  NOTACANTHIFORMES  AND  ANGUILLIFORMES 


95 


HALO- 
5AURU5 

HAL05AUR- 
0P5I5 

ALDRO- 
VANDIA 

POLY  AC  AN - 
TH0N0TU5 

LIPO- 
0ENY5 

NOTA- 
C  A  NTH  US 

HAL05AURINAE 


NOTACANTHINAE 


HAL05AUR0P5INAE 


HALOSAURIDAE 


NOTACANTHIDAE 


Fig.  44.     Hypothesis  of  relationships  within  the  Notacanthiformes. 


Fig.  45.     Leptocephali  of  Neoconger  (above)  and  Moringua  (below)  (Moringuidae). 


96 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  46.     Heads  of  leptocephali  of  Dysommatidae  (above)  and  Syn- 
aphobranchidae  (below),  showing  telescopic  eye. 


pharyngidae,  Eurypharyngidae,  Monognathidae),  or  as  cryptic 
forms  with  modified  lateral-line  and  gill-arch  characters  (My- 
rocongridae,  Xenocongridae,  Muraenidae).  Two  clear  associa- 
tions are  evident  within  this  group.  One  contains  the  Myrocon- 
gridae,  Xenocongridae,  and  Muraenidae.  These  three  families 
are  relatively  generalized  externally  but  share  a  marked  reduc- 
tion in  gill-arch  elements  and  in  the  lateral  line.  The  second 
association  contains  the  three  families  Saccopharyngidae,  Eu- 
rypharyngidae, and  Monognathidae,  the  so-called  gulper  eels. 
These  are  highly  modified  midwater  eels  with  a  greatly  enlarged 
mouth  and  an  elongated,  posteriorly  directed  suspensorium. 
The  gulpers  show  extreme  reduction  in  all  the  skeletal  elements, 
and  their  relationship  to  other  eels  is  difficult  to  determine. 
Among  the  remaining  families,  the  Anguillidae  is  quite  primi- 
tive morphologically,  but  it  seems  to  have  no  advanced  char- 
acters clearly  linking  it  to  any  of  the  other  families.  The  Mo- 


ringuidae  and  Heterenchelyidae  are  fossorial  forms  that  never- 
theless show  substantial  internal  differences  from  each  other 
(Smith  and  Castle,  1972).  Their  resemblances  may  simply  be 
convergent  adaptations  to  a  similar  way  of  life.  The  Nemichthy- 
idae  and  Cyematidae  both  have  prolonged,  nonocclusible  jaws 
studded  with  liny  recurved  teeth,  but  they  differ  markedly  in 
almost  every  other  character;  their  traditional  association  must 
be  questioned. 

Larval  characters  have  so  far  proved  more  useful  in  eluci- 
dating relationships  within  families  than  between  them.  Some 
examples  will  illustrate  the  contribution  that  larvae  have  made 
to  systematics. 

The  Moringuidae  consists  of  two  genera,  Moringna  and  Neo- 
congcr.  Although  both  are  basically  fossorial  forms,  they  differ 
enough  in  external  appearance  that  for  more  than  a  century  they 
were  placed  in  different  families.  It  was  only  the  striking  simi- 
larity of  the  larvae  (Fig.  45)  that  prompted  a  critical  comparison 
of  the  adults  (Smith  and  Castle,  1972).  In  this  case,  the  larvae 
show  the  relationship  much  more  clearly  than  do  the  adults. 

The  close  relationship  between  the  Synaphobranchidae  and 
Dysommatidae  is  supported  by  a  unique  feature  of  the  larvae— 
the  telescopic  eye  (Fig.  46). 

The  genus  Hoplimnis  has  long  been  placed  in  the  family  Mu- 
raenesocidae  because  of  its  possession  of  a  pectoral  fin  and  its 
enlarged  median  vomerine  teeth.  Saiirenchelys  was  always  con- 
sidered a  nettastomatid  because  it  lacked  a  pectoral  fin.  Smith 
and  Castle  (1982)  showed  that  the  larvae  of  these  genera  are 
indistinguishable  (Fig.  47).  On  that  basis  and  because  of  many 
similarities  in  the  adults,  Hophinnis  and  Saurenchelys  were  shown 
to  be  closely  related  and  to  belong  in  the  Nettastomatidae.  The 
two  characteristic  swellings  in  the  gut  of  larval  Hoplunnis  and 
Saurenchelys  are  also  found  in  the  larvae  of  Nettastoma  and 
Nettenchelys. 

The  major  problem  in  eel  systematics  today  is  the  relationship 
between  the  families,  and  here  larvae  provide  little  help.  Sim- 
ilarities occur  between  larvae  of  families  which  otherwise  show 
no  evidence  of  close  relationship.  For  example,  the  larvae  of 
the  Anguillidae  and  Derichthyidae  are  quite  similar  (the  larva 
of  Derichthys  was  even  named  Lcplocephalus  angiulloides).  but 
the  two  families  do  not  seem  especially  close  and  fall  on  opposite 
sides  of  the  fused-frontals  vs.  divided-frontals  dichotomy.  The 
larvae  of  the  Heterenchelyidae  resemble  those  of  certain  con- 
grids,  but  heterenchelyids  have  divided  frontals  and  congrids 
have  fused  frontals.  Larvae  of  the  congrid  genus  Acromycter 


Fig.  47.     Leptocephali  of  Hoplunnis  tenuis  (above)  and  Saurenchelys  sp.  (below)  (Nettastomatidae). 


SMITH:  ELOPIFORMES.  NOTACANTHIFORMES  AND  ANGUILLIFORMES 


97 


Fig.  48.     Leptocephali  of  Cyema  alrum  (Cyemalidae)  (above)  and  Nemichthys  scolopaceus  (Nemichthyidae)  (below). 


98 


ONTOGENY  AND  SYSTEMATICS  OF  HSHES- AHLSTROM  SYMPOSIUM 


ELOPIDAE 


MEGAL - 
OP  I  DAE 


ALBUL- 
IDAE 


HALO- 
SAURIDAE 


NOTA- 
CANTHIDAE 


EELS- 
21  FAM5. 


ELOPIFORMES 


ANGUILLOIDEI 


Fig.  49.     Hypothesis  of  relationships  between  major  groups  of  elopomorphs. 


(Fig.  52E)  have  a  looped  gut  and  superficially  resemble  certain 
ophichthids  (Fig.  51  A);  on  the  other  hand,  some  ophichthid 
larvae  (for  example,  Basicanichthys.  Fig.  52D)  have  a  weakly 
looped  gut  and  superficially  resemble  congrid  larvae  (Smith  and 
Leiby,  1980). 

A  contraindication  of  relationship  may  be  shown  by  the  larvae 
of  the  Nemichthyidae  and  Cyematidae.  It  was  mentioned  above 
that  these  two  families  differ  in  many  characters  and  that  their 
traditional  association  must  be  questioned.  The  larvae  of  these 
families  are  as  different  from  each  other  as  any  two  leptocephali 
can  be.  Nemichthyid  larvae  are  long  and  slender  with  a  simple 
gut  that  reaches  almost  to  the  tip  of  the  tail.  Cyematid  larvae, 
on  the  other  hand,  are  high  and  deep  and  their  gut  contains 
several  characteristic  loops  (Fig.  48).  Some  observers  have  no- 
ticed a  resemblance  between  cyematid  larvae  and  saccopha- 
ryngoid  larvae  and  have  suggested  that  these  families  are  related 
(Benin,  1937;  Raju,  1974). 

Despite  the  caveats  that  must  be  invoked  when  dealing  with 
the  systematic  implications  of  leptocephali,  these  larvae  play  an 
important  role  in  systematic  studies  of  eels.  They  provide  ad- 
ditional characters  to  be  used  in  systematic  analysis,  and  they 
are  often  more  readily  accessible  than  adults.  The  cryptic  or 
burrowing  habits  of  most  adult  eels  make  them  difficult  to  collect 
in  large  numbers.  The  larvae,  on  the  other  hand,  live  in  open 
water  near  the  surface  and  can  easily  be  collected  with  plankton 
nets  or  midwater  trawls.  In  many  cases,  larvae  provide  data  on 
distribution  and  species  structure  that  are  unavailable  from  adults 
(Smith  and  Castle,  1972,  1982). 


Elopomorphs 

The  Notacanthiformes  and  Anguilliformes  belong  to  a  group 
of  fishes  called  elopomorphs,  along  with  the  Megalopidae,  Elo- 
pidae,  and  Albulidae  (including  Pterothrissidae).  Current  con- 
cepts of  the  interrelationships  of  the  major  groups  of  elopo- 
morphs are  illustrated  in  Fig.  49  (Greenwood,  1977;  Patterson 
and  Rosen,  1977;  Lauder  and  Liem,  1983).  The  trichotomy 
exists  because  there  seem  to  be  no  derived  characters  that  clearly 
link  any  of  the  three  main  branches  with  any  of  the  others. 

Elops  and  Megalops  (including  Tarpon)  seem  more  similar 
to  each  other  than  either  is  to  Alhida.  but  this  may  be  because 
they  are  both  midwater  feeders  with  terminal  mouths,  whereas 
A/hula  is  a  bottom  feeder.  Alhula  has  several  specializations 
(enlarged  cephalic  canals,  prolonged  snout)  that  are  lacking  in 
Elops  and  Megalops.  Most  if  not  all  of  the  resemblances  between 
Elops  and  Megalops  may  be  explained  either  as  primitive  char- 
acters or  as  adaptations  to  a  similar  way  of  life.  Megalops  has 
several  derived  characters  not  found  in  Elops,  most  notably  the 
vascular  air  bladder  and  the  otophysic  connection.  Elops  does 
not  seem  to  have  any  feature  that  is  derived  relative  to  other 
elopomorphs. 

Several  synapomorphies  can  be  cited  to  link  the  Notacanthi- 
formes and  the  Albulidae  (Nelson,  1973;  Greenwood,  1977). 
The  eels  are  usually  placed  on  the  albulid  branch  as  well,  but 
this  is  still  an  open  question.  The  Anguilliformes  and  Notacan- 
thiformes share  a  similar  elongate  body  form,  but  this  feature 
has  evolved  so  many  times  in  fishes  that  it  means  little  by  itself 


SMITH:  ELOPIFORMES,  NOTACANTHIFORMES  AND  ANGUILLIFORMES 


99 


Fig.  50.     Caudal  structure  of  an  anguilliform  leptocephalus  (above)  and  a  notacanthiform  leptocephalus  (below). 


The  only  real  character  seems  to  be  the  swim-bladder  mor- 
phology of  the  two  groups  (Marshall,  1962).  but  a  critical  com- 
parison with  the  swim  bladders  of  Elops  and  Megalops  has  not 
been  made.  Until  that  is  done,  it  cannot  be  determined  whether 
the  swim  bladders  of  eels  and  notacanthiforms  represent  a  syn- 
apomorphy  or  simply  a  general  condition  of  elopomorphs. 

Larvae  probably  cannot  resolve  the  trichotomy.  A  classifi- 
cation based  on  larvae  would  also  yield  three  groups,  but  they 
would  not  be  the  same  three  groups.  The  three  main  groups  of 
larvae  are  the  fork-tailed  group,  the  notacanthiform  group,  and 
the  anguilliform  group.  These  simply  represent  the  condition  in 
the  adults.  The  forked  tail  is  a  primitive  condition  retained  in 
the  Elopidae.  Megalopidae.  and  Albulidae. 

Larvae  do  not  reveal  much  about  relationships  within  the 
fork-tailed  group  either.  The  larvae  of  Elops  and  Megalops  re- 
semble each  other  more  than  they  do  that  of  Albula.  They  are 
smaller,  the  gut  is  shorter,  and  the  dorsal  fin  is  above  or  nearly 
above  the  anal  fin.  Albula  shows  a  trend  toward  elongation, 
although  the  myomeres  are  no  more  numerous  than  those  of 
Elops.  The  gut  is  very  long,  ending  under  the  hypural,  and  the 
dorsal  fin  is  much  farther  forward  than  the  anal  fin.  Pterothrissus 
is  even  more  elongated  and  grows  larger  before  metamorphosis 
than  Albula.  In  albulids  the  myomeres  are  more  V-shaped  than 
W-shaped.  If  the  primitive  condition  is  small  size  and  relatively 


short  larval  life,  then  Megalops  has  the  most  primitive  larva.  It 
is  the  smallest  known  leptocephalus,  metamorphosing  before  it 
reaches  30  mm  standard  length,  at  an  age  of  two  to  three  months 
(Smith,  1980).  Larvae  of  Elops  are  closer  in  size  and  form  to 
those  oi  Megalops  than  to  Albula,  but  this  does  not  necessarily 
demonstrate  that  the  two  former  genera  are  more  closely  related 
to  each  other  cladistically  than  either  is  to  Albula.  It  could  simply 
mean  that  Elops  and  Megalops  retain  a  more  primitive  larval 
form  and  that,  once  again,  they  merely  lack  a  specialization 
found  in  albulids. 

The  larvae  of  the  Notacanthiformes  and  Anguilliformes  do 
not  indicate  a  particularly  close  relationship  between  the  two 
groups.  The  elongated  form  simply  reflects  the  condition  in  the 
adults,  and  in  several  respects  the  two  groups  are  quite  different. 
The  short-based  dorsal  fin  and  the  presence  of  pelvic  fins  in 
notacanthiform  larvae  immediately  separate  them  from  an- 
guilliform larvae.  Eels  lack  pelvic  fins  and  their  dorsal  fin  is  long 
and  confluent  with  the  caudal  and  anal  fins.  In  both  these  char- 
acters the  notacanthiforms  show  the  more  primitive  state.  In 
the  structure  of  the  tail,  however,  the  notacanthiforms  are  more 
highly  modified.  Eels,  despite  their  elongate  form,  retain  a  caudal 
fin  complete  with  hypural  plates  and  caudal  fin  rays.  To  be  sure, 
the  caudal  fin  is  greatly  reduced  and  shows  much  fusion  of 
elements,  but  it  clearly  exists,  in  larvae  as  well  as  adults  (Fig. 


100 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  5 1 .     Leptocephali  of  (A)  Callechelys  sp.;  (B)  Catesbya  pseudomuraena,  and  (C)  Kaupichlhys  hypoproproides. 


50,  top).  Notacanthiforms  have  no  true  caudal  fin.  In  adult 
notacanthiforms  the  vertebrae  approaching  the  tip  of  the  tail 
become  progressively  less  ossified,  the  centra  being  reduced  to 
rings  around  the  notochord  separated  from  the  neural  and  hemal 
arches.  Finally  the  vertebrae  disappear,  leaving  the  notochord 
freely  exposed  (McDowell,  1973).  There  is  no  hypural  structure, 
and  caudal  fin  rays,  if  they  exist,  are  indistinguishable  from  the 
posterior  anal  fin  rays.  The  notacanthiform  larva  likewise  has 
no  caudal  fin  (Fig.  50,  bottom);  the  notochord  ends  freely,  but 
there  are  two  structures  that  may  be  hypural  elements.  Posterior 
to  these  and  to  the  notochord  is  a  single  filament  that  trails 
freely  for  a  variable  distance  and  might  represent  a  caudal  fin 
ray.  The  anal  fin  occupies  the  short  space  between  the  anus  and 
the  end  of  the  tail  proper  (excluding  the  caudal  filament).  The 
important  point  here  is  that  lumping  notacanthiform  and  an- 
guilliform  larvae  as  pointed-tail  leptocephali  is  unwarranted, 
because  the  caudal  structure  is  quite  different  in  the  two  groups. 
Returning  to  the  diagram  in  Fig.  49,  the  fork-tailed  leptocephali 
can  be  viewed  as  the  primitive  type  of  leptocephalus  present  in 
the  elopid  and  megalopid  branches  and  retained  in  the  Albulidae 
as  well.  Two  pomts  of  transformation  occur,  one  in  the  nota- 
canthiform line  and  one  in  the  anguilliform  line.  The  modifi- 
cations in  each  reflect  modifications  in  the  adults  and  by  them- 
selves are  not  indications  of  a  special  relationship.  Additional 
leptocephali  illustrations  were  prepared  and  are  presented  here 
without  further  comment  (Figs.  51,  52). 


Relationships  between  Elopomorphs 

AND  OTHER  TeLEOSTS 

A  widely  favored  view  today  is  that  the  teleosts  consist  of 
four  major  groups  in  a  cladistic  sense:  the  Osteoglossomorpha, 
Elopomorpha,  Clupeomorpha,  and  Euteleostei  (Greenwood  et 
al.,  1966;  Greenwood,  1973;  Nelson,  1973;  Patterson  and  Ro- 
sen, 1977).  These  groups  are  arranged  in  a  hierarchy  with  the 
Osteoglossomorpha  as  the  sister  group  of  the  remaining  three, 
the  Elopomorpha  as  the  sister  group  of  the  remaining  two,  and 
the  Clupeomorpha  as  the  sister  group  of  the  Euteleostei  (Fig. 
53).  This  classification  is  based  on  a  few  characters  that  are 
thought  to  represent  synapomorphies.  It  is  essential,  therefore, 
to  evaluate  these  characters  carefully,  because  the  whole  clas- 
sification stands  or  falls  on  their  reliability. 

The  Elopomorpha  is  united  by  three  characters:  I )  the  pres- 
ence of  rostral  and  prenasal  ossicles;  2)  the  initial  fusion  of  the 
angular  and  retroarticular  bones  in  the  lower  jaw;  3)  the  presence 
of  a  leptocephalus  larva.  It  is  not  certain  that  eels  have  rostral 
ossicles.  Considering  the  extreme  fusion  that  has  taken  place  in 
the  anterior  extremity  of  the  skull  in  eels,  it  should  not  be 
surprising  if  the  rostral  ossicles  were  lost  as  well.  Still,  it  means 
that  the  character  may  not  be  wholly  inclusive  of  the  group.  The 
second  character,  the  fusion  of  the  angular  and  retroarticular, 
seems  to  hold  for  eels  (Leiby,  1979b)  and  appears  to  be  a  true 
synapomorphy.  That  leaves  the  leptocephalus,  and  its  role  is 


SMITH:  ELOPIFORMES.  NOTACANTHIFORMES  AND  ANGUILLIFORMES 


Fig.  52.     Leptocephali  of  (D)  Bascamchthys  sp.;  (E)  Acromycter  sp.;  (F)  Hildebrandia;  and  (G)  Dysomma  anguillare. 


crucial.  If  it  is  a  synapomorphy,  then  the  congruence  between 
it  and  the  lower-jaw  character  reinforces  the  naturalness  of  the 
Elopomorpha.  Furthermore,  it  is  a  more  complex  character,  thus 
less  likely  to  show  parallelism  than  a  simple  process  like  the 
fusion  of  two  bones  in  the  lower  jaw  (which,  indeed,  has  hap- 
pened independently  in  some  osteoglossomorphs). 


To  explore  this  matter,  we  must  first  establish  clearly  what  a 
leptocephalus  is.  If,  as  some  have  maintained,  it  were  simply  a 
ribbon-like  larva  with  a  posterior  anus  and  a  dorsal  fin  that 
moves  forward  at  metamorphosis,  then  it  would  tell  us  little 
about  elopomorph  phylogeny.  Many  lower  teleosts  have  such 
larvae.  A  leptocephalus  is  considerably  more  than  this,  however. 


102 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


0STE0GL05  - 
SOMORPHA 


ELOPO- 
MORPHA 


CLUPEO- 
MORPHA 


EUTEL  - 
EOSTEI 


Fig.  53.     Hypothesis  of  relationships  between  major  groups  of  Tele- 

ostei. 


The  unique  structure  of  a  leptocephalus  can  be  appreciated  best 
in  cross  section  (Fig.  54,  left).  The  viscera  lie  along  the  ventral 
margin  in  a  narrow  strand.  The  notochord,  dorsal  nerve  cord, 
and  dorsal  aorta  lie  together  in  the  longitudinal  axis  of  the  body 
about  midway  between  the  dorsal  and  ventral  margins.  The 
myomeres  form  a  thin  layer  on  the  outside.  Filling  the  rest  of 
the  interior  of  the  body  is  an  acellular  mucinous  material  bound- 
ed by  a  continuous  layer  of  epithelial  cells.  The  mucinous  pouch 
separates  the  viscera,  the  notochord  and  the  two  sides  of  the 
body  musculature  from  each  other  and  gives  form  and  rigidity 
to  the  body.  The  characteristic  shrinkage  of  the  leptocephalus 
at  metamorphosis  is  due  to  the  loss  (presumably  by  resorption) 
of  the  internal  mucinous  material.  A  typical  clupeid  larva  such 
as  Elnaneus  teres  (Fig.  54,  right)  is  constructed  much  differently. 
Here  there  is  no  mucinous  pouch.  The  notochord  occupies  a 
large  part  of  the  cross-sectional  area  and  is  surrounded  imme- 
diately by  the  thick  axial  musculature  to  form  a  solid,  compact 
structure.  The  viscera  lie  immediately  below  the  dorsal  aorta. 
Leptocephali  have  a  small  head  and  a  set  of  long,  sharp  teeth 
whose  function  is  uncertain,  since  leptocephali  do  not  seem  to 
be  predatory.  The  basic  structure  of  a  leptocephalus  is  the  same 
whether  it  is  an  elopiform,  notacanthiform  or  anguilliform.  A 
leptocephalus  larva  is  known  for  every  family  of  elopomorphs 
except  the  rare,  monotypic  Myrocongridae,  so  the  character 
seems  entirely  inclusive  of  the  group.  Nothing  even  remotely 
comparable  is  found  outside  the  Elopomorpha. 


Fig.  54.  Cross  section  through  the  bodies  of  a  leptocephalus  (Meg- 
alops  allanlictis)  (left)  and  a  clupeid  larva  (Elrumeus  teres)  (right).  DA, 
dorsal  aorta;  NC,  notochord;  SC,  spinal  cord. 


The  leptocephalus,  then,  must  be  considered  a  true  synapo- 
morphy  and  powerful  evidence  in  favor  of  the  monophyly  of 
the  Elopomorpha.  Perhaps  nowhere  else  in  fish  systemalics  have 
larval  stages  played  a  more  important  role. 

The  Marine  Biomedical  Institute,  The  University  of  Texas 
Medical  Branch  at  Galveston,  200  University  Boule- 
vard, Galveston,  Texas  77550. 


Ophichthidae:  Development  and  Relationships 


M.  M.  Leiby 


THE  family  Ophichthidae,  comprising  approximately  250 
nominal  species  and  53  recognized  genera,  is  arranged  in 
six  tribes  and  two  subfamilies  (McCosker.  1977)  (Fig.  55).  The 
subfamilies,  Myrophinae  and  Ophichthinae,  are  separated  by  a 


number  of  characters.  All  adult  Myrophinae  have  a  well-de- 
veloped caudal  fin  which  is  continuous  with  the  dorsal  and  anal 
fin.  Adult  Ophichthinae,  except  for  Echelus  in  the  tribe  Ophich- 
thini  and  Lcptenchelys  in  the  tribe  Bascanichthyini,  lack  a  caudal 


LEIBY:  OPHICHTHIDAE 


103 


Family    Ophichthidoe 


Subfomily     Myrophmoe 


Sub  fomily     Ophichthmae 


I— II 


Ttib« 

Myrophin 


Tribe 
Ophichthini 


Tribe 
Sphogebranchir 


Tribe 
Basconichthyini 


Tribe 
Callechelyini 


£ 

>■ 

C 

£ 

o 
■6 

3 

w 

£ 

3 

m 

■^ 

t 

o 

:^ 

o 

e 

V 

b 

■a 

3 

c 

o 

c 
at 

CT 

i 

3 
tl 

W 

3 
3 

c 

i 

3 

E 

a 
o 

o 

E 

o 

o 

o 

« 

w 

• 

3 

u 

a. 

u 

a> 

(D 

Z 

3 

<3 

s 

VI 

o 

tn 

O 

O 

a. 

u 

UJ 

^ 

-* 

o      =      •) 
w      ^      tt 


-o     ^      •»      ly 


Callechelyin 
Ancestor 


Bosconichf  hy  s-like 
Ancestor 


Moderately    Specolized 
Optiictittiin-like    Ancestor 


oderotely    Speciolized 
Ophictttttin-like    Ancestor 


Moltfolioptiis     Or 

E  vips-  like     Ancestor 


Ouossiremus-like    Ancestor 


Ance  strol     Myroph 


Ancestrol     Optlictlttlin 


=  Tribe     BenThenchelyin 


Congrid-like    Ancestor 


Fig.  55.     Hypothesized  relationships  of  the  subfamilies  and  genera  of  the  eel  family  Ophichthidae. 


tin  having  instead  a  hardened  tail  tip  with,  at  most,  a  few  ru- 
dimentary caudal  rays  embedded  in  the  flesh  of  the  tail.  The 
monotypic  genus  Leptenche/ys.  known  only  from  the  1 1 5  mm 
type  specimen,  has  caudal-fin  rays,  but  they  are  weakly  devel- 
oped compared  to  those  of  a  myrophin  (McCosker,  1977).  Since 
all  ophichthid  larvae  have  a  well-developed  caudal  fin  until  the 
onset  of  metamorphosis,  the  presence  of  weakly  developed  rays 
in  the  only  known  specimen  of  Leplenchelys  may  be  an  anomaly 
resulting  from  incomplete  resorption  during  metamorphosis. 
The  well  developed  caudal  fin  of  Echelus  has  prompted  most 
earlier  authors  to  place  it  in  the  family  Echelidae  (=Ophichthi- 
dae,  in  part)  or  to  ally  it  with  the  subfamily  Myrophinae  (e.g.. 
Dean,  1972;  Blache,  1977);  however,  the  osteology  of  the  genus 
(McCosker,  1977)  and  its  larval  morphology  (Blache,  1977:  Figs. 
72  and  74)  clearly  place  Echelus  in  the  subfamily  Ophichthinae 
and  ally  it  with  the  tribe  Ophichthini. 

Adult  Myrophinae  have  four  to  seven  branchiostegal  rays 
attached  to  the  epihyal  and  ceratohyal  and  1 3-45  free  (unat- 
tached) branchiostegal  rays  which  originate  posterior  to  the  tips 


of  the  epihyals.  Most  adult  Ophichthinae  have  the  majority  of 
their  branchiostegal  rays  attached  to  the  epihyal  and  ceratohyal. 
The  free  branchiostegal  rays  of  all  Ophichthinae  originate  an- 
terior to  the  tips  of  the  epihyals. 

The  ceratohyal,  epihyal  and  hypohyal  of  both  the  Myrophinae 
and  the  Ophichthinae  originate  from  a  single  block  of  cartilage 
with  the  first  center  of  ossification  being  a  thin  strip  along  the 
lateral  face  of  the  cartilage  (Leiby,  1979a,  b;  1981).  When  de- 
velopment is  complete,  the  ceratohyal  of  the  Myrophinae  is  a 
simple  bone  which  terminates  about  midpoint  along  the  lateral 
face  ofthe  epihyal  (Dean,  1972;  McCosker,  1977;  Leiby,  1979b). 
The  ceratohyal  ofthe  Ophichthinae  has  a  slender,  elongate  distal 
portion  which  terminates  about  midpoint  along  the  lateral  face 
of  the  epihyal  and  a  medial  portion  which  is  attached  to  the 
proximal  end  ofthe  epihyal  by  a  cartilage  (McCosker,  1977; 
Leiby,  1981). 

The  urohyal  ofthe  Myrophinae  and  Ophichthinae  ossifies  in 
a  bifurcated  medial  ligament  which  is  attached  to  the  developing 
hypohyals.  In  the  Myrophinae,  the  urohyal  is  generally  limited 


104 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


L  L- 


Fig.  56.  (Upper.)  Anterior  portion  of  Myrophis  punclalus  larva  depicting  typical  myrophin  gut  morphology.  Abbreviations:  LL|_j,  liver  lobes 
1-3;  GB,  gall  bladder.  (Lower.)  Anterior  portion  of  Neenchelvs  microlrelus  larva  depicting  gut  morphology.  Abbreviations:  LL,.,,  liver  lobes  1- 
2;  GB.  gall  bladder. 


to  a  basal  plate  which  ossifies  from  the  hypohyal  to  the  bifur- 
cation of  the  ligament.  The  urohyal  of  the  Ophichthinae  gen- 
erally ossifies  to  include  a  spike  which  extends  well  posterior  to 
the  area  of  the  bifurcation. 

The  gill  openings  of  the  Myrophinae  are  midlateral  and  con- 
stricted. Ophichthine  gill  openings  are  variable  in  position,  their 
major  axis  ranging  from  midlateral  to  ventral,  but  always  un- 
constricted. 

Leptocephali  belonging  to  five  of  the  nine  myrophin  genera 


have  been  identified.  Larvae  of  four  of  these  five  genera  have 
three  unconnected  liver  lobes  with  the  gall  bladder  on  the  third 
lobe  (Fig.  56-upper).  Larvae  of  the  fifth  genus,  Neenchelvs.  which 
differ  trenchantly  from  all  other  ophichthid  larvae,  have  two 
unconnected  liver  lobes  with  the  gall  bladder  on  the  second  lobe 
(Fig.  56-lower).  Leptocephali  belonging  to  twenty  of  the  forty- 
four  ophichthin  genera  have  been  identified.  All  twenty  of  these 
genera  have  two  connected  liver  lobes  with  the  gall  bladder  on 
the  second  lobe  (Fig.  57-upper). 


LEIBY:  OPHICHTHIDAE 


105 


5mm 


Fig.  57.  (Upper.)  Anterior  portion  of  Ophichthus  gomesi  larva  depicting  typical  ophichthin  gut  morphology.  Abbreviations:  LL|_2.  liver  lobes 
1-2;  GB,  gall  bladder.  (Lower.)  Middle  portion  of  Ophichthus  gomesi  larva  depicting  position  of  nephros  relative  to  anus  in  some  members  of 
the  Ophichthus  lineage  of  the  tribe  Ophichthini.  Abbreviations:  N,  nephros;  A,  anus. 


The  dorsal  fin  of  known  myrophin  lai^ae  has  well-developed 
pterygiophores  and  fin  rays  prior  to  the  onset  of  metamorphosis 
and  migrates  only  a  few  myomeres  anteriorly  (4-6)  during  meta- 
morphosis to  reach  its  adult  position.  The  dorsal  fin  of  known 
ophichthin  larvae,  which  is  weakly  developed  having  only  pte- 
rygiophores and  rudimentary  rays  in  its  anterior  portion  prior 
to  metamorphosis,  must  migrate  5-20  myomeres  anteriorly  dur- 
ing metamorphosis  in  species  having  the  dorsal  fin  antenor  to 
the  branchial  aperture  as  adults,  and  20-50  myomeres  in  species 
having  the  dorsal  fin  posterior  to  the  branchial  aperture  as  adults, 
and  is  resorbed  m  species  which  are  finless  as  adults. 

The  subfamily  Myrophinae  contains  two  tribes  (sensu 
McCosker,  1977),  the  Myrophini  and  the  Benthenchelyini.  Os- 
teological  examination  of  adults  in  the  tribe  Myrophini  indi- 
cated the  presence  of  three  lineages  consisting  of  Pseudomyro- 
phis  and  Neenchelys;  Myrophis,  Ahlia.  and  a  currently 
undescribed  genus;  and  Muraemchlhys  and  its  allies.  The  My- 
rophis and  Muraemchlhys  lineages  share  a  common  ancestor 


(Fig.  55).  Larval  morphology  oi  Myrophis,  Ahlia  and  Muraen- 
ichthys  is  very  similar  and  supports  the  determination  of  a  close 
relationship  for  the  two  lineages.  Larvae  of  these  three  genera 
have  three  unconnected  liver  lobes,  similar  gut  and  opistho- 
nephros  morphology,  and  similar  body  length  to  depth  ratios 
(Fahay  and  Obenchain,  1978;  Leiby,  1979b;  Ochiai  and  No- 
zawa,  1980).  Pseudomyrophis  larvae  have  three  unconnected 
liver  lobes  and  a  body  length  to  depth  ratio  which  is  similar  to 
that  of  the  Myrophis  and  Miiraenichthys  lineages,  but  gut  and 
opisthonephros  morphology  is  significantly  different  from  that 
seen  in  the  Myrophis  and  Muraemchthys  lineages  and  supports 
the  conclusion  drawn  from  adult  data  that  the  Pseudomyrophis 
lineage  is  distinct  from  the  Myrophis  and  Muraenichthys  lin- 
eages. Nelson  ( 1 966a)  suggested  that  Pseudomyrophis  micro- 
pinna,  the  type  of  the  genus,  was  congeneric  with  Neenchelys 
hiutendijki,  but  that  P.  nimius,  while  belonging  to  the  same 
lineage,  was  separable  at  the  generic  level  from  either  of  the 
other  two  species.  Dean  (1972)  also  felt  that  the  differences 


106 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


between  P.  micropinna  and  P.  nimiiis  warranted  a  separate  ge- 
nus for  P.  nimius.  However,  McCosker  (1977,  1982)  demon- 
strated that  Pseudomyrophis  and  Neenchelys  are  both  valid  gen- 
era and  that  P.  micropinna,  P.  nimius,  P.  atlanlicus  and  an 
undescribed  Pseudomyrophis  from  the  eastern  Pacific  are  con- 
generic. Dean  (1972)  indicated  that  Myrophis  frio  properly  be- 
longs in  the  Pseudomyrophis  lineage.  Evidence  from  larval  mor- 
phology supports  McCosker's  (1977,  1982)  recognition  of 
Pseudomyrophis  and  Neenchelys  as  valid  genera,  and  supports 
the  recognition  of  P.  micropinna,  P.  nimius,  P.  atlanlicus,  the 
undescribed  Pseudomyrophis  from  the  eastern  Pacific,  two  un- 
described Pseudomyrophis  known  only  from  their  larvae  in  the 
western  Atlantic,  one  undescribed  Pseudomyrophis  from  the 
eastern  Atlantic  known  only  from  its  larva  and  erroneously 
identified  as  P.  nimius  (Blache,  1977),  and  Myrophis  frio  as 
congeneric.  Pseudomyrophis  larvae  are  readily  distinguishable 
from  all  other  ophichthid  larvae  by  a  combination  of  the  fol- 
lowing characters:  three  unconnected  liver  lobes,  undulating  gut 
and  nephros,  characteristic  head  shape,  and  pigmentation 
(Blache,  1977;  Leiby,  in  press  a).  Neenchelys  larvae  differ  tren- 
chantly from  Pseudomyrophis  larvae  in  having  two,  rather  than 
three,  unconnected  liver  lobes,  a  gut  lacking  the  marked  un- 
dulations seen  in  Pseudomyrophis  larvae,  and  a  much  deeper 
body  than  any  other  known  ophichthid  (Castle,  1 980;  this  paper. 
Fig.  56-lower).  Studies  of  adult  Pseudomyrophis  and  Neenchelys 
have  clearly  demonstrated  that  the  two  genera  are  more  closely 
related  to  each  other  than  either  is  to  any  other  genus  ( McCosker, 
1977,  1982).  In  the  light  of  this  information,  the  most  parsi- 
monious interpretation  of  the  data  on  the  larval  morphology  of 
the  two  genera  is  that  Neenchelys  was  derived  from  Pseudo- 
myrophis or  a  Pseudomyrophis-hke  ancestor.  Pseudomyrophis 
and  all  other  known  myrophin  larvae  except  Neenchelys  have 
three  unconnected  liver  lobes  and  similar  body  length  to  depth 
ratios.  It  seems  likely,  therefore,  that  larvae  of  the  ancestral 
myrophin  also  had  three  unconnected  liver  lobes  and  a  similar 
body  length  to  depth  ratio.  Neenchelys  larval  morphology  can 
be  easily  derived  from  this  proposed  ancestral  larval  morphol- 
ogy by  significantly  deepening  the  body  and  foreshortening  the 
gut  so  that  one  liver  lobe  is  lost.  Derivation  of  Pseudomyrophis 
larval  morphology  from  a  Neenchelys-Wkc  ancestor  requires  a 
change  from  the  ancestral  larval  morphology  body  plan  to  the 
Neenchelys  larval  body  plan  and  a  later  re-emergence  of  the 
ancestral  larval  myrophin  body  plan  in  Pseudomyrophis. 

Benthenchelys  cartieri.  a  highly  specialized  pelagic  eel  (Castle, 
1972)  is  the  sole  member  of  the  tribe  Benthenchelyini.  The 
larvae  of  this  species  have  not  yet  been  described,  but  based  on 
the  hypothesized  evolutionary  history  of  the  Ophichthidae  (Fig. 
55),  it  seems  likely  that  the  larvae  of  5.  cartieri  will  have  three 
unconnected  liver  lobes,  a  well-developed  dorsal  fin  which  mi- 
grates little  during  metamorphosis,  and  a  body  length  to  depth 
ratio  that  is  typical  of  the  Ophichthidae.  Discovery  of  these 
larvae  should  help  clarify  relationships  within  the  Myrophinae. 

The  subfamily  Ophichthinae  contains  four  tribes  (sensu 
McCosker,  1977);  the  Ophichthini,  Sphagebranchini,  Bascan- 
ichthyini  and  Callechelyini.  The  tribe  Ophichthini  lies  at  the 
evolutionary  base  of  the  subfamily  Ophichthinae,  and  contains 
the  most  primitive,  least  specialized  members  of  the  subfamily. 
The  ancestral  ophichthin  was  probably  Ophichthus-hke.  The 
tribe  Ophichthini,  which  contains  two  lineages,  and  the  tribe 
Sphagebranchini  can  be  easily  derived  from  the  generalized 
ophichthin  character  states  which  are  represented  in  the  genus 


Ophichihus  (sensu  McCosker,  1977).  One  lineage  in  the  tribe 
Ophichthini  appears  to  be  directly  derived  from  the  generalized 
Ophichthus  condition.  The  genus  Echelus  has  been  represented 
as  belonging  to  its  own  unique  lineage  in  the  Ophichthinae  and 
has  been  considered  the  most  primitive  member  of  the  tribe 
Ophichthini  because  in  addition  to  having  all  the  primitive 
characters  of  its  closest  relative  Ophichthus.  it  possesses  a  well- 
developed  caudal  fin.  A  re-examination  of  adult  Echelus  char- 
acters in  conjunction  with  the  larval  characters  oi  Echelus  sug- 
gests, however,  that  Echelus  belongs  to  the  Ophichthus  lineage 
and  that  the  caudal  fin  of  Echelus  is  either  a  case  of  character 
reversal  or  paedomorphosis  which  resulted  in  Echelus  retaining 
the  larval  caudal  fin  rather  than  losing  it,  as  is  apparently  the 
case  in  all  other  members  of  the  Ophichthinae.  In  addition  to 
the  generalized  genera  Echelus.  Ophichthus,  and  Ophisurus,  the 
Ophichthus  lineage  contains  two  groups  of  specialized  genera 
which  are  closely  tied  to  Ophichthus  by  a  nearly  continuous 
character  series.  The  Pisodonophis-Myrichthys-Cirrhimuraena 
group  differ  from  the  basic  Ophichthus  body  plan  by  having  an 
increased  number  of  branchiostegals,  multiserial  dentition,  and 
individual  sp)ecializations  found  in  each  genus.  The  second  group, 
containing  Mystriophis  and  seven  allied  genera,  are  specialized 
for  the  capture  of  large  active  prey  by  having  a  strengthened 
suspensorium  and  enlarged  dentition.  The  close  relationship  of 
this  group  to  Ophichthus  is  emphasized  by  similar  adaptations 
in  some  species  of  Ophichthus  (McCosker,  1977).  The  close 
relationship  of  the  Ophichthus  lineage  is  further  emphasized  by 
the  unique  positioning  of  the  nephros  relative  to  the  anus  found 
in  many  members  of  this  lineage.  Larvae  from  seven  of  the 
fourteen  genera  in  the  Ophichthus  lineage  have  been  identified. 
While  there  is  considerable  inter-  and  intrageneric  variability 
in  the  general  morphology  of  these  larvae,  five  of  the  seven 
genera  (Echelus.  Ophichthus,  Ophisurus,  Echiophis,  and  Apla- 
tophis)  are  generally  characterized  by  having  larvae  with  a  neph- 
ros which  terminates  4-14  myomeres  anterior  to  the  anus  on 
the  next  to  last  gut  loop  or  between  the  last  and  next  to  last  gut 
loop  (Fig.  57-lower).  This  condition  has  not  been  observed  in 
any  genera  of  the  Ophichthinae  outside  of  the  Ophichthus  lin- 
eage of  the  Ophichthini.  The  larvae  of  Myrichthys,  one  of  the 
specialized  genera  in  the  Ophichthus  lineage,  has  a  nephros  which 
terminates  above  or  just  anterior  to  the  anus  (Leiby,  in  press 
a).  Blache  (1977)  identified  a  series  of  larvae  as  Brachysomophis 
atlanlicus.  This  series  of  larvae  differs  from  the  larvae  of  the 
closely  related  genus  Aplalophis  in  having  the  nephros  termi- 
nating above  or  just  anterior  to  the  anus.  Larvae  of  the  western 
Pacific  species  of  Brachysomophis  have  not  yet  been  identified. 
Consequently,  it  is  unknown  whether  this  nephric  position  is  a 
secondarily  derived  character  of  the  genus  Brachysomophis  or 
whether  it  is  limited  to  the  eastern  Atlantic  species  B.  atlanlicus. 
The  other  lineage  to  arise  from  the  generalized  Ophichthus- 
hke  ancestor  contains  eight  genera  including  Quassiremus  and 
Malvoliophis  (Fig.  55),  which  are  characterized  by  various  re- 
ductions and  modifications  of  the  generalized  Ophichthus-Vike 
condition  such  as  reduced  gill  arches,  cephalic  lateralis  systems, 
and  pectoral  fins.  This  lineage  probably  gave  rise  to  the  Sphag- 
ebranchini and  subsequent  lineages  by  continued  modification, 
reduction,  and  specialization  of  the  ophichthin  condition 
(McCosker,  1977).  The  larvae  of  the  Quassiremus- Malvoliophis 
lineage  are  virtually  unknown.  Leiby  (in  press)  tentatively 
identified  three  larvae  as  Quassiremus  produclus,  but  no  other 
larvae  from  this  lineage  have  been  identified.  There  is  a  natural 


LEIBY:  OPHICHTHIDAE 


107 


progression  in  larval  morphology  from  some  Ophichthus  spp. 
through  Quassiremus  morphology  to  sphagebranchin  mor- 
phology which  tends  to  support  McCosker's  (1977)  hypothesis 
that  the  other  ophichthin  lineages  arose  through  modification, 
reduction,  and  specialization  of  the  ancestral  Ophichthus-like 
condition.  Quassiremus  larvae  look  much  like  the  larvae  of 
some  Ophichthus  spp.,  but  differ  in  having  the  nephros  termi- 
nate over  or  just  anterior  to  the  anus,  and  in  having  reduced 
gill  arches. 

The  tribe  Sphagebranchini  is  distinguished  from  the  other 
tribes  of  the  Ophichthinae  by  a  combination  of  the  following 
adult  characters:  the  pectoral  girdle  is  reduced;  the  pectoral  fin 
is  absent;  the  gill  openings  are  low  to  entirely  ventral;  the  neu- 
rocranium  is  elongate  (neurocranium  depth  going  4  or  more 
times  into  its  length),  generally  depressed,  and  truncate  poste- 
riorly; the  gill  arches  are  generally  much  reduced;  the  body  is 
equal  to  or  shorter  than  the  tail;  the  tail  tip  is  sharply  pointed; 
and,  the  cephalic  lateralis  system  is  generally  better  developed 
than  in  other  tribes  (McCosker,  1977).  Larval  characters  which 
distinguish  this  tribe  from  other  tribes  in  the  Ophichthinae  or 
which  distinguish  lineages  within  the  tribe,  are  reflections  of  the 
adult  characters  (e.g.,  reduced  gill  arches,  short  gut,  dorsal  fin 
origin)  (Leiby,  1982).  As  yet,  there  are  no  independent  larval 
characters  which  confirm  the  monophyletic  origin  of  this  tribe 
or  which  confirm  the  proposed  lineages  within  the  tribe,  al- 
though the  larval  morphology  is  similar  to,  and  sometimes  dif- 
ficult to  distinguish  from,  the  larval  morphology  of  some  Oph- 
ichthini  and  is  consistent  with  the  hypothesis  of  modification, 
reduction,  and  specialization  of  the  ancestral  ophichthin  con- 
dition which  has  been  proposed  based  on  adult  data. 

The  tribe  Bascanichthyini,  apparently  derived  from  a  mod- 
erately specialized  ophichthin-like  ancestor  (McCosker,  1977), 
is  distinguishable  from  the  other  tribes  of  the  Ophichthinae  by 
a  combination  of  the  following  adult  characters:  the  body  is 
equal  to,  or  longer  than  the  tail;  the  gill  openings  are  low  lateral 
and  crescentic,  never  entirely  ventral;  dorsal-fin  origin  is  on  the 
head  in  most  genera;  the  pectoral  fin  is  reduced  or  absent;  the 
cephalic  lateralis  system  is  reduced;  and,  the  gill  arches  are 
generally  much  reduced  (McCosker,  1977).  The  genus  Dalophis 
is  provisionally  placed  in  the  Bascanichthyini  despite  its  pos- 
session of  a  gill  arch  skeleton  and  a  body  length  which  are  more 
ophichthin  than  bascanichthyin,  due  to  its  reductions,  general 
cephalic  appearance  and  several  osteological  characters  (Mc- 
Cosker, 1977).  If  this  placement  oi  Dalophis  is  correct,  it  seems 
likely  that  the  ancestral  bascanichthyin  was  similar  in  appear- 
ance to  Dalophis.  Larval  characters  which  distinguish  this  tribe 
from  other  tribes  in  the  Ophichthinae  are  reflections  of  adult 
characters  (e.g.,  reduced  gill  arches,  relatively  long  gut  and  opis- 
thonephros,  and  dorsal-fin  origin).  Larvae  have  been  identified 
from  each  of  the  three  proposed  bascanichthyin  lineages  [e.g., 
Dalophis  (Blache,  1 977;  Palomera  and  Fortuno,  1981),  Bascan- 
ichth\'s(B\?Lc\\e,  \971\  Leiby,  1981),  Gordiichthys (Leiby.  in  press), 
Caralophia  (Leiby,  in  press)],  but  there  are  currently  no  clear 
larval  characters  which  are  useful  for  elucidating  relationships 
within  the  Bascanichthyini.  With  one  exception,  all  of  the  larvae 
assigned  to  the  Bascanichthyini  are  characterized  by  extremely 
low  to  moderately  developed  gut  loops  and,  except  for  gut  length, 
nephros  length  and  dorsal-fin  origin,  look  much  like  larvae  of 
the  Sphagebranchini.  One  larval  form  which  cannot  yet  be  as- 
signed to  a  genus,  has  tentatively  been  placed  in  the  Bascani- 
chthyini based  on  gill  arch  and  caudal  osteology  although  its 


gut  morphology  is  more  like  some  Callechelyini  than  Bascani- 
chthyini (Leiby,  in  press).  Discovery  of  the  adults  of  this  species 
may  help  clarify  relationships  within  the  Bascanichthyini. 

The  tribe  Callechelyini  is  apparently  derived  from  a  bascan- 
ichthyin-like  ancestor.  Adults  of  this  tribe  are  distinguished  by 
a  short  neurocranium  (neurocranium  depth  >  33%  of  its  length); 
the  dorsal-fin  origin  on  the  head  or  nape;  the  body  longer  than 
the  tail;  absence  of  a  pectoral  fin;  low  lateral  to  entirely  ventral 
anteriorly  convergent  gill  openings;  reduced  gill  arches;  reduced 
cephalic  lateralis  system;  laterally  compressed  body;  small  eyes; 
and,  a  stout  hyoid  (McCosker,  1977).  Larvae  of  three  of  the  five 
known  Callechelyin  genera  have  been  identified  (Leiby,  1984) 
and  are  readily  distinguishable  from  larvae  of  the  other  ophich- 
thin tribes.  Callechelyin  larvae  are  characterized  by  moderate 
to  pronounced  gut  loops;  variable  but  distinctive  pigmentation 
(see  Leiby,  in  press  b,  for  full  descriptions);  anterior  dorsal-fin 
origin;  nephric  myomeres  more  than  56%  of  total  myomeres;  a 
distinct  fourth  hypobranchial  which  may  be  separate  from  or 
united  with  a  reduced  fifth  ceratobranchial  (a  remnant  of  the 
fourth  hypobranchial  united  with  a  reduced  fifth  ceratobranchial 
may  occasionally  be  found  in  gill  arches  of  larval  Sphagebran- 
chini and  Bascanichthyini;  a  distinct  fourth  hypobranchial  is 
found  in  some  larval  Ophichthini,  but,  when  present,  is  united 
with  a  well  developed  fifth  ceratobranchial);  and  usually  two 
hypurals  rather  than  the  three  seen  in  other  ophichthids. 
McCosker  and  Rosenblatt  (1972)  and  McCosker  (1977)  recog- 
nized the  presence  of  subgeneric  lines  in  the  genus  Callechelys. 
Evidence  from  larval  morphology  confirms  the  presence  of  two 
subgeneric  lineages  in  Callechelys  (Leiby,  1984).  Adults  of  one 
subgenus  have  a  split  urohyal  and  two  rod-shaped  elements  in 
the  pectoral  girdle.  The  larvae  of  this  subgenus  have  pronounced 
gut  loops;  the  fourth  hypobranchial  free  from  the  fifth  cerato- 
branchial; most  or  all  of  the  myosepta  without  pigment;  most 
or  all  of  the  anal  pterygiophores  without  pigment;  no  pigment 
on  the  esophagus;  pigment  on  the  dorsal  surface  of  each  gut 
loop  but  no  pigment  between  gut  loops;  pronounced,  round 
pigment  patches  in  the  body  wall  lateral  to  each  gut  loop;  and, 
three  to  five  pronounced,  circular  postanal  pigment  patches  which 
consist  of  subcutaneous  and  body-wall  pigment.  Adults  of  the 
second  subgenus  have  a  simple  urohyal  and  one  or  two  rod- 
shaped  elements  in  the  pectoral  girdle.  The  larvae  of  this  sub- 
genus have  moderate  gut  loops;  the  fourth  hypobranchial  united 
with  the  fifth  ceratobranchial;  dark  pigment  every  third  to  elev- 
enth myoseptum,  or  light  pigment  on  every  myoseptum;  round 
or  saddle-shaped  patches  of  pigment  in  the  body  wall  on  the 
ventral  margin  of  the  tail  extending  onto  the  anal  pterygio- 
phores, or  pigment  on  every  anal  pterygiophore  but  none  in  the 
ventral  body  wall;  pigment  on  the  esophagus,  on  the  dorsal 
surface  of  each  gut  loop,  and  between  each  gut  loop;  occasionally 
some  body-wall  pigment  lateral  to  each  gut  loop;  four  to  seven 
irregular,  subcutaneous  pigment  patches  on  the  tail,  usually  not 
flanked  by  body-wall  pigment. 


Relationships  to  other  taxa 

The  family  Ophichthidae  is  generally  considered  to  be  a  co- 
hesive group  which  is  the  sole  member  of  the  superfamily  Oph- 
ichthoidea.  The  unique  nature  of  ophichthid  larvae  supports 
this  allocation.  Most  workers  (e.g.,  Gosline.  1951;  Nelson,  1966b; 
McCosker,  1 977)  consider  the  Ophichthidae  to  be  a  specialized 
offshoot  of  the  Congridae,  although  Dean  (1972)  decried  the 


108 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


value  of  the  characters  used  to  associate  the  Ophichthoidea  with 
the  Congroidea  and  implied  that  the  Ophichthidae  could  just 
as  easily  be  a  specialized  offshoot  of  the  Anguilloidea.  While 
the  only  known  larvae  which  could  be  confused  with  the 
Ophichthidae  are  members  of  the  family  Congridae  (e.g.,  Ac- 
romycter  larvae  have  pronounced  gut  loops.  Nystactichthys  lar- 
vae have  a  gut  which  expands  abruptly  between  the  esophagus 
and  intestine),  there  are  no  known  larval  characters  which  un- 


equivocally establish  the  evolutionary  relationships  of  the 
Ophichthidae.  Careful  osteological  studies  of  ontogenetic  series 
of  eel  larvae  from  the  various  families  may  eventually  clear  the 
currently  clouded  picture. 

Department  of  Natural  Resources,  Marine  Resources 
Laboratory,  100  Eighth  Avenue  Southeast,  Saint  Pe- 
tersburg, Florida  33701. 


Clupeiformes:  Development  and  Relationships 
M.  F.  McGowAN  AND  F.  H.  Berry 


THE  order  Clupeiformes  contains  four  families  of  fishes:  the 
herrings,  Clupeidae;  the  anchovies.  Engraulidae;  the  wolf- 
herrings,  Chirocentridae;  and  the  denticle  herring,  Denticipiti- 
dae  (Nelson,  1976).  Denticeps  clupeoides.  the  monotypic  den- 
ticipitid,  occurs  in  freshwater  in  southwest  Nigeria  (Clausen, 
1959).  Two  species  of  Chirocentrus  occur  in  marine  waters  of 
the  Indo-Pacific  region  from  the  Red  Sea  to  the  western  Pacific 
(Whitehead,  1972).  They  are  unusual  among  the  Clupeiformes 
in  that  they  are  piscivorous.  The  herrings  and  anchovies  are,  in 
general,  small  schooling  planktivores  of  marine  coastal  waters. 
The  Indo-Pacific  shad,  Tenualosa  reevesii.  reaches  509  mm 
standard  length;  the  West  African  riverine  species,  Thrattidion 
noctivagns  and  Sierrathrissa  leonensis.  are  mature  at  18  mm 
(Wongratana,  1980).  There  are  192  species  of  clupeids  in  62 
genera  and  122  species  of  engraulids  in  16  genera  (Table  24) 
based  on  our  review  of  the  literature.  Herrings  and  anchovies 
are  most  speciose  in  the  tropics,  and  individual  species  are  most 
abundant  in  cold  temperate  regions  and  eastern  boundary  cur- 
rents (Longhurst,  1971).  Some  are  found  in  fresh  or  brackish 
water;  some  are  anadromous.  They  support  major  fisheries 
worldwide.  Their  biology  has  been  reviewed  most  recently  by 
Blaxter  and  Hunter  (1982). 

Development 

The  eggs  and  the  larvae  of  Chirocentrus  are  known  (Delsman, 
1923,  1930b);  the  egg  and  larva  oi  Denticeps  are  unknown;  and 
the  eggs  or  larvae  of  at  least  one  species  in  a  genus  have  been 
described  for  approximately  one-half  the  genera  of  herrings  and 
anchovies  but  for  only  one-third  of  all  species.  Ontogenetic 
stages  of  herrings  and  anchovies  are  best  known  for  species  of 
commercial  interest  or  potential  commercial  interest  in  regions 
with  low  clupeoid  diversity  such  as  the  northeast  Atlantic  (e.g., 
Chtpea,  Sprattus.  Sardina.  Engraulis)  and  the  California  current 
(e.g.,  Sardinops,  Etrumeus.  Engraulis).  The  ontogeny  of  mor- 
phology and  behavior,  and  the  requirements  for  growth  and 
survival  of  the  herring,  Cliipea  harengus,  and  the  anchovy,  En- 
graulis mordax,  are  well  known  (Blaxter  and  Hunter,  1982). 
Very  little  detailed  information  exists  for  clupeids  from  species- 
rich  areas,  especially  western  African  freshwaters  and  the  New 
World  tropics.  Descriptive  taxonomy  is  still  needed  in  these 
areas.  Table  25  lists  the  clupeiform  fishes  for  which  we  found 
some  information  about  eggs  and  larvae. 

Published  descriptions  of  clupeoid  eggs  and  larvae  may  not 


be  adequate  for  systematic  studies  for  a  variety  of  reasons.  When 
there  are  few  species  in  an  area  with  which  to  confuse  the  de- 
scribed species,  only  the  key  identifying  features  are  described. 
When  eggs  are  hatched  but  the  larvae  are  not  reared  to  meta- 
morphosis, usually  an  atypical  starving  early  larva  is  described. 
When  a  well-described  series  of  field-caught  larvae  is  compared 
with  a  laboratory-reared  series  there  may  be  differences  in  pig- 
mentation and  size  at  a  particular  stage  of  development  due  to 
the  rearing  environment.  Future  descnptions  should  describe 
the  eggs  and  yolk-sac  larvae  thoroughly  because  these  stages 
have  characters  other  than  those  such  as  meristics  which,  be- 
cause they  are  shared  with  the  adults,  are  redundant  for  system- 
atic purposes.  Future  descriptions  should  also  try  to  describe 
the  development  of  characters  which  are  of  phylogenetic  im- 
portance in  adult-based  classifications  because  the  ontogenetic 
transformation  of  a  character  provides  information  about  the 
polarity  of  states  of  that  character  (Nelson,  1978). 

Because  the  eggs  and  larvae  of  so  many  clupeiform  genera  are 
undescribed  and  because  existing  descriptions  vary  in  com- 
pleteness, it  is  premature  to  attempt  a  phylogenetic  classification 
of  the  Clupeiformes  based  on  early  life  history  stages.  However, 
because  many  species'  eggs  and  larvae  have  been  described  it 
is  possible  to  identify  and  describe  characters  of  taxonomic  and 
phylogenetic  value,  to  discuss  their  distribution  among  the  Clu- 
peiformes, and  to  point  out  some  similarities  and  conflicts  be- 
tween the  distribution  of  egg  and  larval  characters  and  current 
hypotheses  of  clupeiform  phylogeny. 

Taxonomic  characters  of  eggs  and  larvae 

The  taxonomic  characters  of  clupeoid  eggs  include  size,  shape, 
chorion  thickness  and  sculpturing,  width  of  perivitelline  space, 
degree  of  yolk  segmentation,  number  and  size  of  oil  globules  if 
present,  whether  they  are  pelagic  or  demersal,  whether  they  are 
adhesive  or  not,  and  whether  they  are  spawned  in  fresh,  brackish 
or  full  seawater. 

The  egg  of  Chirocentrus  is  1.60-1.65  mm  in  diameter,  has  a 
very  small  perivitelline  space,  is  pelagic,  spherical,  and  is  abun- 
dant near  shore,  especially  around  river  mouths  (Delsman, 
1930b).  The  egg  of  Chirocentus  nudus  has  a  chorion  with  fine 
hexagonal  sculpturing  (unique  among  clupeiforms)  and  up  to  9 
small  oil  globules,  while  the  egg  of  C.  dorab  has  a  smooth  cho- 
rion and  may  have  a  single  oil  globule  (Delsman,  1923,  1930b). 

The  eggs  of  clupeids  are  all  globular  and  they  range  in  size 


McGOWAN  AND  BERRY:  CLUPEIFORMES 


109 


Table  24.  Families,  Subfamilies,  Genera,  and  Species  of  Clupeiformes  with  Selected  Meristics.  Classification  follows  Whitehead  (1972) 
and  Nelson  (1976)  for  subfamilies;  Wongratana  (1980,  1983)  and  Nelson  (1983)  where  pertinent  for  genera  and  species;  otherwise  the  nomenclature 
is  that  of  the  author  cited  in  the  table.  Data  compiled  by  F.  H.  Berry  for  species  presumed  valid.  A;  Atlantic;  P:  Pacific;  c:  central;  e:  east;  n: 
north;  s:  south;  w:  west;  FW:  Freshwater;  IcP:  Indo-central  Pacific;  IwP:  Indo-west  Pacific;  1:  India;  Aust;  Australia;  Philipp:  Philippines;  US: 
United  States  of  America;  Braz:  Brazil,  Venz:  Venezuela;  Arg;  Argentina. 


Localion 

Dorsal 

Anal 

P2 

Gillrakers 

Vertebrae 

Upper 

Lower 

Reference 

DENTICIPITIDAE 

Denticeps 

clupeoides 

Nigeria 

9 

26-27 

5 

10 

41 

Clausen,  1959 

CHIROCENTRIDAE 

Chirocenlrus 

dorab 

IcP-Aust 

72- 

-74 

Delsman,  1923:  White- 
head, 1973 

nudus 

IwP 

CLUPEIDAE 

Clupeinae 

Sardinelta 

longiceps 

I 

17-19 

14-18 

9 

117-241 

150-253 

Wongratana,  1980 

neglecta 

se  Africa 

17-19 

16-18 

9 

108-166 

143-188 

Wongratana,  1983 

lemuru 

China-Aust 

17-19 

15-19 

9 

51-153 

77-188 

Wongratana,  1980 

Jussieui 

China-Aust 

19-20 

19-21 

8 

52-61 

88-101 

Wongratana,  1980 

sindensis 

I 

17-20 

17-21 

8 

16-46 

38-77 

Wongratana,  1980 

gibbosa 

IwP 

17-20 

17-22 

8 

16-36 

38-66 

Wongratana,  1980 

fimbriata 

IwP 

18-20 

19-22 

8 

27-47 

54-82 

Wongratana,  1980 

albella 

IwP 

18-20 

18-23 

8 

20-36 

41-68 

Wongratana,  1980 

dayi 

1 

18-19 

19-20 

8 

51-103 

87-134 

Wongratana,  1980 

fijiense 

N.  Guinea 

17-18 

18-19 

8 

33-40 

61-74 

Wongratana,  1980 

la  Wilis 

Philipp 

18-19 

1-22 

Wongratana,  1980 

hauliensis 

Taiwan 

18-20 

19-22 

8 

Wongratana,  1980 

brachysoma 

1-Aust 

17-20 

18-22 

8 

25-39 

48-67 

Wongratana,  1980 

richardsoni 

China 

18-19 

18-22 

8 

36-42 

63-74 

Wongratana,  1983 

zunasi 

China-Japan 

17-19 

17-21 

8 

21-23 

42-58 

Wongratana,  1980 

marquesensis 

Marquesas 

16-18 

17-21 

7-8 

15-58 

27-85 

42- 

-44 

Wongratana,  1980 

melanura 

IcP 

16-18 

17-20 

8 

20-41 

38-74 

Wongratana,  1980 

alncauda 

se  Asia 

18-19 

17-18 

8 

20-26 

39-43 

Wongratana,  1980 

aurita 

wAeA 

17-20 

16-18 

9 

56-81 

95-132 

45- 

-47 

Wongratana,  1980 

hrasiliensis 

wA 

17-18 

18-20 

9 

>150 

46 

Hildebrand,  1963d; 

Whitehead,  1973; 

Berry 

inaderensis 

eA 

8 

>70 

Whitehead,  1981 

rouxi 

ecA 

8 

34-40 

Whitehead,  1981 

Amblygasler 

sirm 

IwP 

18-20 

17-22 

14-18 

36-43 

Wongratana,  1980 

clupeoides 

wP 

18-19 

17-19 

12-14 

26-31 

Wongratana,  1980 

leiogaster 

IwP 

,19 

17-20 

13-16 

31-33 

Wongratana,  1980 

Herk/olsichlhys 

quadrimaculalus 

IwP-Aust 

18-20 

16-21 

13-17 

30-37 

Wongratana,  1980 

konigsbergeh 

wP-Aust 

18-19 

19-22 

15-17 

30-34 

Wongratana,  1980 

caslelnaui 

wP-AusI 

17-20 

17-22 

18-22 

39-52 

Wongratana,  1980 

gotoi 

N.  Guinea 

19 

17 

16 

34 

Wongratana,  1983 

lossei 

Persian  G. 

18-19 

15-18 

12-15 

29-35 

Wongratana,  1983 

spilura 

I 

17-19 

15-18 

12-15 

29-34 

Wongratana,  1980 

punclatus 

Red  Sea 

17-20 

13-18 

12-17 

31-39 

Wongratana,  1980 

dispilonotus 

se  Asia 

17-20 

16-19 

14-17 

34-38 

Wongratana,  1980 

Escualosa 

elongala 

Thailand 

16 

19 

26 

41 

Wongratana,  1983 

thoracata 

IwP-Aust 

15-17 

17-21 

16-25 

29-40 

Wongratana,  1980 

Opisthonema 

bidleri 

eP 

18-21 

20-23 

8-9 

35-47 

65-83 

46- 

-48 

Berry  and  Barrett,  1963 

medirasire 

eP 

17-20 

19-23 

8-9 

70-99 

110-156 

45- 

-48 

Berry  and  Barrett,  1963 

herlangai 

Galapagos 

19-20 

19-22 

8-9 

75-117 

133-171 

46- 

-48 

Berry  and  Barrett,  1963 

liherlale 

eP 

17-20 

19-22 

8-9 

1-149 

161-224 

44-48 

Berry  and  Barrett,  1963 

oglinum 

wA 

18-22 

22-25 

8-9 

43-60 

72-107 

45- 

-49 

Berry  and  Barrett,  1963 

captivai 

Colombia  A 

19-20 

18-21 

8 

(c25-28) 

49 

Rivas,  1972;  Berry 

110 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  24.    Continued. 


Location 

Dorsal 

Anal 

P2 

Gillraker 

Vertebrae 

Upper 

Lower 

Reference 

Harengula 

humeralis 

wA 

18 

16 

8 

13-15 

26-29 

40-41 

Whitehead,  1973;  Berry 

clupeota 

wA 

18 

18 

8 

14-16 

27-31 

41-42 

Whitehead,  1973;  Berry 

jaguana 

wA 

17-18 

17-18 

7-8 

16-20 

31-35 

41-43 

Whitehead,  1973;  Berry 

peruana 

esP 

18-19 

15-17 

8 

15-19 

31-51 

40-42 

Berry 

thrissina 

enP 

16-20 

14-17 

8-9 

9-18 

24-33 

40-43 

Hildebrand,  1946;  Berry; 
Miller  and  Lea,  1972 

Ramnogaster 

arcuata 

Arg 

7 

Whitehead,  1973,  1965 

melanostoma 

Arg 

Whitehead,  1965 

pallida 

Arg 

Whitehead,  1965 

Platanichthys 

platana 

Braz 

14 

16 

7 

13 

25 

Whitehead,  1973 

Sardinops 

sagax  sagax 

esP 

17-20 

17-20 

8 

49-54 

Ahlstrom 

sagax  caerulea 

enP 

17-20 

17-20 

8 

21-23 

44-45 

48-54 

Berry;  Miller  and  Lea, 

1972 

neopilchardus 

Aust 

18-20 

17-21 

58-93 

50-52 

Berry 

melanosticta 

e  Asia 

ocellata 

s  Africa 

8 

Whitehead,  1981 

Sardina 

pilchardus 

enA 

17-18 

17-18 

8 

44-106 

50-53 

Whitehead,  1981 

Rhinosardinia 

amazomca 

Guyanas 

13-16 

15-19 

8 

ca.  20 

33-43 

Hildebrand,  1963d; 
Whitehead,  1973; 
Berry 

bahiensis 

Braz 

17 

18 

Hildebrand,  1963d 

Lile 

piqmtinga 

wcA 

15-17 

17-19 

7-8 

12-17 

30-36 

38-41 

Whitehead,  1973;  Berry 

stolifera 

eP 

17-18 

17-23 

8 

13-18 

32-36 

42-44 

Hildebrand,  1946 

Clupea 

harengus 

nA 

16-20 

16-20 

37-52 

53-60 

Hildebrand,  1963d; 
Wheeler,  1969 

pallasi 

nP 

13-21 

14-20 

20 

45 

46-55 

Berry,  1964b,  Ahlstrom; 
Miller  and  Lea,  1972 

bentincki 

Chile 

Whitehead,  1965 

Sprallus 

spraltus 

enA 

16-19 

18-20 

7-8 

46-49 

Whitehead,  1965; 
Wheeler.  1969 

antipodum 

Aust 

8 

Whitehead,  1965 

muelten 

Aust 

8 

Whitehead,  1965 

hassensis 

Aust 

8 

46 

Whitehead,  1965 

fuegensis 

Chile 

8 

49-51 

Whitehead.  1965 

Clupeonella 

cultiventris 

Whitehead,  1965 

grimmi 

Whitehead,  1965 

engraulifonnis 

Whitehead.  1965 

abrau 

Whitehead.  1965 

Dussumieriinae 

Eirumeus 

teres 

Cosmop. 

18-22 

10-19 

8-9 

12-15 

28-35 

48-50 

Wongratana.  1980; 
Miller  etal..  1979; 
Miller  and  Lea,  1972 

whiteheadi 

S.  Africa 

18-20 

12-13 

8 

16-18 

36-39 

54-56 

Wongratana,  1983 

Dussumieria 

elopsoides 

IcP 

18-23 

14-18 

8 

11-16 

21-32 

54-55 

Wongratana,  1980; 
Delsman,  1925 

acuta 

1-China 

19-22 

14-18 

8 

11-15 

19-26 

54-55 

Wongratana,  1980; 
Delsman,  1925 

McGOWAN  AND  BERRY:  CLUPEIFORMES 
Table  24.    Continued. 


Ill 


Location 

Dorsal 

Anal 

P2 

Gillrakcrs 

Venebrae 

Upper 

Lower 

Reference 

Spratelloidinae 

Spralelloides 

gracilis 

IwP  Aust 

11-14 

11-14 

S 

10-12 

28-37 

Wongratana,  1980 

lewisi 

N.  Guinea 

11-13 

10-13 

8 

9-11 

28-32 

Wongratana,  1983 

delkatulus 

IwP  Aust 

10-14 

9-11 

8 

9-12 

26-32 

44- 

-45 

Wongratana,  1980; 
Miller  etal.,  1979 

robustus 

Aust 

12-13 

10-11 

8 

9-11 

28-35 

Wongratana,  1980 

Jenkinsia 

lamprolaenia 

wcA 

12-13 

13-16 

8 

19-24 

39- 

-40 

Whitehead.  1973;  Berry; 
Cervigon  and 
Velazquez,  1978 

stolifera 

wcA 

9-12 

13-16 

18-25 

Whitehead,  1973 

majua 

wcA 

11-13 

21-28 

Whitehead,  1973 

parvula 

Venz 

10-13 

12-16 

20-24 

38- 

-39 

Cervigon  and  Velaz- 
quez, 1978 

Dorosomatinae 

Clupanodon 

ihrissa 

wP 

16 

21-26 

8 

(190-480) 

(200-420) 

Wongratana,  1980 

Konosirus 

punctatus 

China 

16-19 

21-25 

8 

(145-270) 

(160-250) 

Wongratana,  1980 

Nematalosa 

erebi 

Aust 

14-16 

19-22 

8 

(155-370) 

(145-370) 

Wongratana,  1980 

chanpole 

IwP 

15-17 

22-26 

8 

(250-315) 

(255-355) 

Wongratana,  1980 

arabica 

I 

17-19 

18-20 

8 

(145-335) 

(180-390) 

Wongratana,  1980 

come 

I-Aust 

17-18 

20-24 

8 

(175-245) 

(170-250) 

Wongratana,  1980 

nasus 

I-wP 

15-19 

20-26 

8 

(155-310) 

(165-315) 

Wongratana,  1980 

japonica 

wP 

16-18 

19-22 

8 

149-205 

156-193 

Wongratana,  1980 

vlaminghi 

Aust 

16-17 

19-25 

8 

216-300 

239-328 

Wongratana,  1980 

paubuensis 

N.  Guinea 

14-16 

22-27 

8 

72-342 

82-309 

Wongratana,  1980 

flyensis 

N.  Guinea 

14-16 

21-26 

8 

152-553 

195-508 

Wongratana,  1983 

Gonialosa 

whitcheadi 

Burma 

15 

27 

8 

(92) 

90-93 

Wongratana,  1983 

mammmna 

I 

14-16 

22-27 

8 

87-160 

96-166 

Wongratana,  1980 

modesta 

Burma 

15-17 

24-28 

8 

(125-170) 

(140-185) 

Wongratana,  1980 

Anodontostoma 

chacunda 

IwP 

17-21 

17-22 

8 

52-98 

54-96 

Wongratana,  1980 

selangkat 

wP 

18-20 

17-21 

8 

129-186 

100-166 

Wongratana,  1980 

ihailandiae 

IwP 

17-20 

18-23 

8 

43-125 

46-140 

Wongratana,  1983 

Dorosoma 

cepedianum 

wnP 

10-13 

25-36 

7-8 

(ca.  300- 
400) 

48- 

-51 

Miller,  1960;  Berry 

petenense 

wnA 

11-14 

17-27 

7-8 

(ca.  300- 
400) 

40- 

-45 

Miller  1960;  Berry 

anale 

eMexico 

29-38 

Miller,  1960 

chavesi 

eNicaragua 

12-14 

(22-31) 

Miller,  1960 

smithi 

wMexico 

9-13 

(22-31) 

43- 

-46 

Hildebrand,  1963d; 

Miller,  1960 
Berry 

Congothnssinae 

Congothrissa 

gossei 

Congo 

14-16 

15-17 

7-8 

ca. 

40 

Poll,  1964 

Alosinae 

Hilsa 

kelee 

IwP 

16-19 

17-22 

8 

(45-105) 

(70-180) 

Wongratana,  1980 

Tenualosa 

toli 

IwP 

17-18 

15-21 

8 

(38-55) 

(60-95) 

Wongratana,  1980 

macrura 

Java 

19 

21-22 

8 

(46-52) 

(63-74) 

Wongratana,  1980 

reevesii 

wP 

17-19 

16-20 

8 

53-131 

80-248 

Wongratana,  1980 

ilisha 

wP 

17-20 

18-23 

8 

46-196 

62-272 

Wongratana,  1980 

thibaudeaui 

Thailand 

16-18 

19-23 

8 

(170-248) 

(205-320) 

Wongratana,  1980 

112 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES- AHLSTROM  SYMPOSIUM 


Table  24,    Continued. 


Upper 


Gadusia 
chapra 
variegata 

Alosa 

sapidissima 

pseudoharengus 

mediocris 

chn^sochloris 

alabamae 

aestivalis 

fallax 

alosa 


Pakistan 
Burma 

wnA-enP 

eUS-Canada 

eUS 

eUS 

eUS 

eUS,  Canada 

enA 

enA 


14-18 
16-17 

17-20 
15-19 
15-20 
16-21 
16-20 
15-20 
18-21 

18-21 


21-25 

25-27 

20-23 
17-21 
19-23 
18-21 
19-22 
16-21 
19-23 

20-26 


(160-235)     (170-255) 
(250-270)     (252-270) 


Wongratana.  1980 
Wongratana,  1980 


59-73 

54-59 

Hildebrand,  1963d; 

Berry 

38-43 

46-50 

Hildebrand,  1963d; 

Berry 

18-23 

54-55 

Hildebrand,  1963d; 

Berry 

20-24 

53-55 

Hildebrand,  1963d; 

Berry 

42-48 

55 

Hildebrand.  1963d; 

Berry 

41-51 

49-53 

Hildebrand,  1963d; 

Berry 

20-40 

55-59 

Whitehead,  1981; 
Wheeler,  1969 

55-85 

57-58 

Whitehead,  1981; 
Wheeler,  1969 

Ethmalosa 

fimbriata 

eA 

18 

22 

8 

53 

136 

44 

Whitehead,  1981;  Berry 

Brevoortia 

aurea 

Braz 

gunteri 

Gulf  Mexico 

17-20 

20-25 

7 

144 

42- 

-44 

Hildebrand,  1963d 

patronus 

Gulf  Mexico 

17-21 

20-23 

7 

138-142 

42- 

-48 

Hildebrand,  1963d;  Berry 

smilhi 

eUS 

18-20 

22-23 

7 

151 

45- 

-47 

Hildebrand,  1963d;  Berry 

tyrannus 

eUS,  Canada 

18-22 

18-24 

7 

137-145 

45- 

-50 

Hildebrand,  1963d;  Berry 

Ethmidium 

chilcae 

Chile-Peru 

18-23 

15-18 

7-8 

123-129 

147-159 

48- 

-50 

Hildebrand,  1946;  Berry 

Pellonulinae 

Ehirava 

fluvial  His 

I 

14-16 

12-18 

8 

12-14 

24-30 

Wongratana,  1980 

madagascarensis 

Nelson,  1970 

Dayella 

malabanca 

I 

14 

17 

8 

10-11 

24-27 

Wongratana,  1980 

Clupeoides 

borneensis 

Borneo 

15-18 

15-19 

8 

9-12 

18-24 

Wongratana,  1980 

hypselosoma 

Borneo 

14-15 

16-18 

8 

10 

12-19 

Wongratana,  1980 

paupensis 

Borneo 

13-16 

17-22 

8 

9-11 

15-19 

Wongratana,  1980 

venulosus 

N.  Guinea 

13-15 

20-22 

8 

Corica 

laciniata 

Borneo 

15-17 

13-16  +  2 

8 

10-13 

23-27 

Wongratana,  1980 

soborna 

I 

15-16 

14-15  +  2 

8 

9-11 

19-21 

Wongratana,  1980 

Pellonulinae 

Laevisculella 

dekimpet 

Nelson,  1970 

Odaxothrissa 

losera 

Nelson,  1970 

Potamothrissa 

aculiroslris 

Nelson,  1970 

Spratellomorpha 

bianalis 

Nelson,  1970 

Pristigasterinae 

llisha 

sirishai 

I 

17-18 

39-43 

8-12 

22-26 

Wongratana,  1980 

novacula 

Burma 

16 

43-45 

10-12 

21-23 

Wongratana,  1980 

megaloplera 

1 

16-19 

38-53 

8-11 

19-23 

47- 

-48 

Wongratana,  1980;  Berry 

elongala 

1-China 

16-20 

43-53 

9-13 

21-25 

Wongratana.  1980 

filigera 

I 

17-21 

46-52 

9-12 

19-23 

50- 

-52 

Wongratana,  1 980;  Berry 

macrogaster 

I 

18-19 

49 

11-12 

23-25 

Wongratana,  1980 

pristigaslroides 

Java 

17-18 

47-48 

9-10 

17 

Wongratana,  1980 

kampeni 

1 

16-18 

38-46 

9-12 

20-24 

Wongratana,  1983 

striatula 

1 

15-18 

40-48 

10-13 

21-24 

Wongratana,  1980 

melastoma 

IwP 

15-18 

35-48 

10-13 

21-25 

Wongratana,  1983 

McGOWAN  AND  BERRY:  CLUPEIFORMES 


113 


Table  24.    Continued. 


Location 

Dorsal 

Anal 

P2 

Gillrakcrs 

Venebrae 

Upper 

Lower 

Reference 

obfuscala 

I 

16 

39-42 

7 

12-13 

27-28 

Wongratana,  1980 

afncana 

ecA 

15 

47 

Whitehead,  1981 

amazonica 

Braz 

20 

34 

6 

15 

29 

Hildebrand,  1963d 

furlhii 

ecP 

15-17 

46-50 

11-12 

20-25 

50-52 

Peterson,  1956; 
Hildebrand,  1946; 
Meek  and  Hildebrand, 

Neoopisthoplerus 

1923 

cubanus 

Cuba 

12-15 

39-43 

0 

10 

17-19 

47 

Hildebrand,  1963d,  Berry 

tropicus 

15 

43-48 

0 

8 

20 

45-47 

Peterson,  1956; 
Hildebrand,  1946 

Pellonulinae 

Clupeichthys 

hleekeh 

Borneo 

14-15 

16-18  +  2 

8 

8-10 

16-18 

Wongratana,  1980 

aesarnensis 

Thailand 

13-15 

14-16  +  2 

8 

8-10 

17-19 

Wongratana,  1983 

goniognathus 

Thailand 

14-15 

15-17  +  2 

8 

8 

15-16 

Wongratana,  1980 

perakensis 

Malaya 

13-15 

14-17  +  2 

7 

5-9 

13-15 

Wongratana,  1980 

Pellonula 

leonensis 

ecA 

8 

20-30 

Whitehead,  1981 

vorax 

ecA 

Whitehead,  1981 

Microthrissa 

royauxi 

Nelson,  1970 

Poecilothrissa 

congica 

Nelson,  1970 

Hyperlophus 

villala 

Nelson,  1970 

Cynolhnssa 

ansorgii 

Whitehead,  1981 

memo 

Potamalosa 

richmondia 

Wongratana,  1980 

Gitchnstella 

aestuarius 

Wongratana,  1980 

Limtwlhrissa 

mtodon 

Wongratana,  1980 

Stolothrissa 

tanganicae 

Wongratana,  1980 

Pristigasterinae 

Prist  igaster 

cayana 

Brazil 

13-16 

44-55 

0 

10 

20-23 

43-44 

Hildebrand,  1963d;  Berry 

Opisthoplerus 

valenaermesi 

China 

16-18 

54-65 

7 

9-12 

23-25 

Wongratana,  1980 

lardoore 

I 

14-17 

51-63 

7 

8-12 

22-28 

50-52 

Wongratana,  1980;  Berry 

dovii 

ecP 

12-13 

53-62 

0 

17-18 

51-52 

Meek  and  Hildebrand, 
1923;  Ahlstrom 

equalorialis 

esP 

11-12 

59-62 

0 

10 

25 

46-47 

Hildebrand,  1946; 
Ahlstrom 

Raconda 

russehana 

I 

0 

81-92 

8-11 

23-27 

62 

Wongratana,  1980;  Berry 

Pellona 

ditchela 

I-Aust 

16-19 

34-41 

7 

10-14 

22-27 

42 

Wongratana,  1980;  Berry 

day! 

I 

17-18 

35-42 

7 

9-11 

20-21 

Wongratana,  1983 

altamazonica 

Braz 

18 

37-40 

6-7 

9 

12-14 

Hildebrand,  1963d;  Berry 

castelnacana 

Braz-Venz 

18-20 

34-42 

6-7 

13-14 

24-25 

45-46 

Hildebrand.  1963d; 
Whitehead,  1973;  Berry 

flavipinnis 

Braz-Arg 

17-21 

38-47 

7 

14-15 

28-31 

43 

Hildebrand,  1963d; 
Whitehead,  1973;  Berry 

harroweri 

wcA 

14-17 

36-42 

5-6 

12-13 

24-28 

38-40 

Hildebrand,  1963d; 
Whitehead,  1973;  Berry 

114 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  24.    Continued. 


Location 

Dorsal 

Anal 

P2 

Gillrakers 

Vertebrae 

Upper 

Lxjwer 

Reference 

Odontognathus 

mucronatus 

wsA 

10-12 

74- 

-85 

0 

7-9 

22-26 

53- 

-54 

Hildebrand,  1963d; 
Whitehead,  1973;  Berry 

compressus 

wcA 

10-14 

52- 

-62 

0 

9 

18-23 

46- 

-47 

Hildebrand,  1963d; 
Whitehead,  1973;  Berry, 
Meek  and  Hildebrand, 

1923 

panamensis 

ecP 

11-12 

61-68 

0 

ca.  21 

51- 

-53 

Peterson,  1956; 

Meek  and  Hildebrand, 

Chirocentrodon 

1923 

bleekenanus 

wcA 

14-16 

38- 

-45 

6-7 

4-6 

15-17 

44- 

-45 

Hildebrand.  1963d; 
Whitehead,  1973;  Berry 

Pliosteostoma 

lutipinnis 

ecP 

49-51 

0 

50-51 

Peterson,  1956;  Berry 

macrops 

CLUPEIDAE 

Status  not  verified 

Alosinae 

Caspialosa 

maeolica 

Nelson,  1970 

Clupeinae 

Clupeonella 

delicalula 

Nelson,  1970 

Dorosomatinae 

Nematatosa 

horm 

Nelson,  1970 

Thratlidion 

noctivagus 

Sierrathrissa 

leonensis 

ENGRAULIDAE 

Coilinae 

Coilia 

ramcarati 

I 

14-16 

9-10 

21-23 

29-30 

Wongratana,  1980 

borneensis 

Borneo 

14-15 

7 

21-23 

32 

Wongratana,  1980 

reynaldi 

I 

13-14 

7 

20-27 

28-36 

Wongratana,  1980 

coomansi 

Borneo 

14 

7 

21-24 

31-33 

Wongratana,  1980 

rebentischii 

Borneo 

14-15 

7 

15-19 

22-27 

Wongratana,  1980 

neglecta 

I 

13-15 

7 

17-19 

21-27 

Wongratana,  1980 

dussumieri 

I 

13-15 

7 

17-20 

23-26 

Wongratana,  1980 

rendahli 

China 

13-15 

7 

grayii 

I-China 

13-14 

7 

21-23 

28-31 

Wongratana,  1980 

lindmam 

Thailand 

12-15 

7 

18-25 

29-34 

Wongratana,  1980 

macrognalhos 

Borneo 

14-15 

7 

15-16 

22-24 

Wongratana,  1980 

mystus 

China 

13-15 

79- 

-89 

6-7 

17-22 

25-29 

Wongratana,  1980 

nasus 

China-Japan 

13-15 

87- 

-117 

7 

16-20 

23-26 

Wongratana,  1980 

Engraulinae 

Engraulis 

japonicus 

IwP 

14-17 

14-22 

22-34 

26-39 

Wongratana,  1980 

(=australis) 

eA 

(=encrasicolus) 

eA 

Wongratana,  1980 

(=capensis) 

sAfrica 

Wongratana,  1980 

anchoita 

swA 

Whitehead,  1973 

euryslole 

nwA 

15-16 

16- 

-19 

7 

28-31 

43- 

-45 

Whitehead.  1973 

ringens 

seP 

15-18 

19- 

-24 

35-43 

38-48 

46- 

-49 

Hildebrand,  1946;  Berry 

mordax 

neP 

14-19 

19- 

-26 

28-41 

37-45 

43- 

-47 

Miller  and  Lea.  1972 

"juruensis" 

Amazon 

Whitehead,  1973 

A  nchovia 

clupeoides 

swA 

14 

31 

7 

105 

41 

Whitehead,  1973 

rastralis 

eP 

12-14 

26- 

-30 

ca.  50 

Meek  and  Hildebrand, 
1923;  Whitehead,  1973 

tnnilatis 

cubana 
parva 

lamprotaenia 

hepselus 

filfera 

lyok'pis 

ginsburgi 

tricolor 

choerosloma 

januaria 

mitchilli 

pecloralis 

cayorum 

argenteus 

argentivitlala 

ischana 


McGOWAN  AND  BERRY:  CLUPEIFORMES 

Table  24.    Continued. 


115 


Location 

Dorsal 

Anal 

P2 

GUItakei^ 

Vertebrae 

Upper                   Lower 

Reference 

surinamensis 
macrolepidota 

magdalenae 

cwA 
eP 

neP 

13-15 
12-14 

25-28 
27-33 

7 

47-62 
ca.  95 

40-42 

Whitehead,  1973 
Meek  and  Hildebrand, 

1923;  Whitehead,  1973; 

Peterson,  1956 

A nchoa 
spinifer 

wcA-ecP 

15-17 

30-40 

7 

12-16            12-18 

19-21 ± 

Hildebrand,  1963c; 

Venz 

wA 
wcA 

wA 

wA 

wcA 

wA 

Venz 


pananwnsis 

ecP 

compressa 

mundeoloides 

walkeri 

anatis 

curta 

ecP 

delicatissima 

P 

helleri 

P 

slarksi 

ecP 

clarki 
eigenmanma 

P 

ecP 

scofteldi 

P 

lucida 

ecP 

13-15 


14-15 


26-32 


14-16 

20-24 

15-16 

21-25 

13-16 

19-27 

13-16 

18-24 

13-15 

19-23 

12-16 

19-27 

18-22 


wsA 

14-16 

18-22 

Bermuda 

13-15 

22-24 

wsA 

14-15 

21-24 

wnA 

14-16 

24-30 

Braz 

14-16 

25-27 

wA 

13-15 

25-29 

Venz 

16 

32 

ecP 

18-20 

enP 

18-21 

ecP 

12 

32-26 

14-19 


19-21 


16-18 


23-25 
13-15  23-26  18-21         23-26 

20-23 


32-40 


20-22 


21-22 


41 


7 
7 

17-23 
17-20 

23-33 

23-28 

42-43 
38-41 

7 

13-18 

16-22 

39-42 

7 

15-21 

19-25 

40-44 

7 

17-19 

20-26 

39-40 

7 

16-23 

20-27 

41-43 

44-45 


Whitehead,  1973; 

Peterson,  1956; 

Cervigon,  1966; 

Nelson,  1983 
Whitehead.  1973; 

Cervigon.  1966; 

Hildebrand,  1963c 
Whitehead,  1973 
Whitehead,  1973; 

Hildebrand,  1964 
Whitehead.  1973; 

Hildebrand,  1964 
Whitehead,  1973; 

Hildebrand,  1964 
Whitehead,  1973; 

Hildebrand,  1964 
Whitehead,  1973; 

Cervigon,  1966; 

Hildebrand,  1963c 
Cervigon,  1966; 

Hildebrand,  1963c 


25-28 


18-22 

24-28 

40- 

-42 

Hildebrand.  1963c 

17-20 

23-26 

41- 

-42 

Hildebrand,  1963c 

20-23 

23-26 

41- 

-42 

Hildebrand,  1963c 

15-19 

20-26 

38- 

-44 

Hildebrand,  1963c 

13-14 

17-19 

4 

2 

Hildebrand,  1963c 

13-15 

15-17 

43 

Hildebrand,  1963c 

14 

19 

Hildebrand,  1963c 

17-21 

24-25  + 
19-22 

Peterson,  1956;  Nelson, 
1983 

19-21 

22-24  + 
19-21 

Peterson,  1956;  Nelson, 
1983 

22-24 

18-20  + 
21-24 

18-19  + 
20-22 
18-20  + 
21-23 
18-20  + 
21-24 
17-19  + 
20-23 

Peterson,  1956;  Nelson. 

1983;  Hildebrand, 

1946 
Nelson,  1983 

Nelson,  1983 

Nelson,  1983 

Nelson,  1983 

22-25 

19-22  + 
19-22 

Peterson,  1956;  Nelson. 
1983 

23-26 

19-21  + 
19-21 
20-23  + 
18-21 

Nelson.  1983;  Miller 
and  Lea,  1972 

Nelson,  1983;  Miller 
and  Lea,  1972 

22-26 

20-22  + 
19-21 
21  +  21 

Peterson,  1956;  Nelson, 

1983 
Nelson,  1983 

12-13 

17-21  + 
20-25 
20-22  + 
21-23 

Peterson,  1956;  Nelson, 

1983 
Nelson,  1983 

19-22 

17-20  + 
19-22 

Peterson,  1956;  Nelson, 
1983 

116 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 

Table  24.    Continued. 


Location 

Dorsal 

Anal 

Gil  I  rakers 

Vertebrae                             Refere 

'2                  Upper 

Lower 

ice 

naso 

ecP 

14-16 

23-27          21 

-24         23-27 

19-21  + 
19-22 

Peterson,  1956;  Nelson, 
1983;  Hildebrand, 

1946 

chamensis 

eP 

21  +  22 

Nelson,  1983 

nasus 

ecP 

15-16 

21-27          21 

-25         24-28 

20-21  + 
20-22 

Nelson,  1983; 
Hildebrand, 

1946 

exigua 

ecP 

17-22 

23-25 

43-45 

Peterson,  1956;  Nelson, 

1983 

Anchovielta 

leptdentostole 

wsA 

14-16 

22-25 

7              17-18 

19-23 

Whitehead,  1973; 
Cervigon,  1966 

urevirostris 

wsA 

16-18 

18-20 

7 

24-27 

Whitehead,  1973 

guianensis 

wcsA 

14-15 

18-20 

7 

23-26 

40         Whitehead,  1973 

cayennensis 

wcA 

13 

16 

7 

30 

Whitehead,  1973 

nattereri 

Braz 

12 

25-29 

Whitehead,  1973; 
Cervigon,  1966 

perfasciata 

wnA 

12-15 

15-19 

18-23 

25-28 

42-44      Cervigon,  1966 

elongata 

Panama  A 

13-14 

22-24 

17-18 

22-24 

39         Cervigon,  1966 

blackbumi 

Venz 

13-15 

25-27 

10-12 

15-17 

43         Cervigon,  1966 

jainesi 

Braz 

12-13 

19-21 

12-13 

20-21 

40         Cervigon,  1966 

vaillanti 

23 

19 

Whitehead,  1973 

carrikeri 

17-18 

14-15 

Whitehead.  1973 

Slolephorus 

indicus 

IwP 

14-17 

17-22 

16-20 

20-28 

20-23+  Wongratana, 
19-21 

980 

commersonii 

IwP 

15-17 

20-23 

12-27 

21-35 

Wongratana, 

980 

brachycephalus 

Papua 

16-17 

22-25 

15-17 

20-22 

Wongratana, 

983 

chinensis 

China 

16-18 

20-23 

18-19 

26-27 

Wongratana, 

980 

wailei 

1-Aust 

15-17 

19-24 

14-17 

1-4 

Wongratana, 

980 

holodon 

seAfr 

15-18 

20-23 

17-22 

24-29 

Wongratana, 

980 

andhraensis 

el-Papua 

15-17 

19-23 

14-15 

20-21 

Wongratana, 

980 

lysoni 

Papua 

15-17 

21-25 

15-18 

21-25 

Wongratana. 

983 

insulahs 

I-China 

14-17 

19-23 

16-20 

22-28 

Wongratana, 

980 

dubwsus 

I 

14-16 

19-24 

19-24 

25-31 

Wongratana, 

980 

baganensis 

I 

14-16 

20-23 

16-19 

20-24 

Wongratana, 

980 

iri 

Thailand 

14-15 

19-22 

15-17 

19-22 

Wongratana, 

980 

oligobranchus 

Philipp 

14-16 

18 

7              13-14 

17-18 

Wongratana, 

983 

Thryssa 

baelama 

IwP 

15 

29-34 

14-20 

19-26 

Wongratana, 

980 

chefuensis 

China 

14 

29-34 

23-28 

27-30 

Wongratana, 

980 

rastrosa 

N.  Guinea 

14-15 

32-35 

39-44 

55-61 

Wongratana, 

980 

scratchteyi 

N.  G.-Aust 

14 

33-36 

15-18 

18-20 

Wongratana, 

980 

aesluaha 

N.  G.-Aust 

13-15 

32-36 

22-25 

27-29 

Wongratana, 

980 

kammalcnsis 

Thailand 

14-15 

32-37 

23-27 

28-32 

Wongratana, 

980 

kammalensoides 

I 

14 

34-35 

18 

24-25 

Wongratana, 

983 

vilrirostris 

e  Africa 

13-15 

34-43 

14-17 

20-23 

Wongratana, 

980 

adetae 

China 

13-14 

38-44 

13-16 

20-22 

Wongratana, 

980 

dussumieri 

I-Taiwan 

12-15 

34-38 

13-16 

17-19 

Wongratana, 

980 

mysto-x 

I-China 

13-15 

35-39 

9-11 

13-16 

Wongratana, 

980 

polybranchialis 

I 

13-15 

38-42 

18-21 

25-27 

Wongratana, 

983 

gualamiensis 

I 

13-15 

36-40 

11-13 

17-19 

Wongratana, 

980 

malabarka 

I 

13-15 

37-41 

14-16 

17-19 

Wongratana, 

980 

hamiltonii 

IwP 

13-15 

35-41 

7-10 

11-15 

Wongratana, 

980 

whiteheadi 

Pers.  G. 

12-14 

42-46 

13-15 

18-20 

Wongratana, 

983 

purava 

I 

12-14 

42-47 

14-16 

18-19 

Wongratana. 

980 

stenosoma 

I 

12-14 

43-48 

13-15 

17-19 

Wongratana, 

983 

dayi 

I 

13-14 

44-49 

10-13 

14-18 

Wongratana, 

983 

spinidens 

I-Thai 

12-14 

44-48 

9-11 

13-15 

Wongratana, 

980 

setirostris 

I-China 

13-15 

32-39 

5-6 

10-12 

Wongratana, 

980 

Encrasicholina 

purpurea 

Hawaii 

12-15 

14-18 

7             1 5-22 

23-29 

41-44      Miller  etal.,  1 
Wongratana 
Nelson,  198 

979; 
,  1980; 
3 

McGOWAN  AND  BERRY:  CLUPEIFORMES 

Table  24.    Continued. 


117 


Localion 

Dorsal 

Anal 

P2 

Gillrakers 

Vertebrae 

Upper 

Lower 

Reference 

punclifer 

IwP 

12-16 

14-17 

7 

15-22 

23-29 

24-25  + 
17-20 

Miller  etal.,  1979; 
Wongratana,  1980; 
Nelson.  1983 

heterolobus 

IwP 

13-15 

15-19 

20-25 

23-29 

22-24  + 
19-21 

Miller  etal.,  1979; 
Wongratana,  1980; 
Nelson,  1983 

devisi 

I-Aust 

13-16 

17-21 

17-18 

20-22 

21-23  + 
19-21 

Miller  etal..  1979; 
Wongratana,  1980; 
Nelson,  1983 

ronquilloi 

Philipp 

15-17 

19-22 

20-21 

28-30 

Wongratana,  1980 

Pterengraulis 

alherinoides 

wcA 

12-14 

29-35 

7 

10-12 

12-15 

43-45 

Lycengraulis 

hatesii 

wcA 

14-16 

27-30 

7 

9-13 

12-15 

47 

Whitehead,  1973; 
Cervigon,  1966 

grossidens 

wcA 

14-16 

24-28 

7 

13-19 

17-23 

41-48 

Whitehead,  1973; 
Cervigon,  1966 

poeyi 

eP 

13-15 

22-27 

14-18 

18-23 

43 

Whitehead,  1973; 
Peterson,  1956; 
Meek  and  Hildebrand,  1923 

Cetengraulis 

edenlulus 

wcsA 

13-16 

21-27 

7 

45-53 

Whitehead,  1973; 
Meek  and  Hildebrand,  1923 

mysticelus 

ecP 

13-17 

18-26 

40-58 

43-60 

39-43 

Peterson,  1956; 

Hildebrandichthys 
seliger 

Venz 

12 

25 

Papuengraulis 
micropinna 

N.  Guinea 

5-6 

54-56 

Lycolhrissa 
crocodilus 

China 

10-13 

47-51 

Setipinna 
tenuifilis 
papuensis 
melanochir 
taty 

wheeleri 
phasa 
brevifilis 

I-China 

N.  G.-Aust 

China 

I-China 

Burma 

I 

I 

14-16 
14-15 
13-15 
13-15 

14 
13-15 
13-15 

49-59 
54-57 
48-53 
48-58 
72-77 
69-82 
68-75 

Heterothrissa 
breviceps 

I-China 

17-18 

59-64 

Status  not  verified: 

Thrissa 

grayi 

Lycengraulis 
barboun 
olidus 

Cetengraulis 
juruensis 

Amazon-FW 

20-22 

Anchoa 
arenicola 

Anchoviella 
hubbsi 
pallida 
balboae 

llisha 
indica 

ca.  23 


ca.  33 


6-7 


15-16 

25-27 

6-7 

8-10 

10-12 

13-17 

11 

15 

7-10 

9-12 

13-17 

18-20 

16-18 

21-22 

15-16 

18-19 

14-15 

17 

7-8 


11-12 


Hildebrand,  1946;  Meek 
and  Hildebrand,  1923; 
Miller  and  Lea,  1972 

Cervigon,  1966; 
Schultz,  1949 

Wongratana,  1980 

Wongratana,  1980 

Wongratana,  1980 
Wongratana,  1980 
Wongratana,  1980 
Wongratana,  1980 
Wongratana,  1983 
Wongratana,  1980 
Wongratana,  1980 

Wongratana,  1980 


Nelson,  1970 

Nelson,  1970 
Nelson,  1970 


20  +  20  Nelson,  1984 

Nelson,  1970 

Nelson,  1970 
Nelson,  1970 
Nelson,  1970 

Nelson,  1970 


118 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


from  0.59-0.75  mm  in  Sardinella  jussiem  (Bensam,  1970)  to 
2.5-3.8  mm  in  Alosa  sapidissima  (Jones  et  al.,  1978).  Most 
clupeid  eggs  are  1-2  mm  in  diameter.  All  have  a  segmented 
yolk.  The  chorion  is  not  ornamented  or  sculptured.  The  peri- 
vitelline  space  varies  in  thickness  among  species.  It  may  be  as 
large  as  45%  of  the  egg  diameter  (Sardinella  zunasi)  or  as  small 
as  5-10%  (Anodontostoma.  Opisthoplerus).  The  egg  yolk  may 
shrink  relative  to  the  egg  diameter  when  preserved  (Bensam, 
1967)  and  the  yolk  decreases  in  size  during  the  development  of 
the  embryo.  Oil  globules  are  present  in  the  eggs  of  most  clupeids. 
One  is  often  present  (e.g..  Sardinella.  Harengula,  Sardinops); 
Escualosa  thoracata  has  nine  (Delsman,  1932a,  described  as 
Clupeoides  Hie).  The  eggs  of  clupeids  which  lay  demersal  adhe- 
sive eggs  (Clupea.  Dorosoma,  Spratelloides)  have  a  gelatinous 
covering  around  the  egg.  The  pelagic  egg  of  Tenualosa  ilisha  is 
also  covered  by  a  gelatinous  sheath.  In  Dorosoma  petenense  the 
adhesive  layer  is  composed  of  transformed  ovarian  follicular 
epithelium,  an  unusual  feature  among  teleosts  (Shelton,  1978). 

Eggs  of  anchovies,  family  Engraididae.  range  in  size  from  0.7 
mm  (Lycengraidis)  to  1.75  mm  (Slolephorus,  long  axis).  Their 
shape  varies  from  globular  to  extremely  elliptical.  The  ratio  of 
the  long  axis  of  the  ellipse  to  the  short  axis  has  been  used  to 
identify  anchovy  eggs  (Peterson,  1956;  Phonlor,  1978).  Some 
Slolephorus  species  have  a  distinct  knob  on  one  end  of  the  egg 
surrounding  the  micropyle.  A  perivitelline  space  is  present  but 
smaller  and  less  noticeable  than  in  clupeid  eggs  because  of  the 
elliptical  shape.  Oil  globules  are  absent  except  in  the  genera 
Coilia  and  Setipinna,  which  have  spherical  eggs  like  clupeids, 
and  the  Indo-Pacific  species  of  Slolephorus.  Fig.  58  illustrates 
representative  eggs  of  clupeiforms. 

Yolk-sac  larvae  are  characterized  by  their  size  at  hatching  (2- 
5  mm),  which  is  related  to  yolk  size;  whether  the  yolk-sac  is 
rounded  or  pointed  posteriorly,  the  number  and  position  of  oil 
globules,  number  of  myomeres  and  pigmentation.  Larvae  from 
demersal  adhesive  eggs  may  hatch  with  pigmented  eyes  (Clupea 
harengus);  those  from  pelagic  eggs  hatch  with  unpigmented  eyes. 
Oil  globules  may  be  present  in  the  anterior,  ventral,  or  posterior 
part  of  the  yolk  sac.  Multiple  oil  globules  in  early  embryos 
coalesce  into  a  single  large  one  before  hatching  in  Seiipinna 
phasa  (John,  1 95 1  a).  A  spherical  yolk  sac  usually  remains  spher- 
ical although  shrinking  in  size  during  development  (Sardinella 
zunasi),  while  a  yolk  sac  which  is  pointed  posteriorly  may  be- 
come more  rounded  as  yolk  is  utilized  (Coilia  sp.).  Larval  clu- 
peiforms are  slender  and  elongate  with  long  straight  guts.  Series 
of  melanophores  are  variously  arranged  above  and  below  the 
gut  and  along  the  ventral  body  wall.  Subtle  differences  in  pig- 
mentation are  very  useful  for  identifying  co-occuring  larval  clu- 
peoids  prior  to  fin  development.  Larvae  of  Engraulis  mordax. 
Sardinops  sagax.  and  Etrumeus  leres  are  illustrated  for  com- 
parison in  Moser  (1981).  Median  dorsal  melanophores  in  clu- 
peid embryos  migrate,  reaching  their  characteristic  ventral  po- 
sitions soon  after  hatching  (Orion,  1 953a).  In  engraulids,  pigment 
cells  are  presumed  to  migrate  similarly  but  they  don't  become 


pigmented  until  after  hatching.  Melanophores  are  commonly 
present  ventrally  just  anterior  to  the  pectoral  symphysis  in  small 
larvae,  (e.g.,  Opislhonema.  Harengula,  Engraulis,  Sardinops, 
Etrumeus).  During  development  external  rows  of  melanophores 
become  dark  streaks  and  internal  melanophores  may  increase 
in  size  and  number  at  first  but  disappear  or  become  occluded 
at  transformation.  A  thorough  description  of  pigment  devel- 
opment of  laboratory-reared  Opislhonema  oglinum  larvae  com- 
plete with  dorsal,  lateral,  and  ventral  illustrations  is  given  by 
Richards  et  al.  (1974).  Preanal  myomere  number  is  taxonom- 
ically  useful  but  it  does  not  correspond  exactly  with  precaudal 
vertebral  count  in  the  adult  because  of  changes  during  trans- 
formation. Pectoral  fin  buds  and  a  continuous  dorsal-caudal- 
anal  finfold  are  present  at  hatching.  Fin  rays  first  appear  in  the 
caudal  fin  then  in  the  dorsal,  then  the  anal,  next  the  pelvic,  and 
last  the  pectoral  fin.  Ossification  of  fin  rays  proceeds  in  the  same 
order.  A  full  complement  of  fin  rays  is  not  attained  until  trans- 
formation, which  occurs  at  approximately  20  mm  standard  length 
(e.g.,  Harengula  jaguana.  Houde  et  al.,  1974;  Opislhonema  og- 
linum Richards  et  al.,  1974).  Figs.  59  and  60  illustrate  yolk  sac 
larvae  of  herrings  and  anchovies. 

The  most  useful  single  character  for  identifying  larval  clu- 
peiforms is  total  myomere  or  vertebral  number.  Pigment  pat- 
terns are  useful  when  vertebral  counts  overlap.  The  relative 
positions  of  dorsal  and  anal  fins  and  the  length  of  the  gut  can 
be  used  to  separate  clupeids  from  engraulids:  clupeids  have  a 
longer  gut  relative  to  body  length  and  there  is  a  gap  between 
the  posterior  margin  of  the  dorsal  fin  and  the  anterior  margin 
of  the  anal  fin;  engraulids  have  a  shorter  gut  and  tend  to  have 
the  posterior  margin  of  the  dorsal  over  the  anterior  insertion  of 
the  anal  fin.  The  number  of  myomeres  between  dorsal  and  anal 
fins  has  been  used  as  a  taxonomic  character  in  larvae  of  certain 
size  classes  (Houde  and  Fore,  1973)  and  in  clupeid  adults  (Sve- 
tovidov,  1963).  During  metamorphosis  the  position  of  the  gut 
and  the  dorsal  and  anal  fins  shift  forward  relative  to  myomere 
number.  The  dorsal  insertion  moves  10  myomeres  forward  in 
Sardinops  sagax  (Ahlstrom,  1968);  it  moves  eight  myomeres 
in  Harengula  jaguana  (Houde  et  al.,  1974).  The  migration  of 
the  fin  takes  place  at  approximately  the  time  when  the  fin  ray 
number  stabilizes.  The  pelvic  fin  migrates  posleriad  in  Clupea 
harengus  (Lebour,  1921).  Because  of  these  dramatic  changes  in 
morphology  during  development  different  characters  must  often 
be  used  at  different  stages  to  separate  species.  However  some 
morphometric  characters  show  a  small  but  consistent  difference 
between  species  at  all  sizes  as  between  .4losa  pseudoharengus 
and  .4.  aestivalis  (Chambers  et  al.,  1976).  Additional  care  must 
be  taken  when  using  information  from  laboratory-reared  spec- 
imens to  identify  field  samples.  Fin  development  began  at  4 
mm  in  laboratory-reared  Opislhonema  oglinum.  but  was  not 
observed  in  wild-caught  larvae  less  than  7  mm  long  (Richards 
et  al.,  1 974).  Shrinkage  due  to  preservation  and  handling  (Thei- 
lacker,  1980a)  also  presents  problems  when  comparing  devel- 
opment of  larvae  based  on  length.  Meristic  characters  in  Clupea 


Fig.  58.  Eggs  of  Clupeiformes  illustrating  taxonomic  characters:  number  and  size  of  oil  globules,  width  of  perivitelline  space,  degree  of  yolk 
segmentation,  shape,  size.  (A)  Chirocemrus  nudus.  1.56  mm.  Delsman,  1923;  (B)  Etrumeus  leres.  1.35  mm,  Ahlstrom  and  Moser.  1980;  (C) 
Opisthoplerus  tardoore,  0.85  mm,  Bensam,  1967;  (D)  Dussumiena.  1.5  mm,  Delsman,  1925;  (E)  .Anodontostoma  chacunda.  0.92  mm,  Delsman, 
1926c;  (F)  Sardinops  melanosticta.  1.60  mm,  Mito,  1961;  (G)  Coilia.  1.04  mm,  Delsman.  1932b;  (H)  Setipinna  phasa.  1.10  mm,  Jones  and 
Menon,  1950;  (I)  Anchoa  mitchilli.  0.84  x  0.65,  Kuntz,  1914b;  (J)  Engraulis  mordax.  1.40  x  0.74,  Bolin,  1936a;  (K)  Slolephorus  msulans.  1.92 
X  0.69,  Delsman,  1931;  (L)  Slolephorus  indicus  or  commersonii.  1.15  x  0.81,  Delsman,  1931.  All  redrawn  by  J.  Javech. 


McGOWAN  AND  BERRY:  CLUPEIFORMES 


,19 


H 


K 


120 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  59.  Yolk-sac  larvae  of  Clupeidae  and  Chirocentrus  illustrating  taxonomic  characters:  number,  size,  and  position  of  oil  globules;  shape  of 
yolk  sac;  degree  of  segmentation  of  yolk;  preanal  myomeres.  (A)  Sardinella  zunasi.  2.1  \  mm,  Takita,  1966;  (B)  Sardmelta  :unasi,  4.79  mm, 
Takita,  \9(>(>.(C)  Elrumeus  teres.  AM  mm. Mao,  \9(i\:(D)  llisha  elongata.  5.59  mm,  Sha  and  Ruan,  \9i\:{E)  Dussumieria.  3.17  mm,  Delsman, 
1925;  (F)  Chirocentrus  mtdus.  3.79  mm,  Delsman,  1923.  All  redrawn  by  J.  Javech. 


McGOWAN  AND  BERRY:  CLUPEIFORMES 


121 


Fig.  60.  Yolk-sac  larvae  of  Engraulidae  illustrating  taxonomic  characters:  oil  globules,  shape  of  yolk  sac,  yolk  segmentation,  preanal  myomeres. 
(A)  EngrauUs  japomcus.  3.02  mm,  Mito,  1961;  (B)  Coilia.  2.83  mm,  Takita,  1967;  (C)  Coilia.  2.46  mm,  Delsman,  1932b;  (D)  Slolephorus 
msularis.  2.19  mm.  Delsman,  1931;  (E)  Thryssa  hamiltomi.  2.42  mm,  Delsman,  1929a;  (F)  Cetengraulis  mysticetus,  1.99  mm,  Simpson,  1959. 
All  redrawn  by  J.  Javech. 


122 


ONTOGENY  AND  SYSTEM ATICS  OF  FISHES- AHLSTROM  SYMPOSIUM 


harengus  larvae  were  shown  to  be  affected  by  temperature  and 
salinity  (Hempel  and  Blaxter,  1961);  morphometric  characters 
in  Gikhristella  aestuarius  adults  were  found  to  differ  between 
estuaries  with  different  types  of  prey  items  (Blaber  et  al.,  1981). 

There  are  several  characters  which  may  be  useful  in  system- 
atics  when  they  are  described  for  more  clupeiform  species.  The 
melanophores  on  the  caudal  fin  dorsal  and/or  ventral  to  the 
notochord  tip  in  small  larvae  have  been  described  for  a  few 
species.  Harengida  jaguana  has  dorsal  melanophores  only  at 
first,  then  both  dorsal  and  ventral  (Houde  et  al.,  1974).  Opis- 
thonema  oglinum  has  ventral  ones  (Richards  et  al.,  1974).  Sar- 
dinella  brasiliensis.  S.  maderensis  and  S.  zunasi  have  just  ven- 
tral melanophores  but  Sardine/la  rouxi  has  both.  Slight 
differences  in  pigmentation  over  the  brain  and  on  the  mid-dorsal 
and  mid-ventral  postanal  body  midline  have  been  used  to  iden- 
tify scombrid  larvae.  Small  scombrid  larvae  are  otherwise  very 
similar  to  each  other  as  are  clupeoid  larvae.  The  development 
of  free  neuromasts  and  the  lateral  line  has  been  described  for  a 
few  species  (Blaxter  et  al.,  1983).  Development  of  the  swim- 
bladder  and  its  unique  connection  with  the  inner  ear  should  be 
useful  (Hoss  and  Blaxter,  1982).  Ephemeral  basihyal  teeth  were 
observed  on  Opisthonema  oglinum  and  Harengula  jaguana  lar- 
vae (Richards  et  al.,  1974;  Houde  et  al.,  1974).  Two  patterns 
of  nasal  epithelium  cells  have  been  observed  with  scanning  elec- 
tron microscopy  (Yamamoto  and  Ueda,  1 978).  Harengula,  Sar- 
dinops  and  Konosirus  had  one  pattern  while  Etrumeus  (a  clu- 
peid)  had  the  same  pattern  as  Engraulis,  an  engraulid. 

Although  the  eggs  and  yolk-sac  larvae  of  clupeiforms  have 
many  characters  of  potential  systematic  importance,  the  taxo- 
nomic  characters  of  the  older  larvae  (meristics,  fin  position,  and 
pigmentation)  will  tend  to  be  redundant  with  the  same  adult 
characters.  However,  clupeoids  are  easily  reared  in  the  labo- 
ratory so  direct  experimental  evaluation  of  the  polarity  of  adult 
character  states  by  comparative  developmental  studies  is  pos- 
sible. 

Relationships 

The  clupeiform  fishes  are  considered  a  well-defined  mono- 
phyletic  group  based  on  their  unique  otophysic  connection,  the 
caudal  skeleton,  and  other  characters  (Greenwood  et  al.,  1966). 
The  distribution  of  species  within  genera,  genera  within  subfam- 
ilies, and  number  and  taxonomic  rank  of  categories  within  the 
group  are  not  agreed  upon  (Gosline,  1971,  1980;  Miller,  1969; 
Nelson,  1967,  1970,  1973;  Whitehead,  1972,  1973).  J.  S.  Nelson 
(1976)  lists  the  families  Chirocentridae,  Denticipitidae,  Clu- 
peidae,  and  Engraulidae.  He  gives  seven  subfamilies  of  herrings 
(Dussumieriinae,  Clupeinae,  Pellonulinae,  Alosinae,  Doroso- 
matinae,  Pristigasterinae,  and  Congothrissinae)  and  two 
subfamilies  of  anchovies  (Engraulinae  and  Coilinae).  Spralel- 
loides  is  separated  from  the  Dussumieriinae  and  given  subfamily 
rank  by  Whitehead  (1972.  1973).  Jenkinsia  is  the  western  At- 
lantic spratelloidin. 

Based  on  the  gill  arches  Nelson  (1967)  concluded  that  the 
Dussumieriinae  (including  Spratelloides  and  Jenkinsia)  were  the 
most  primitive  clupeid  family;  the  Pristigasterinae  were  also 
primitive  but  with  distinctive  specializations;  the  Clupeinae  were 
more  advanced,  but  linked  to  the  Dussumieriinae  by  Clupea 
and  Sprattus;  the  Alosinae  and  Dorosomatinae  were  closely 
related  and  perhaps  both  derived  from  the  Clupeinae;  and  the 
Pellonulinae,  lacking  the  specializations  of  the  Alosinae  and 
Dorosomatinae,  most  resembled  the  Clupeinae.  Expanding  his 


study  of  gill  arches  in  the  Clupeidae  to  the  hyobranchial  ap- 
paratus in  the  Clupeiformes,  Nelson  (1970)  divided  the  order 
into  the  superfamilies  Chirocentroidae,  Engrauloidae.  Pristi- 
gasteroidae,  and  Clupeoidae.  The  Clupeoidae  were  suggested  to 
consist  of  two  families:  the  Clupeidae  composed  of  the  Dus- 
sumieriinae, Pellonulinae,  and  Alosinae  in  part;  and  the  Do- 
rosomatinae composed  of  the  Dorosomatinae  plus  Hilsa  from 
the  Alosinae  and  Harengida  and  Sardinella  from  the  Clupeinae. 
Sardina  and  Alosa  were  aligned  with  Clupea,  Polamalosa,  and 
Etrumeus  in  his  tree  depicting  relationships  of  representative 
genera  (Nelson,  1970:  Fig.  1 1). 

Whitehead  (1972,  1973)  acknowledged  that  radical  changes 
in  clupeid  classification  could  be  expected  but  retained  the 
subfamilies  Dussumieriinae,  Spratelloidinae,  Clupeinae,  Pel- 
lonulinae, Alosinae,  Dorosomatinae,  and  Pristigasterinae  in  his 
works  which  were  chiefly  concerned  with  the  identification  of 
genera  and  species. 

The  most  recent  comprehensive  work  is  that  of  Wongratana 
(1980)  on  the  Clupeidae  and  Engraulidae  of  the  Indo-Pacific. 
He  examined  over  14,000  specimens  and  considered  many  me- 
ristic  and  morphological  characters  including  gill  rakers,  epi- 
branchial  organs,  predorsal  bones,  caudal  osteology,  circumor- 
bital  bones,  gut  form,  the  gas  bladder,  scale  striae,  and  the  patterns 
of  scale  distribution  on  the  body.  No  numerical,  cladistic,  or 
phenetic  analyses  were  done.  Taxonomic  characters  were  dis- 
cussed with  respect  to  apparent  evolutionary  trends  and  relative 
importance.  Wongratana  retained  the  subfamilies  of  Whitehead 
(1972).  The  Spratelloidinae  were  diagnosed  by  a  bony  process 
on  the  6th  and  1 2th  principal  caudal  rays.  Spratelloides  is  also 
unique  among  Indo-Pacific  clupeids  in  having  a  single  epural. 
Jenkinsia,  the  spratelloidin  in  the  Western  Atlantic,  also  has  a 
single  epural  (Hollister,  1936).  The  Alosinae  and  Dorosomatin- 
ae were  kept  separate  and  the  Pristigasterinae  were  accorded 
subfamily  status  although  considered  quite  distinct  from  the 
other  clupeids.  The  Dussumieriinae  and  Pellonulinae  were  con- 
sidered the  most  primitive  groups,  the  Alosinae  and  Doroso- 
matinae the  most  advanced,  and  the  Spratelloidinae  and  Clu- 
peinae were  considered  intermediate.  Within  the  anchovies,  the 
Coiliinae  have  one  epural  while  the  Engraulinae  have  two  {En- 
graulis) or  three  (Papuengraulis).  The  Coiliinae  were  considered 
primitive  relative  to  the  Engraulinae  although  specialized  in 
many  respects. 

Wongratana  ( 1 980)  found  that  the  number  of  predorsal  bones 
varies  from  one  to  thirty  in  the  clupeids  and  engraulids  (Chi- 
rocenlrus  has  none).  Some  engraulids  and  pellonulins  have  a 
gap  between  the  posterior  predorsal  bone  and  the  first  dorsal 
pterygiophore  which  he  interpreted  as  evidence  that  the  dorsal 
fin  has  migrated  posteriad  during  evolution.  It  would  be  inter- 
esting to  compare  the  patterns  of  dorsal  bones  and  the  anteriad 
migration  of  the  dorsal  fin  during  larval  metamorphosis.  The 
"dorsal  scutes"  of  Clupanodon  ihrlssa  were  found  to  be  the 
exposed  tips  of  predorsal  bones  (Wongratana,  1980).  The  only 
double-armored  herrings  known  now  are  Polamalosa  and  Hy- 
perlophus  in  the  Pellonulinae,  and  Elhmidium  in  the  Alosinae. 
Dorsal  scutes  are  interesting  because  they  occurred  in  herring- 
like fossils  (Diplomystus,  Knightia,  and  Gasteroclupea)  which 
resemble  pristigasterins  (Nelson  1970). 

Because  he  examined  so  many  species  from  such  a  wide  area 
Wongratana  (1980)  was  able  to  clear  up  many  nomenclatural 
questions  and  to  correct  previous  misidentifications  which  had 
been  based  on  limited  material.  He  also  described  24  new  species 


McGOWAN  AND  BERRY:  CLUPEIFORMES 


123 


(Wongratana,  1 983)  and  provided  keys  to  all  Indo-Pacific  species 
(Wongratana,  1980).  However  no  direct  comparison  between 
his  classification  and  that  of  Nelson  (1967,  1970,  1973)  is  pos- 
sible because  he  only  examined  Indo-Pacific  material  while  Nel- 
son included  West  African  and  New  World  material. 

Evidence  from  eggs  and  larvae 

There  are  two  major  problems  with  using  characters  of  eggs 
and  larvae  to  criticize  classifications  based  on  adult  characters. 
First,  the  planktonic  stages  of  fishes  are  exposed  to  different 
selective  pressures  than  the  adults  so  they  may  show  patterns 
of  specializations  for  planktonic  life  which  are  not  congruent 
with  the  distribution  of  adult  character  states.  Second,  relatively 
few  genera  of  clupeiform  fishes  have  had  the  eggs  or  larvae 
described  for  even  one  species  in  the  genus.  The  first  problem 
must  be  dealt  with  the  same  as  any  character  complex  in  a  group 
with  more  than  one  character  complex.  More  knowledge  of  the 
ecology  of  the  larvae  in  the  sea  would  indentify  species  with 
different  funtional  requirements  for  their  larvae.  The  second 
problem  may  be  resolved  by  using  the  available  evidence  in  a 
parsimonious  fashion. 

Eggs  and  young  larvae  are  similar  within  genera.  Seven  species 
of  Sardine/la  (Table  25)  all  have  moderately  sized  clupeid-type 
eggs  with  a  wide  perivitelline  space  and  a  single  oil  globule.  The 
egg  described  by  Takita  (1966)  and  Chang  et  al.  (1981)  as  that 
oi Harengida  ziinasi  is  similar.  Wongratana  ( 1 980)  places  zunasi 
in  Sardinclla. 

Within  subfamilies  there  is  little  apparent  consistency  in  egg 
morphology  among  genera.  Etruineus  has  no  oil  droplet  but 
Dussumieria  does.  Brevoortia  has  eggs  1.3  mm  or  larger  with  a 
single  oil  globule;  HHsa  kelee  has  1.00-1.07  mm  eggs  with  sev- 
eral small  oil  droplets.  Clupea  has  demersal  adhesive  eggs  while 
Sprattus  has  pelagic  eggs  with  a  small  perivitelline  space.  The 
Indo-Pacific  pristigasterin  species  of  Ilisha  have  large  eggs  with 
adhesive  coatings  and  a  single  large  oil  globule  but  Opislhopterus 
tardoore  and  the  eastern  Pacific  O.  dovii  have  small  eggs  with 
small  perivitelline  spaces  and  no  oil  droplets. 

The  functional  significance  of  egg  characters  is  unknown.  Sep- 
arate lineages  within  the  group  which  have  radiated  into  several 
habitats  could  show  parallel  adaptations  such  as  oil  droplets  for 
buoyancy  or  nutrition,  adhesive  coating  for  retention  nearshore 
or  demersally.  and  egg  size  as  a  trade-off  between  broadcasting 
and  parental  investment.  Alternatively,  different  types  of  eggs 
within  taxonomic  categories  could  also  support  splitting  the 
category.  The  anchovy  genus  Stolephorus  contains  species  with 
eggs  which  range  from  oval  with  no  oil  globule  to  varying  degrees 
of  eccentricity  with  an  oil  droplet,  to  unusually  shaped  eggs  with 
knobs  on  one  end  (Delsman,  1931).  Nelson  (1983)  separated 
Stolephorus  into  two  groups,  a  Stolephorus  group  with  1 3  species 
and  an  Encrasicholina  (new  usage)  group  of  5  species  which  he 
considered  more  closely  related  to  New  World  anchovies  than 
to  the  1 3  Stolephorus  species.  The  three  Encrasicholina  species 
whose  eggs  are  known  have  an  oval  egg  without  a  knob.  One 
of  the  three,  E.  hetcrolobus.  was  reported  by  Delsman  (1931) 
to  have  a  small  oil  droplet  and  to  be  relatively  more  abundant 
near  shore  than  Stolephorus  zolingeri.  The  other  two,  E.  pur- 
purcus  and  E.  punctifer  (^buccanceri,  Strasburg,  1960;  =zolin- 
geri.  Delsman,  1931),  occur  in  Hawaii  and  neither  has  an  egg 
with  an  oil  droplet.  New  World  anchovies  don't  have  eggs  with 
knobs  or  oil  droplets;  therefore,  the  evidence  from  eggs  supports 


Nelson's  revision  and  in  addition  provides  some  basis  for  zoo- 
geographic  speculation. 

Whether  the  pristigasterins  should  be  given  equal  rank  with 
the  clupeids  and  engraulids  cannot  be  answered  with  the  avail- 
able ontogenetic  information.  There  are  two  very  different  egg 
types  in  the  group,  small  with  small  perivitelline  space  and  large 
with  gelatinous  coating,  both  of  which  could  be  considered  spe- 
cializations. Etrumeus.  Jenkmsia.  Spratelloides,  Clupea.  Sprat- 
tus, and  Potamalosa  were  linked  based  on  a  foramen  in  the 
fourth  epibranchial  (Nelson,  1970).  Eggs  of  Spratelloides  and 
Clupea  are  both  demersal  and  adhesive.  The  planktonic  eggs  of 
Etrumeus  and  Sprattus  both  have  narrow  perivitelline  spaces 
and  lack  oil  globules.  Eggs  of  Potamalosa  and  Jenkinsia  are 
unknown.  Jenkinsia  is  related  to  Spratelloides  and  has  demersal 
larvae  (Powles,  1977)  so  it  may  have  demersal  eggs.  The  de- 
velopmental osteology  of  these  genera  could  be  studied  to  de- 
termine if  the  shared  foramen  is  phylogenetically  homologous. 
The  egg  of  Anodontostoma,  Dorosominae,  is  similar  to  eggs  of 
the  Alosinae  in  that  it  has  multiple  small  oil  droplets.  Otherwise 
both  the  Alosinae  and  Dorosomatinae  contain  species  with  de- 
mersal adhesive  eggs  and  species  with  buoyant  planktonic  eggs. 

Other  suggestions  of  Nelson  (1970)  that  Sardinclla.  Opistho- 
nema.  Harengula.  and  Herklotsichthys  should  be  placed  with 
the  Dorosomatinae  and  Sardina  and  Sardinops  with  the  Alo- 
sinae and  then  that  the  Alosinae  and  Dorosomatinae  should  be 
combined  leaving  just  Clupeinae  and  Dorosomatinae  cannot  be 
critically  evaluated  with  existing  ontogenetic  data.  These  hy- 
potheses could  be  tested  by  comparing  the  osteological  devel- 
opment of  the  characters  used  by  Nelson,  augmented  by  other 
early  life  history  characters. 

Relationships  of  the  Clupeiformes 
Greenwood  et  al.,  (1966)  placed  the  Clupeomorpha  and  Elo- 
pomorpha  together  in  their  Division  One  but  gave  serious  con- 
sideration to  the  possibility  that  the  Clupeomorpha  should  be 
recognized  as  a  separate  division.  Using  information  on  the  gut 
and  lower  jaw.  Nelson  (1973)  proposed  that  the  Clupeomorpha 
were  distinct  from  the  Elopomorpha  but  perhaps  related  to  the 
non-osteoglossomorph  teleosts.  Gosline  (1980)  concluded  that 
the  clupeiform  fishes  should  be  grouped  with  the  elopiform,  the 
salmoniform,  gonorynchiform,  and  ostariophysine  fishes;  sep- 
arated on  one  side  from  the  osteoglossiform  fishes  and  from  the 
iniomous— acanthopterygian  teleosts  on  the  other.  His  conclu- 
sions were  based  on  five  morphological  character  complexes: 
the  caudal  skeleton,  the  swim  bladder-ear  connection,  the  post- 
cleithrum,  the  structures  associated  with  pectoral  fin  movement, 
and  the  various  types  of  premaxillary  movements  and  jaw  pro- 
trusion (Gosline,  1980). 

Gosline  (1980)  considered  the  elopomorphs  to  be  an  early 
offshoot  from  a  basal  lower  teleostean  group.  He  considered  the 
gonorynchiforms  and  ostariophysines  to  be  more  closely  related 
to  each  other  than  to  the  clupeiforms.  A  clupeiform— osteo- 
glossiform link  has  also  been  mentioned  (Greenwood,  1973).  J. 
S.  Nelson  (1976),  who  put  the  superorders  Clupeomorpha  (Clu- 
peiformes) and  Elopormorpha  (Elopiformes,  Albuliformes,  An- 
guilliformes)  into  Division  Taeniopaedia,  slated  succinctly  that 
"the  relation  of  superorders  recognized  here  is  poorly  known 
and  they  are  essentially  "loose  ends."  "  Lauder  and  Liem  (1983) 
provisionally  follow  Nelson  (1970)  for  most  groups  within  the 
Clupeomorpha  but  represent  the  interrelationships  of  clupeoid 
lineages  as  an  unresolved  polychotomy.  Lauder  and  Liem  (1983) 


124 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  25.    Sources  of  Early  Life  History  Information  for  Clupeiformes.  Reviews  and  readily  available  works  with  superior  illustrations 

are  cited  rather  than  original  descriptions  in  some  cases. 


Genus  species 


Eggs 


Lar- 
vae 


Ju- 

ven-       Mor-     Mens- 

lies     phology     tics 


Pig- 
menta-    Oste- 
Fins       lion       ology 


Keys 
or  com- 
pan- 
sons  Wild- 

Fe-      Spawn-  Spawn-     with     Reared  caught 
cun-         ing  ing       others     speci-     speei- 

dity      region    season   species    mens      mens 


References 


X 
X 
X 

X 
X 


X 
X 


X 
X 


X 
X 
X 

X 
X 


X 
X 
X 
X 
X 

X 
X 

X 
X 


Chirocentnis  dorab  X        X 

Chirocentrus  nudus  X        X 

Sardinella  zunasi  X        X 


Sardinella  jussieui  XXX 

Sardinella  aurila  XXX 

Sardinella  albella  X        X 

Sardinella  fimbriata  X        X 

Sardinella  brachysoma  X        X 
Sardinella  brasiliensis  X        X 

Sardinella  longiceps  X        X 
Sardinella  maderensis  X 

Sardinella  rouxi  X 

Clupea  harengus  XXX 

Clupea  pallasi  XXX 
Clupea  bentincki  X        X 

Spratlus  sprattus  XXX 

Sprattus  antipodurn  X 


Elrumeus  teres  X        X  X        X 

Elrumeus  whiteheadi  XX  XX 

Dussumieria  sp.  XX  XX 

Spratelloides  delicatulus  X        X  X        X        X 

Jenkinsia  lamprolaenia  X  X        X        X 

Konosirus  punctatus  XX  XX 

Anodontostoma  chacunda  X        X  X        X        X 

Dorosoma  pelenense  X        X  X        X        X 


Amblygaster  leiogasler  XX  XX 

Amblygaster  sirm  X        X 

Escualosa  thoracata  XX  X 

Opisthonema  lihenate  X 

Opisthonema  oglinum  X        X  X        X        X 

Harengula  jaguana  X        X  X        X        X 


Harengula  peruana  X 

Sardinops  sagax  caerulea  X        X 

Sardinops  sagax  musica  X        X 


Sardinops  melanosticta  XX  X 

Sardinops  ocellata  X        X  X        X        X 

Sardina  pilchardus  X        X  X        X        X 

Lile  stolifera  X 

Dorosoma  cepedianum  X        X  X        X        X 


Hilsakelee  XX  XX 

Tenualosa  itisha  XX  XX 

Alosa  sapidissima  X        X  X        X        X 

Alosa  pseudoharengus  X        X  X        X        X 


X 
X         X 


X 
X 
X 
X 


X 

X 


X 
X 


X 
X 


X 
X 


X 

X 


X 
X 
X 
X 


X  X 
X  X 
X         X 


X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 
X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

XXX 
XXX 

X 

X        X 


X 

X 


X 

X 

X 

X 
X 

X 

X 

X 

X 
X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X  X 
XXX 
XXX 


X 

X        X 

X 
X  X 
X         X 


X 

X 


X 
X 


Delsman,  1923,  1930b 
Delsman,  1923,  1930b 
Takita,  1966;  Chang 

et  al.,  1981 
Bensam,  1970 
Jones  et  al.,  1978; 

Houde  and  Fore,  1973 
Delsman,  1933b 
Delsman,  1926 
Delsman,  1926 
Matsuura.  1975 
Nair,  1960 
Conand,  1978;  Conand 

and  Fagetti,  1971 
Conand,  1978 
Jones  et  al..  1978; 

Fahay,  1983 
Wang,  1981 
Orcllana  and  Balbontin, 

1983 
Saville,  1964 
Russell,  1976;  Robert- 
son, 1975a 
Mito,  1961a 
Brownell,  1979;  OToole 

and  King,  1974 
Delsman,  1925 
Uchidaet  al.,  1958; 

Miller  et  al.,  1979 
Powles,  1977 
Mito,  1961a 
Delsman,  1933a 
Shelton  and  Stephens, 

1980;  Jones  etal., 

1978 
Delsman,  1926b 
John,  1951a 
Delsman,  1926c,  1934a 
Peterson,  1956 
Richards  et  al..  1974; 

Jones  et  al.,  1978 
Houde  etal..  1974; 

Gorbunova  and 

Zvyagina  1975; 

Houde  and  Fore,  1973 
Peterson,  1956 
Ahlstrom,  1943;  Miller, 

1952 
Santander  and  de  Castillo, 

1977;  Orellanaand 

Balbontin,  1983 
Mito,  1961a 
Brownell,  1979;  Louw 

and  OToole,  1977 
SaviUe,  1964;  Russell, 

1976 
Peterson,  1956 
Shelton  and  Stephens, 

1980;  Jones  etal., 

1978;  Cooper,  1978 
Rao,  1973 
Kulkami,  1950 
Bainbridge,  1962; 

Jones  etal.,  1978 
Jones  etal.,  1978; 

Chambers  et  al.,  1976 


McGOWAN  AND  BERRY:  CLUPEIFORMES 


125 


Table  25.    Continued. 


Genus  species 


Eggs 


Ur- 
vae 


ven-       Mor-     Mens- 

iles     phology     tics        Fins 


Pig-  Fe- 

menta-    Oste-      cun- 

tion       ology       dity 


Keys 
or  corn- 
pan - 

sons  Wild- 

Spawn-  Spawn-     with     Reared   caught 
ing  ing       others    speci-     speci- 

region    season  species    mens     mens 


References 


Alosa  mediae ns 

Alosa  aestivalis 
Caspialosa  sp. 
Elhmalosa  fimbriala 
Brevoortia  aurea 

Brevoortia  patronus 
Ethmidium  macutata 

Gilchristella  aesluanus 
Laevisculella  dekimpei 
PeUonula  vorax 

Ilisha  elongata 

Ilisha  melasloma 
IHsha  afncana 
Ilisha  furthi 

Neoopislhopterus  tropicus 
Opisthopterus  tardoore 
Opisthopterus  do\i 
Opisthopterus  equatorialis 
Odontognathus  panamensis 
Anchoa  ischana 
Anchoa  panamensis 
A  nchoa  curta 
Anchoa  tucida 
Anchoa  naso 
Anchoa  exigua 
A  nchoa  arenicola 
Anchoa  marinii 
Anchoa  hepsetus 
Anchoa  mitchilli 
Anchovia  macrolepidota 
Engraulis  japo  nicus 

Engraulis  eur\'slole 
Engraulis  anchoita 
Engraulis  inordax 


Engraulis  encrasicolus 
Engraulis  ringens 


Slolephorus  purpureus 
Stolephorus  buccaneeri 

Stolephorus  heterolobus 
Slolephorus  tri 

Thryssa  hamiltonii 
Thry'ssa  sp. 
Lycengraulis  poeyi 
Lycengraulis  gross  idens 
Celengraulis  mysticetus 
Setipinna  melanochir 
Selipmna  taty 
Setipinna  phasa 
Heterothrissa  breviceps 
Coilia  sp. 
Coilia  sp. 


X 
X 
X 
X 


X 

X 

X 
X 


X 
X 


X 

X 


X 
X 


XXX 
XXX 
X        X 


X 
X 


X 

X 


XXX 
X         X 


X 
X 


X 
X 


X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 
X 

X 


X         X         X         X         X         X        X 
X         X        X         X         X  X 

X         X 
X         X 

X        X 

X         X 

X 

X         X 

X         X 

XX  X 

XX  X 

XX  X 

XX  X 

XXX  X 

XX  X 


X 
X 

X 
X 


X 

X 

X 

X 

X 

X 

X 

X 
X 


X 
X 

X 
X 


X 
X 
X 
X 
X 


X        X      Jones  etal.,  1978; 

Chambers  et  al.,  1976 
X        X      Jones  etal.,  1978 

Pertseva,  1936 
X        X       Bainbndge,  1961 
X  Conand  1978;  de 

Ciechomski.  1968 
X      Houde  and  Fore,  1973 
X      Orellana  and  Balbontin, 

1983 
X       Brownell,  1979 
X      Conand,  1978 

Bainbndge.  1962; 
Conand,  1978 
X        X       Delsman,  1930a;  Uchida 

etal..  1958 
X  Delsman,  1930a 

X      Dessier,  1969 
Peterson.  1956 
Peterson,  1956 
X  Bensam,  1967 

Peterson,  1956 
Peterson,  1956 
Peterson,  1956 
Peterson,  1956 
Peterson,  1956 
Peterson,  1956 
Peterson,  1956 
Peterson,  1956 
Peterson,  1956 
Peterson,  1956 
de  Ciechomski,  1968 
Jones  etal.,  1978 
Jones  etal..  1978 
Peterson,  1956 
Mito,  1961a;  Brownell, 

1979;  Russell,  1976 
Jones  etal.,  1978 
de  Ciechomski,  1965 
Bolin.  1936a;  Ahlstrom, 
1965;  Ahlstrom, 
unpublished 
X      D'Ancona,  1931a;  Saville, 
1964;  Marchal,  1966 
X        X       Orellana  and  Balbontin, 
1983;  Fischer,  1958b; 
Einarsson  and  Rojas 
de  Mendiola,  1963 
X       Miller  etal.,  1979 
X      Delsman,  1931;  John, 

1951a 
X       Delsman,  1931 
X       Delsman,  1931;  John, 

1951a 
X       Delsman,  1929a 
X      John, 1951a 
X       Peterson,  1956 
X       Phonlor,  1978 
X      Simpson,  1959 
X      Delsman,  1932a 
X      Delsman,  1932 
X      Jones  and  Menon,  1950 
X      Delsman,  1932a 
X  Takita,  1967 

X  Delsman,  1932b 


X 
X 
X 
X 
X 
X 
X 
X 
X 
X 
X 


X 
X 
X 


126 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES -AHLSTROM  SYMPOSIUM 


place  the  clupeomorpha  nearer  to  the  next  most  advanced  clade, 
the  Euteleostei,  than  to  the  next  least  advanced  clade,  the  Elo- 
pomorpha. 

Evidence  from  eggs  and  larvae 

Relevant  ontogenetic  evidence  concerning  the  relationships 
of  the  Clupeiformes  is  meager.  Elopiform  eggs  are  unknown. 
Anguilliform  eggs  resemble  clupeid  eggs  in  having  perivitelline 
spaces,  segmented  yolks,  and  may  have  oil  droplets.  Eel  eggs 
can  be  much  larger  than  herring  eggs:  5.5  mm  diameter  in  A/m- 
raena.  2.43  mm  in  an  anguillid  (Ahlstrom  and  Moser,  1980). 
Osteoglossomorphs  have  pelagic  or  demersal  eggs  which  may 
be  0.5-4.0  mm  in  diameter,  may  be  dark  blue,  and  may  have 
a  very  wide  perivitelline  space  as  in  Hiodon  (Breder  and  Rosen, 
1966).  The  coincidence  of  demersal  adhesive  eggs  in  both  the 
osteoglossomorphs  and  the  Dorosomatinae  is  extremely  un- 
likely to  be  a  shared  derived  character  from  a  common  ancestor. 
Clupeid  and  anguillid  eggs  are  considered  unspecialized  relative 
to  eggs  of  the  higher  teleosts  (Ahlstrom  and  Moser,  1980).  Very 
little  else  may  be  said.  Perhaps  electron  microscopy  will  reveal 
patterns  of  chorion  sculpturing  which  will  be  informative. 

The  larvae  of  Clupeiformes  are  unspecialized  and  undergo  a 
fairly  uneventful  metamorphosis.  The  migration  of  the  dorsal 
fin  during  transformation  also  occurs  in  the  elopiforms.  The 
larva  of  Chanos.  a  primitive  gonorynchiform  (Fink  and  Fink, 
1981),  superficially  resembles  clupeids  or  engraulids  but  appar- 
ently does  not  have  the  same  migration  of  the  dorsal  fin  (Rich- 
ards, this  volume). 

If  the  Elopomorpha  and  the  Clupeomorpha  share  a  common 
ancestor  it  is  possible  that  the  Clupeomorpha  retained  the  un- 
specialized, rapidly  developing  larvae  while  the  adults  evolved 
towards  a  specialized  schooling  planktivore  body  plan.  The  lep- 
tocephalus  found  in  the  elopiforms,  albuliforms,  and  anguilli- 
forms  could  have  evolved  for  dispersal  or  to  reduce  predation 
or  to  take  advantage  of  larval  drift  the  way  Angntlla  does  in  the 
North  Atlantic  and  the  way  herring  do  in  the  North  Atlantic 
with  their  circuit  of  migration  (Cushing,  1977).  The  leptoceph- 
alus  could  have  arisen  in  the  common  ancestor  of  anguilliforms 
and  elopiforms  or  in  parallel,  in  response  to  the  same  selective 
influence,  after  the  adult  eels  had  begun  their  divergence  from 
the  still  unspecialized  elopiform  fishes.  The  leptocephalus  is 
considered  a  specialized  character  by  Forey  (1973a),  who  sug- 
gested that  it  arose  before  the  elopid-albulid  dichotomy.  Trans- 
forming elopoid  leptocephali  resemble  transforming  clupeiform 
larvae  (A/e^a/ops— Harrington,  1958:  Plate  1;  f/ops— Sato  and 
Yasuda,  1980:  Fig.  1;  ,4//)j//a-Hildebrand,  1963b:  Fig.  23). 


The  egg  and  larval  evidence  thus  is  consistent  with  a  rela- 
tionship between  the  Elopomorpha  and  the  Clupeomorpha  based 
on  primitive  characters  but  is  not  helpful  in  aligning  this  Di- 
vision (J.  S.  Nelson's  usage,  1976)  closer  to  any  other. 

Summary  and  recommendations 

The  eggs  and  early  larval  stages  of  the  Clupeiformes  provide 
many  taxonomic  characters  with  potential  value  for  testing  phy- 
logenetic  hypotheses.  Most  of  the  discrete  characters,  such  as 
number  of  oil  globules,  have  more  than  two  states  and  the 
continuous  characters,  such  as  degree  of  egg  eccentricity,  have 
at  least  a  moderate  range  of  values.  Although  the  fraction  of 
species  whose  eggs  and  larvae  have  been  described  is  low  and 
the  descriptions  are  uneven  in  quality  and  not  distributed  uni- 
formly among  taxa,  egg  and  larval  characters  appear  consistent 
within  genera.  Within  nominal  subfamilies  they  are  not  consis- 
tent, but  the  subfamilies  show  parallel  trends  in  adult  characters 
and,  in  addition,  the  distribution  of  genera  in  higher  taxa  is  not 
yet  agreed  upon  by  all  workers. 

Most  descriptions  of  clupeiform  larvae  have  been  to  enable 
identification  of  regional  species.  Differences  between  larvae 
usually  involve  subtle  features  of  pigmentation  or  morphome- 
try, or  counts  of  meristic  characters  which  converge  with  the 
meristics  of  the  adult.  Phylogenetically  significant  characters 
such  as  ephemeral  dentition,  osteological  development,  and  the 
comparative  ontogeny  of  characters  used  in  the  taxonomy  of 
the  adults  are  rarely  mentioned. 

Future  descriptions  of  eggs  and  larvae  should  address  system- 
atic characters  as  well  as  those  needed  for  identification.  Eggs 
and  larvae  of  many  species  should  be  redescribed  to  give  com- 
plete series  through  metamorphosis.  Ontogenetic  characters 
should  be  used  in  revisions  of  the  group.  Classifications  of  the 
Clupeiformes  which  are  based  on  just  a  few  characters  should 
be  tested  by  comparing  the  ontogeny  of  those  characters  because 
there  are  many  apparently  parallel  trends  in  the  group.  Addi- 
tional studies  of  the  physiology  and  ecology  of  the  eggs  and 
larvae  should  be  done  to  determine  the  functional  significance 
of  observed  characters.  It  would  also  be  useful  to  perform  quan- 
titative phenetic  and  cladistic  analyses  now  of  the  Clupeiformes 
for  those  regions  or  taxa  for  which  information  is  already  fairly 
complete. 

National  Marine  Fisheries  Service,  Southeast  Fisheries 
Center,  75  Virginia  Beach  Drive,  Miami,  Florida  33149. 


Ostariophysi:  Development  and  Relationships 


L.  A.  FUIMAN 


OSTARIOPHYSI,  as  regarded  here,  include  all  fishes  whose  3  orders,  about  55  families,  and  more  than  5,000  species,  there- 
four  or  five  anteriormost  vertebrae  are  modified  to  form  by  accounting  for  over  70%  of  the  world's  freshwater  fish  species, 
an  otophysic  connection,  the  Weberian  apparatus  (Rosen  and  Oslariophysans  occupy  most  freshwater  habitats  worldwide,  from 
Greenwood,  1970).  These  primarily  freshwater  fishes  comprise  torrential  Himalayan  streams  to  still  tropical  lakes,  as  well  as 


FUIMAN:  OSTARIOPHYSI 


127 


Fig.  6 1 .  Egg  of  Clenolucius  hujela  ( 1 8  hours  poslfertilization)  show- 
ing the  membranous  pedestal  by  which  the  egg  attaches  to  plants.  Pho- 
tograph by  H.-J.  Franke. 


coastal  marine  waters  (the  latter  by  a  few  characids,  cyprinids, 
and  aspredinids,  as  well  as  all  ariid  and  plotosid  catfishes).  The 
presence  of  a  Webenan  apparatus  has  overshadowed  the  suite 
of  remaining  diagnostic  characters  for  the  group  which  includes 
an  axe-shaped  endochondral  portion  of  the  metapterygoid,  an- 
teriorly bifurcate  pelvic  girdle,  second  hypural  fused  to  the  com- 
pound terminal  centrum,  and  elongate  olfactory  tracts  (all  de- 
tailed by  Fink  and  Fink,  1981).  Additional  characters  include 
a  pheromone-mediated  alarm  reaction  and  homy  dermal  pro- 
jections called  unculi  (Roberts,  1982b). 

According  to  the  classification  of  Fink  and  Fink  (1981),  the 
orders  of  Ostariophysi  (their  Otophysi)  are:  Cypriniformes, 
Characi formes,  and  Siluriformes  (the  latter  including  Siluroidei 
and  Gymnotoidei).  Cypriniforms  (with  over  1,800  species  in  5 
families)  uniquely  share  peculiarities  of  the  following:  kineth- 
moid  bone,  palatine-mesopterygoid  articulation,  fifth  cerato- 
branchial,  and  lateral  process  of  the  second  vertebral  centrum. 


They  lack  jaw  teeth  and  an  adipose  fin.  They  are  found  in  North 
America,  Eurasia,  and  Africa.  Characiforms  (comprising  at  least 
1,000  species  in  14  families)  are  characterized  by  multicuspid 
teeth,  a  prootic  foramen,  dorsomedial  opening  in  the  posttem- 
poral  fossa,  enlarged  lagenar  capsule,  and  a  gap  between  the 
compound  terminal  centrum  and  hypural  1.  They  occur  in  Af- 
rica, South  America,  and  southernmost  North  America.  Silu- 
roids  (with  about  2,000  species  in  3 1  families)  are  distributed 
nearly  worldwide.  Although  quite  diverse  morphologically,  they 
commonly  lack  scales  and  several  bones  (including  the  sym- 
plectic,  subopercle,  and  separate  parietals).  They  show  consid- 
erable fusion  of  portions  of  the  first  five  vertebrae  and  pectoral 
and  dorsal  fin  rays.  The  electrogenic  gymnotoids  are  character- 
ized by  an  extremely  long  anal  fin  and  substantial  reductions  or 
losses,  such  as  the  loss  of  dorsal  and  pelvic  fins,  and  palatine 
and  ectopterygoid  bones.  They  are  confined  to  South  America 
and  southernmost  North  America. 

Development 

Knowledge  of  the  early  life  history  stages  of  ostariophysans 
is  rather  spotty  and  concentrated  on  fishes  from  a  few  geographic 
regions.  Major  descriptive  works  cover  portions  of  the  Soviet 
Union  (Kryzhanovskii,  1949;  Kryzhanovskii  et  al.,  1951;  Kob- 
litskaia,  1981),  Japan  (Okada,  1960;  Nakamura,  1969),  and  the 
United  States  (Jones  et  al.,  1978;  Snyder,  1981;  Auer,  1982; 
Fuiman  et  al.,  1983).  Most  of  these  works  concentrate  on  cy- 
priniforms. Additional  descriptive  data  are  available  as  indi- 
vidual papers  on  Indian  major  carps  (Cyprinidae)  and  Indian 
siluroids  (reviewed  by  Jhingran,  1975).  African  and  South 
American  ostariophysan  eggs  and  larvae  remain  little  known. 

Of  the  six  families  of  cypriniforms,  nothing  is  known  of  the 
eggs  and  larvae  of  the  families  with  fewest  species,  Gyrinocheili- 
dae  and  Psilorhynchidae.  Catostomids  are  known  well.  Cypri- 
nids, cobitids,  and  homalopterids  are  known  to  a  lesser  degree. 
Scattered  notes  are  available  for  nine  characiform  families  but 
only  a  few  descriptions  of  ontogeny  exist.  Brief  descriptions  of 
larvae  of  representatives  from  seven  families  of  siluroids  are 
available,  and  notes  on  eight  additional  families  exist.  Photo- 
graphs of  larvae  of  two  gymnotoids.  Eigenmannia  virescens  anA 
Aptewnotus  leptorhynchus  are  published  (Kirschbaum  and 
Westby,  1975;  Kirschbaum  and  Denizot,  1975;  Kirschbaum, 
1984)  but  without  morphological  descriptions.  Most  informa- 
tion on  ostariophysan  larvae  deals  with  external  morphology. 
Osteological  studies  are  few  (Bertmar,  1959;  Hoedeman,  1960a- 
d). 

Eggs 

Ostariophysan  eggs  vary  considerably  in  their  morphology 
and  the  habitat  they  occupy.  Most  are  spherical,  demersal,  1  to 
5  mm  in  diameter,  with  pale  yellow,  somewhat  granular  yolk 


Table  26.    Larval  Characters  of  Major  Groups  of  Ostariophysans. 


Cypnniformes 


Characiformes 


Siluroidei 


Gymnotoidei 


Size  at  hatching  (mm  XL) 

Yolk-sac  shape 

Gap  between  yolk  sac  and  anus 

Barbels: 

Presence 

Timing  of  development 

Size  at  finfold  absorption  (mm  TL) 


2-10 

pyriform  or  tubular 

absent 

present  or  absent 

late  or  early 

15-25 


2-5 
elliptical 
present 

absent 
10-20 


3-8 
elliptical 
present 

present 
early 
11-23 


elliptical 
absent 

absent 
15 


128 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


A 


FUIMAN:  OSTARIOPHYSI 


129 


Ssur 


Fig.  62.  Representative  cypriniform  larvae.  (A-C)  Cyprinidae:  (A)  Tribolodon  hakonensis  (UMMZ  212151)  9.2  mm  TL;  (B)  Semotilus 
alromaculatus  8.6  mm  TL;  (C)  Barbiis  {  =  Capoela)  tilteya  (UMMZ  212148)  6.0  mm  TL;  (D,  E)  Cobitidae;  (D)  Misgurnus  fossiHs  6.9  mm  TL, 
after  Kryzhanovskii  (1949);  (E)  Acanthophthalmus  cf  kuhni  4.0  mm  TL  (specimen  from  S.  S.  Boggs). 


lacking  oil  globules.  Eggs  may  be  strongly  adhesive  (e.g.,  Cy- 
priniformes:  Nemacheilus  [=Barbatula]  torn  [Kobayasi  and 
Moriyana,  1957];  Characiformes:  Gymnocon'mhus  tenictzi  [pers. 
obs.];  Siluriformes:  Loricana  calaphracta  [pers.  obs.]),  nonad- 
hesive  (e.g.,  Cypriniformes:  Clenopharyngodon  idclla  [Inaba  et 
al.,  1957];  Siluriformes:  Tandanm  landanus  [Lake.  1967]),  or 
weakly  adhesive  (e.g.,  Cypriniformes:  Catoslomus  commersoni 
[pers.  obs.];  Characiformes:  Scrrasalmm  nattercn  [pers.  obs.]; 
Siluriformes:  Baganus  hagarius  [David,  1961]).  Adhesive  fila- 
ments or  other  apparent  modifications  of  the  egg  surface  are 
almost  entirely  unknown. 

Representatives  of  outgroups  (Gonoi^nchiformes,  Clupeo- 
morpha,  "Salmoniformes,"  and  Osteoglossomorpha)  share  the 
spherical  egg  with  yellow,  granular  or  segmented  yolk.  Their 
eggs  are  pelagic  or  demersal,  usually  1 .0  to  1.3  mm  in  diameter. 


adhesive  (in  Osmerus)  or  nonadhesive  (in  Chanos.  Alosa.  and 
Hiodon).  without  oil  globules  (Chanos)  or  with  one  to  several 
(in  Alosa  and  Osmerus). 

Exceptions  to  this  characterization  of  ostariophysan  eggs  exist. 
Among  cypriniforms,  the  cyprinid  subfamily  Acheilognathinae 
(Gosline,  1978)  exhibits  elliptical  to  pyriform  eggs  which  are 
deposited  in  the  mantle  cavity  of  a  bivalve  mollusc  (Kryzhan- 
ovskii et  al.,  1951;  Nakamura,  1969;  Makeeva,  1976).  Their 
irregular  shape  may  be  the  important  mechanism  preventing 
the  eggs  from  being  expelled.  Some  cyprinid  eggs  are  pelagic 
(e.g.,  Hypophthalmichthys  molitrix  [Nakamura,  1969;  Koblit- 
skaia,  1981])  and  have  a  larger  diameter  (ca.  5  to  6  mm)  due 
to  the  considerable  perivitelline  space.  Only  one  ostariophysan, 
the  cypriniform  Cobitis  biwae,  was  reported  to  have  12  to  13 
small  oil  globules  in  the  yolk  (Okada  and  Seiishi,  1938;  Okada, 


Fig.  63.     Representative  cypriniform  larvae  (continued).  Catostomidae:  Hypentetium  etowanum  (upper)  13.1  mm  and  (lower)  15.0  mm  TL. 


130 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


/JS-^y 


Fig.  64.     Representative  characiform  larvae.  Serrasalmidae:  Serrasalmus  nattereri  (UMMZ  211677)  8.2  mm  TL  (upper).  Characidae:  Hy- 
phessobrycon  cf.  callistus  (UMMZ  21 1676)  6.6  mm  TL  (lower). 


1960),  but  this  is  in  doubt  (N.  Komada,  pers.  comm.)  and  has 
not  since  been  confirmed. 

Characiform  eggs  are  poorly  known;  most  information  is  from 
the  aquarium  hobby  literature.  Known  characid  (sensu  Gery, 
1977)  eggs  are  small  (0.8  to  1.2  mm).  However  other  families 
have  eggs  between  2  and  4  mm  (e.g.,  Alestidae,  Anostomidae, 
Curimatidae,  Hepsetidae,  Serrasalmidae).  Apparently  most 
species  have  eggs  that  adhere  to  plants.  Franke  (1981)  described 
adhesive  threads  (gallertigen  Klebfdden)  on  the  surface  of  the 
egg  of  Ctenolucius  hujeta  (Ctenoluciidae)  and  noted  that  this 
was  the  mechanism  by  which  they  attached  to  plants.  My  ex- 
amination of  eggs  supplied  by  Dr.  Franke  found  the  adhesive 
structure  to  be  a  membranous  pedestal  rather  than  adhesive 
threads  (Fig.  6 1 ).  This  is  the  only  known  chorionic  modification 
of  ostariophysans. 

Most  siluroids  have  demersal,  medium  sized  eggs  ( 1  to  4  mm). 
Some  are  tended  by  one  or  both  parents  [e.g.,  Clarias  batrachus 
(Mookerjee,  1946;  Mookerjee  and  Mazumdar,  1950),  Ictalurus 
punctatus  (Tin,  1982c)];  others  are  not  given  parental  care  [e.g., 
Clarias  gariepinus  (HoW,  1968;  Bruton,  1979),  Pangasius  sutchi 
(Varikul  and  Boonsom,  1969)].  The  eggs  are  typically  spherical; 
however,  Clarias  eggs  are  often  slightly  elliptical  (Mookerjee, 
1946;  Greenwood,  1955;  Bruton,  1979).  Some  callichthyids  de- 
posit small  eggs  (ca.  1.0  mm)  in  a  foam  nest  on  the  surface  of 
still  waters  (Kryzhanovskii,  1949).  Parents  in  several  families 
carry  their  eggs.  Some  loricariids  (e.g.,  Loricaria  spp.)  carry  a 
mass  of  eggs  by  means  of  fleshy  appendages  of  the  lower  lip. 
Aspredo  laevis  eggs  apparently  are  attached  by  vascularized  stalks 
to  the  venter  of  the  female  (Wyman,  1859).  Finally,  ariids  are 
oral  incubators  with  perhaps  the  largest  eggs  of  all  oviparous 


teleosts  (10  to  25  mm)  (Chidambaram,  1942;  Gudger,  1912, 
1916,  1918;  and  other  authors).  Although  yolk  is  usually  yellow 
to  slightly  orange  or  brown,  several  species  have  unmistakably 
green  yolk  [e.g.,  Bagarius  bagarius  (David,  1961),  Clarias  ba- 
trachus (Mookerjee,  1946;  Mookerjee  and  Mazumdar,  1950), 
Heteropneustes  fossilis  (Pal  and  Khan,  1969),  Loricariichthys 
sp.  (Taylor,  1983),  Phractura  ansorgei  (Foersch,  1966)].  At  least 
one  siluroid,  the  silurid  Ompok  bimaculatus,  has  reddish  brown 
yolk  (Chaudhuri,  1962).  A  few  species  have  a  jelly-like  coat 
surrounding  the  chorion  [e.g.,  Bagarius  bagarius  (David,  1961), 
Parasilurus  asotus  (Kryzhanovskii  et  al.,  1951),  Phractura  an- 
sorgei (Foersch,  1966),  Trachycorystes  insignis  (Burgess,  1982)]. 

Larvae 

Most  ostariophysans  hatch  in  an  altricial  state  at  about  the 
time  when  pectoral  buds  form,  but  before  the  head  becomes 
free  from  the  yolk  sac  and  retinal  pigment  develops,  although 
there  is  variability  in  the  exact  stage.  The  yolk  sac  is  usually 
large  and  cumbersome,  enforcing  a  stationary  existence  during 
the  first  days,  either  on  the  substrate  (most  commonly)  or  at- 
tached to  plants  by  means  of  a  cephalic  adhesive  mechanism 
(found  in  most  characiforms  and  a  few  cyprinids,  but  structur- 
ally diflTerent  in  these  groups).  Caudal  fin  rays  diflierentiate  first, 
followed  by  nearly  simultaneous  formation  of  dorsal  and  anal 
fin  rays.  Pectoral  and  pelvic  fin  rays  develop  near  the  end  of  the 
larval  period.  The  gonorynchiform  Chanos  hatches  at  about  the 
same  stage  of  development  as  ostariophysans,  but  Atosa  and 
Osmerus  hatch  somewhat  later  (i.e.,  pectoral  buds  and  retinal 
pigment  are  clearly  developed).  These  outgroups  generally  have 
pelagic  larvae  at  hatching.  Fin  rays  in  Chanos  develop  in  the 


FUIMAN:  OSTARIOPHYSI 


131 


same  order  as  described  above,  but  the  sequence  differs  for  Alosa 
and  again  for  Osmerus. 

Within  Ostariophysi,  cypriniform  larvae  (Figs.  62,  63)  are 
largest  at  hatching  (Table  26),  the  largest  sizes  represented  most- 
ly by  catostomids.  The  pyriform  yolk  sac  extends  from  below 
the  head  posteriorly  to  the  anus  (Fig.  62a).  Barbels,  when  pres- 
ent, develop  very  late  in  Cyprinidae  but  early  in  Cobitoidea 
(sensit  Sawada,  1982).  Cyprinids  display  considerable  variation 
in  the  elaboration  of  the  larval  circulatory  system.  Temporary 
networks  of  blood  vessels  invade  portions  of  the  finfolds  and 
the  surface  of  the  yolk  sac  in  a  variety  of  patterns  to  form  the 
larval  respiratory  system  (Kryzhanovskii,  1947).  Cobitoideans 
usually  have  greatly  expanded  finfolds,  especially  those  of  the 
pectoral  buds.  Pronounced  external  gill  filaments  are  known  in 
the  cobitine  genera  Coto/5  (Kryzhanovskii,  1949;Okada,  1960; 
Sterba,  1962),  Lepidocephaliis  (Bhimachar  and  David.  1945), 
and  A/;5^r«wi  (Kryzhanovskii,  1949;  Okada,  1960),  but  not  in 
the  non-cobitine  cobitoidean  genera  Botta.  Lefua,  or  Nemach- 
eilus,  nor  in  other  ostariophysans.  Cyprinids  with  cephalic  ad- 
hesive glands  include:  Ahramis  brama  (Penaz  and  Gajdusek, 
1979);  Brachydanio  rerio  (Frank.  1978);  Cypri niis  carpio  (Hoda 
and  Tsukahara,  1971);  Danio  malabancus  (Jones,  1938);  and 
Notemigonuscrysoleucas (Snyder tXa\.,  \911\  Loosetal.,  1979). 

In  characiforms,  the  yolk  sac  is  short  and  rounded,  not  ex- 
tending to  the  anus  posteriorly  (Fig.  64).  Most  known  characids 
(sensii  stricto)  and  a  hepsetid  (Bertmar,  1959;  Budgett,  1902. 
1 903),  erythrinid  (de  Azevedo  and  Gomes,  1 942),  and  curimatid 
(de  Azevedo  et  al.,  1938)  have  a  temporary  larval  cephalic  ad- 
hesive organ  (more  distinct  than  the  apparent  glandular  mech- 
anism in  cyprinids).  Those  without  such  an  organ  mclude:  Ser- 
rasalmus  nattereri  (pers.  obs.),  Metynnis  maciilatiis  (Azuma, 
1982),  and  Brycinus  longipinnis  (Frank,  1972).  The  adipose  fin 
appears  to  develop  de  novo  toward  the  end  of  the  larval  period, 
not  as  a  remnant  of  the  median  finfold.  However,  the  small  size 
of  the  adipose  fin  and  lack  of  specimens,  photographs,  illustra- 
tions, and  descnptions  of  late  larval  characiforms  prevents  ver- 
ification of  this  inference. 

Although  few  species  are  known  as  larvae,  Siluroidei  may 
contain  the  greatest  diversity  of  larval  characters  among  Ostar- 
iophysi (Fig.  65).  Most  siluroids  hatch  as  altricial  larvae  with  a 
physiognomy  similar  to  that  of  characiforms.  Ictalurids  are  more 
precocial  and  lack  a  postlarval  (sensu  Hubbs,  1 943)  phase.  Ariids 
(Gudger,  1918;  Ward,  1957)  and  some  loricariids  (Lopez  and 
Machado,  1975;  Machado  and  Lopez,  1975)  hatch  in  a  highly 
precocial  state,  resembling  the  adult  in  many  aspects  of  external 
morphology  but  retaining  a  large  yolk  sac  (Fig.  65C).  In  most 
families,  barbels  are  usually  present  at  hatching  or  soon  there- 
after (Fig.  65a).  Cephalic  adhesive  organs  are  usually  absent, 
but  at  least  one  loricariid  (Ancistrus  sp.)  possesses  these  (Franke. 
1979).  Clarias  gariepinus  (=C.  mossambicus)  and  Ompok  bi- 
maculatus  have  an  adhesive  organ  on  the  venter  of  the  yolk  sac 
(Greenwood,  1955,  1956;  Chaudhuri,  1962;  Holl,  1968;Bruton, 
1979).  The  adipose  fin  is  clearly  a  remnant  of  the  median  fin- 
fold,  as  in  "'salmoniforms."  Larvae  of  a  single  gymnotoid,  Ei- 
genmannia  virescens.  are  known  (Fig.  65D,  E;  Table  26;  Kirsch- 
baum  and  Balon,  in  prep.). 

Relationships 

The  Ostariophysi  are  thought  to  be  the  sister  group  of  the 
Gonorynchiformes  (Greenwood  et  al.,  1966;  Rosen  and  Green- 
wood,  1970;  Gosline,   1971;  Fink  and  Fink,   1981).  The  next 


closest  relatives  are  Clupeiformes  (Gosline,  1971)  or  "Salmon- 
iformes"  (Greenwood  et  al.,  1966;  Fink  and  Weitzman,  1982). 
All  concepts  of  Ostariophysi  (those  with  a  Weberian  appa- 
ratus) recognize  four  major  groupings,  "cyprinoids,"  "chara- 
coids,"  "gymnotoids,"  and  "siluroids."  The  traditional  view  of 
relationships  holds  that  "characoids"  are  the  ancestral  stock, 
giving  rise  to  the  remaining  lineages,  with  "gymnotoids"  being 
modified  "characoids,"  and  "cyprinoids"  being  the  closest  rel- 
atives of  the  "characoids"  plus  "gymnotoids."  Fink  and  Fink 
(1981)  gave  a  detailed  history  of  the  classification  schemes  for 
the  Ostariophysi  and  their  relatives  as  an  introduction  to  their 
work  on  the  subject,  which  is  the  only  attempt  to  reconstruct 
the  phylogeny  on  the  basis  of  a  large  set  of  data  ( 1 27  characters). 
Their  proposed  cladistic  phylogeny  differs  significantly  from  the 
traditional  one  by  aligning  "gymnotoids"  with  "siluroids"  as 
the  Siluriformes  (Fig.  66). 

Developmental  characters  in  systematics 

Few  attempts  have  been  made  to  apply  developmental  char- 
acters to  the  systematics  of  ostariophysans.  Kryzhanovskii  (1947) 
grouped  cyprinids  into  four  subfamilies  according  to  details  of 
the  larval  respiratory  system.  He  also  included  characters  re- 
lating to  reproductive  guild  (later  elaborated  in  Kryzhanovskii, 
1948),  original  (ontogenetically)  position  of  the  mouth,  and  rel- 
ative size  of  the  pectoral  buds.  He  supported  these  subfamilial 
designations  with  experimental  results  on  the  morphology  and 
viability  of  larvae  produced  by  artificial  hybridizations  within 
and  among  the  proposed  subfamilies. 

Nakamura  (1969)  dealt  with  the  cyprinids  of  Japan.  In  his 
English  summary,  he  stated  that  currently  proposed  closely  re- 
lated forms  (meaning  genera,  species,  and  subspecies)  have  sim- 
ilar life  history  characteristics.  He  noted  a  few  exceptions,  such 
as  similar  (as  adults)  species  oi  Moroco  whose  early  larvae  differ 
morphologically  and  ecologically.  In  contrast,  he  noted  that  the 
eggs  and  early  larvae  of  Ctenopharyngodon  idella  and  Hypoph- 
thalmichthys  molitri.x  were  very  similar  although  the  species 
were  placed  in  different  subfamilies.  He  used  differences  in  egg 
and  larval  morphology  to  support  the  previously  uncertain  sep- 
aration of  the  genera  Squalidus  and  Gnathopogon. 

In  a  similar  survey.  Loos  and  Fuiman  (1978)  attempted  to 
characterize  the  subgenera  of  the  New  World  cyprinid  genus 
Notropis  in  terms  of  their  egg  and  larval  morphology.  However, 
they  found  substantial  variability  within  the  established  sub- 
genera and  were  unable  to  characterize  them  precisely. 

Each  of  these  attempts  to  apply  developmental  characters  to 
systematics  was  concerned  only  with  establishing  group  mem- 
bership and  not  with  determining  relationships  among  the  groups. 
Further,  none  of  the  work  was  based  on  a  large  data  set  nor  was 
it  approached  in  a  rigorous  manner.  The  difficulties  encountered 
by  Nakamura  ( 1 969),  and  especially  by  Loos  and  Fuiman  (1978), 
probably  were  due  to  the  apparently  convergent  ecomorpho- 
types  expressed  by  unrelated  taxa.  The  low  taxonomic  level 
investigated,  combined  with  the  morphological  similarity  im- 
plied by  von  Baer's  law,  probably  accounted  for  much  of  the 
remaining  difficulty  in  detecting  consistent  differences  among 
taxa. 

Fink  and  Fink's  (1981)  classification  is  based  largely  on  os- 
teological  characters.  The  great  size  and  diversity  of  Ostario- 
physi make  a  detailed  study  of  developmental  osteology  and 
concomitant  investigations  of  bone  homologies  impractical  at 
this  time.  Yet,  available  information  permits  a  preliminary  eval- 


132 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


D 


FUIMAN:  OSTARIOPHYSI 


133 


Fig.  65.  Representative  siluriform  larvae.  (A-B)  Clariidae:  Clanas  gariepinus  (British  Museum  of  Natural  History,  uncataloged)  (A)  6.6  mm 
and  (B)  8.4  mm  TL;  (C)  Loricariidae:  Ancistrus  spinosus  (UMMZ  212152)  8.3  mm  TL;  (D-E)  Rhamphichthyidae;  Eigenmannia  virescens  (D) 
5.0  and  (E)  8.1  mm  TL. 


uation  of  relationships  based  on  developmental  characters.  The 
following  analysis  attempts  to  evaluate  the  contribution  of  se- 
lected developmental  characters  to  ostariophysan  systematics 
by  constructing  an  independent  assessment  of  phylogeny  based 
on  developmental  characters.  That  the  assessment  should  be 
independent  was  attested  by  Moser  and  Ahlstrom  (1974):  "we 
are  increasingly  impressed  with  the  functional  independence  of 
larval  and  adult  characters.  It  is  apparent  that  the  world  of  the 
larvae  and  the  world  of  the  adults  are  two  quite  separate  evo- 
lutionary theaters." 

Representative  ontogenetic  series  of  all  families  of  ostario- 
physans  are  nearly  impossible  to  obtain  because  of  the  large  size 
and  wide  geographic  distribution  of  the  group  and  the  dearth 
of  ichthyologists  studying  larvae.  Consequently,  the  analyses 
employed  here  were  based  on  specimens  generated  from  labo- 
ratory breeding  experiments,  wild-caught  material,  and  data 
published  in  apparently  accurate  accounts  of  ontogeny.  Species 
used  in  the  analyses  included  four  outgroups  to  the  Ostariophysi 
(Gonorynchiformes,  Clupeomorpha,  "Salmoniformes,"  and 
Osteoglossomorpha),  all  characiforms  and  siluriforms  with  suf- 
ficient morphometric  and  developmental  data  for  analysis,  and 
a  sample  of  five  species  from  the  most  primitive  cypriniform 
family,  Cyprinidae.  These  cyprinid  species  possess  different 
combinations  of  larval  characters  (determined  by  their  location 
on  a  Wagner  tree  generated  for  33  larval  cyprinids  [Fuiman, 
1983a]).  Although  not  used  directly,  incomplete  data  on  ap- 
proximately 85  additional  non-cyprinid  ostariophysans  provid- 
ed corroborative  information. 

Species  included  in  the  analysis  of  relationships  and  their 


sources  are  listed  below.  Initials  denote  specimens  borrowed 
from,  or  information  provided  by:  Florida  State  Board  of  Con- 
servation (FSBC),  University  of  Michigan  Museum  of  Zoology 
(UMMZ),  or  Frank  Kirschbaum  (FK). 

OsTEOGLOSSiFORMEs:  Hiodofi  tergisus  [Snyder  and  Douglas 
(1978);  Wallus  (1981,  pers.  comm.)]. 

Salmoniformes:  Osmerus  mordax  [Cooper  ( 1 978);  Tin  ( 1 982b)]. 

Clupeiformes:  Alosa  pseudoharengus  [Jones  et  al.  (1978);  Tin 
(1982a)]. 

Gonorynchiformes:  Chanos  chanos  [Chaudhuri  et  al.  (1978); 
Liaoet  al.  (1979);  Miller  et  al.  (1979)]. 

Cypriniformes:  Cyprinidae— Cvpn>!Wicarp/o  [UMMZ  21 1678; 
Hoda  and  Tsukahara  (1971);  Nakamura  (1969);  Okada 
(I960)];  Leiiciscus  cephaliis  [Cemy  (1977);  Kryzhanovskii 
(1949);  Penaz  (1968);  Prokes  and  Penaz  (1980)];  Opsan- 
ichthys  unciroslris  [Kryzhanovskii  et  al.  (1951);  Makeeva 
and  Ryabov  (1973);  Nakamura  (1951,  1969)];  Parabramts 
pekmensis  [Institute  of  Hydrobiology  (1976);  Kryzhanov- 
skii et  al.  (1951)];  Squalidus  gracilis  [Nakamura  (1969)]. 

Characiformes:  Alestidae— .-l/eirw  haremose  [Durand  and 
Loubens  (1971 )].  Erythrinidae— //op/Zaj^  malabaricus  [FSBC 
8962,  8963,  9593;  de  Azevedo  and  Gomes  (1942);  Hensley 
(1976);  Moreira  ( 1 920);  von  Ihering  et  al.  ( 1 928)].  Charac- 
idae— Hyphessobrycon  cf.  callistiis  [UMMZ  21 1676],  Ser- 
r&sdAmiddie— Serrasalmus  nattereri  [UMMZ  21 1677;  Azu- 
ma(1975)]. 

Siluriformes:  Siluroidei:  Ba.gn6.aQ  —  Mystus  seenghala  [Saigal 
and  Motwani  (1962)];  Rita  rila  [Karamchandani  and  Mot- 


Cypriniformes 


Characiformes 


Siluroidei 


Gymnotoidei 


Fig.  66.     Cladogram  of  ostariophysan  relationships  derived  from  adult  characters  by  Fink  and  Fink  (1981).  Stem  lengths  imply  no  special 
significance. 


134 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Squalidus 
CYPRINIFORMES 


Ictalurus 
21 


Eigenmannia 
*28 


SILURIFORMES 


CHARACIFORMES 


Fig.  67.     Wagner  tree  of  ostariophysan  phylogeny  based  on  larval  characters.  Stem  lengths  are  proportional  to  the  number  of  character-state 
changes  on  a  given  stem. 


wani  (1955)].  Clariidae— Ctor/a^  batrachus  [UMMZ  1 86690. 
209039;  Devaraj  et  al.  (1972);  Mookerjee  (1946);  Mook- 
erjee  and  Mazumdar  (1950)].  Ictaluridae— /rta/Mr«5  neb- 
ulosus  [Armstrong ( 1 962);  Tin  ( 1 982c)].  Pangasiidae-Paw- 
gasius  sutchi  [Varikul  and  Boonsom  (1969)].  Sisoiidae— 
Bagarius  bagarius  [David  (1961)].  Gymnotoidei;  Rham- 
ph'ichVnyiddie— Eigenmannia  virescens  [FK,  Kirschbaum 
and  Westby  (1975)]. 


Phylogenetic  methods 
The  phylogenetic  reconstruction  based  on  developmental 
characters  was  generated  by  the  cladistic  Wagner  tree  method 
(described  by  Kluge  and  Farris.  1969;  Farris.  1970;  Lundberg, 
1972;  and  Jensen,  1981).  Characters  were  chosen  by  virtue  of 
their  availability  in  published  accounts.  Nearly  all  were  recorded 
as  continuous  measures,  but  individual  modes  with  their  neigh- 
boring values  and  disjunct  portions  of  distributions  separated 


Table  27.    Ranges  of  Values  for  Coded  Character  States  of  16  Ostariophysans.  Character  numbers  correspond  to  those  given  in  the 

text.  Primitive  states  are  given  in  boldface  type. 


\*V\z\rz\(^\PT 

Character  slate 

number 

a 

b 

c 

d 

e 

f 

1 

2.27-3.42 

3.74-3.74 

4.74-4.90 

2 

3.93-5.06 

5.65-6.28 

7.06-7.40 

8.86-8.86 

3 

1.35-1.53 

1.92-2.58 

2.98-4.03 

4 

1.19-1.19 

2.09-2.09 

2.46-3.39 

3.78-5.14 

5 

0.97-1.61 

2.18-2.18 

6 

2.06-2.46 

2.62-2.86 

3.25-3.31 

7 

0.13-0.28 

0.36-0.52 

8 

0.14-0.28 

0.42-0.58 

0.71-0.82 

9 

0.72-1.00 

1.29-1.29 

10 

1.26-1.68 

1.97-2.11 

11 

-1.22—0.94 

-0.67—0.23 

0.06-0.06 

12 

0.70-1.04 

1.15-1.40 

13 

0.22-0.39 

0.54-0.84 

14 

15.3-19.0 

21.6-22.5 

25.0-26.3 

28.5-30.3 

32.7-32.7 

15 

8.0-8.0 

12.0-20.0 

25.5-25.6 

29.0-29.0 

38.7-38.7                 45.9-45.9 

16 

0.28-0.29 

0.39-0.44 

0.52-0.70 

0.81-0.81 

17 

0.35-0.35 

0.52-0.55 

0.70-O.96 

1.06-1.10 

18 

0.22-0.38 

0.42-0.56 

19 

0.0-0.05 

0.12-0.28 

0.44-0.44 

20 

0.0-0.03 

0.07-0.15 

0.22-0.33 

21 

0-0 

1-1 

22 

0-0 

1-1 

23 

0-0 

1-1 

24 

0-0 

1-1 

FUIMAN:  OSTARIOPHYSI 


135 


Table  28.  Character-State  Changes  on  Stems  Leading  to 
Hypothetical  Ancestors  (Nodes)  and  Terminal  Taxa  on  the 
Wagner  Tree  of  Ostariophysi.  Numbered  character  states  correspond 
to  those  given  in  Table  27 ,  Uniquely  derived,  unreversed  character  states 

are  given  in  boldface  type.  Reversed  characters  are  noted  by  (r).  Node 
numbers  correspond  to  those  given  in  Fig.  67. 

Node  Characler  state 

1  8c,  lib,  12b,  14b,  20b,  20c  24b 

2  6b 

3  14d,  15b,  18b 

4  13a,  21a 

5  18a(r),  20b(r) 

6  16b 

7  17b 

8  14c,  20a(r),  23b 

9  6b,  19a.  24a(r) 

10  3b,  lla(r),  I2a(r).  14b 

11  lla(r),  22b 

12  6b,  6c 

13  3b,  4c,  12b(r),  14a,  14b,  14c 

14  3a,  4b,  18a(r),  20b(r) 

15  10b,  lib,  24a(r) 

16  12b.  17b.  14b(r) 

17  2b,  6b,  7a.  8b(r) 

18  8a,  16b.  20b{r) 

19  12b,  16a,  18a(r),  20a(r) 

20  2a(r),  6a(r).  lOa(r) 

2 1  5b,  6c,  7b(r),  8b(r),  9b,  1 1  c,  1 5c,  17a,  1 7b.  1 9a 

22  lOa(r),  15c,  15d 

23  14b(r) 

24  lb,  2c,  I7d 

25  6a(r),  1  la(r),  15c(r),  16d,  19c 

26  Ic,  15e,  19a,  20b(r).  22a 

27  2d,  8a,  14b(r),  14c(r),  17b.  17c(r).  18a(r) 

28  4a,  4b,  6c,  10b,  15f 


by  measurable  gaps  were  coded  individually.  Characters  were 
polarized  by  outgroup  comparison  (Table  27).  The  evolutional^ 
transformation  series  for  each  continuous,  multiple  state  char- 
acter was  assumed  to  be  linear  (i.e..  with  one  or  two  adjacent 
states  for  a  given  state).  Consequently,  a  character  coded  with 
n  states  had  n  -  1  different  changes  from  one  state  to  another, 
disregarding  the  direction  of  change.  These  transitions  were 
termed  "two-state  factors."  All  two-state  factors  and  their  states 
for  each  species  were  generated  by  the  FACTOR  computing 
program  (Estabrook  et  al.,  1976).  The  output  from  this  program 
included  an  input  file  for  the  WAGNER  78  computing  program 
which  was  used  to  construct  Wagner  trees.  The  data  deck  was 
resequenced  and  a  new  Wagner  tree  generated  several  times  in 
order  to  identify  the  shortest  (most  parsimonious)  tree  (Jensen, 
1981). 

Characters 
Morphomelhc  characters.— To  develop  morphometric  charac- 
ters for  phylogenetic  analysis,  the  following  lengths  were  mea- 
sured along  the  longitudinal  axis  of  the  fish;  total  length,  preanal 
length,  head  length,  and  eye  diameter.  Two  vertical  measure- 
ments, head  depth  and  body  depth  at  anus,  were  meant  to  rep- 
resent size  and  shape  in  the  dorso-ventral  direction.  All  mea- 
surements were  defined  by  Fuiman  (1979).  They  were  made 
reasonably  independent  of  one  another  by  subtracting  preanal 
length  from  total  length  to  yield  peduncle  length,  and  head  length 
from  preanal  length  to  yield  tnank  length.  Peduncle  length,  trunk 
length,  head  length,  eye  diameter,  body  depth,  and  head  depth 
comprised  the  basic  morphometric  characters. 


(0 

o 

0 

a 
CO 


0) 

E 

3 


3- 


2- 


1- 


\ZZ}  Cypriniformes 

I      I  Characiformes 

^H  Siluroidei 

rXI  Gymnotoidei 


0.25 


0.45 


0.65 


0.85 


Yolk-Sac  Shape  (depth/length) 

Fig.  68.     Frequency  distribution  of  yolk-sac  shape  for  recently  hatched 
ostariophysan  species. 


Body  dimensions  of  larvae  are  strongly  influenced  by  allom- 
etry  (Fuiman.  1983b).  Such  measures  cannot  be  expressed  as 
simple  proportions,  because  the  proportions  are  not  constant 
within  a  species  throughout  the  larval  period.  The  effect  of  size 
on  shape  must  be  eliminated  in  comparisons  of  shape.  Further, 
any  single  measure  which  accounts  for  size  in  one  taxon  may 
be  an  inappropriate  measure  of  size  in  a  distantly  related  taxon. 
Within-group  principal  component  analysis  can  be  used  to  ex- 
tract a  size  component,  PCI  (Humphries  et  al.,  1981),  that  is  a 
linear  combination  of  several  variables,  each  containing  infor- 
mation on  size  and  shape.  Thus,  PC  1  includes  more  information 
on  size  than  any  single  measure  and  is  a  better  comparison  across 
taxa. 

Univariate  and  multivariate  methods  of  allometry  relate  dis- 
tance measures  log-linearly  (Huxley,  1932;  Jolicoer,  1963).  Thus, 
a  within-species  principal  component  analysis  of  the  logarithms 
of  the  six  basic  morphometric  characters,  based  on  the  covari- 
ance  matrix,  was  performed  to  extract  the  size  component  (PCI ). 
The  extreme  PCI  scores  for  all  taxa  were  compared  and  two 
values  (0.00  and  0.60),  one  near  each  end  of  the  larval  period, 
were  chosen  as  standard  sizes  for  comparing  morphometry.  The 
six  morphometric  measures  were  reconstructed  for  each  of  these 
sizes  by  means  of  the  regressions  of  the  logarithm  of  the  char- 
acter on  PC  1 .  By  selecting  two  sizes  to  compare,  the  phylogenetic 
analysis  included  information  on  changing  shape  (allometry)  as 
well  as  static  shape.  The  final  1 2  character  values  were  recorded 
as  predicted  lengths  (in  mm)  for  each  morphometric  measure 
at  each  of  2  standard  sizes.  However,  body  depth  at  the  anus 
contained  no  discontinuous,  phyletic  variability.  The  final  mor- 
phometric characters  were;  (Characters  1  and  2)  Peduncle  length 
(smaller  and  larger  standard  size,  respectively),  (3  and  4)  Trunk 
length,  (5  and  6)  Head  length,  (7  and  8)  Eye  length,  (9  and  10) 
Head  depth.  Three  additional  morphometric  characters  were 


136 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


0.9 


E 
E 

c 

O) 

c 

(V 

-J 

0) 

o 

c 

3 
T3 

(U 
Q. 


CI3  Cypriniformes 
I     1  Characiformes 
fTTTTI  Siluroldel 
— -  Gymnotoidel 


— I 1 1 1 1 1 1 — 

0.0  0.2  0.4  0.6 

Size  (PCI    score) 


Fig.  69.     Morphometric  characters  important  in  defining  major  groups  of  ostariophysan  larvae.  Shaded  areas  and  individual  Imes  enclose  all 
regression-predicted  values  at  two  standard  sizes  (0.0  and  0.6)  of  a  given  taxon. 


included:  (11)  Size  at  hatching  (PCI  score  at  total  length  for 
hatching,  based  on  the  regression  of  PCI  on  the  logarithm  of 
total  length),  (12)  Size  at  complete  finfold  absorption  (PCI  score 
at  total  length  for  complete  finfold  absorption,  based  on  the 
regression  of  PCI  on  the  logarithm  of  total  length),  (13)  Yolk- 
sac  shape  (ratio  of  the  greatest  vertical  length  [depth]  of  the  yolk 
sac  to  its  greatest  horizontal  length  in  recently  hatched  individ- 
uals). 

Meristic characters.— These  include:  ( 1 4)  Preanal  myomeres  (all 
myomeres  at  least  partly  anterior  to  a  vertical  line  projected 
from  the  anus,  including  an  occipital  segment)  and  (15)  Postanal 
myomeres  (all  myomeres  entirely  posterior  to  a  vertical  line 
projected  from  the  anus,  including  a  urostylar  segment). 

Missing  myomere  data  for  Hoplias  malabancus  were  taken 
from  vertebral  counts  made  from  radiographs  of  adults  (UMMZ 
66435).  The  one-to-one  ontogenetic  relationship  of  myosepta 
to  neural  spines  in  monospondylous  fishes  (Lauder,  1980)  per- 
mitted estimation  of  myomere  number  from  vertebral  number 
only  by  inclusion  of  myomeres  for  an  occipital  segment,  a  uro- 
stylar segment,  and  the  four  (five  in  siluroids)  obscured  We- 
berian  vertebrae  (Fuiman,  1982a). 

Ontogenetic  characters. —Size,  rather  than  chronological  age,  is 
most  closely  related  to  development  (Gerking  and  Rausch,  1 979). 
Thus,  total  length  at  the  onset  of  selected  developmental  events 
was  recorded.  To  compare  these  sizes  among  species  with  dif- 
fering initial  lengths  and  ranges  of  lengths  for  the  larval  period. 


the  logarithm  of  the  hatching  length  was  subtracted  from  the 
logarithm  of  the  length  at  a  given  event.  This  difference  was 
divided  by  the  difference  of  the  logarithms  of  length  at  complete 
finfold  absorption  and  at  hatching  (the  criteria  used  here  to 
delimit  the  larval  period).  The  resultant  character  was  the  per- 
centage of  the  larval  period  that  occurred  prior  to  the  event,  an 
estimate  of  relative  timing  of  the  event.  When  characters  were 
present  at  hatching  or  did  not  develop  until  after  complete  fin- 
fold absorption  they  were  coded  as  0.00  or  1.10,  respectively. 
The  following  events  were  recorded:  (16)  Anal  fin  rays  (first 
distinct  ray),  ( 1 7)  All  median  fin  rays  (all  median  fin  rays  present, 
finfolds  may  persist,  fin  margins  may  be  incomplete),  (18)  Yolk 
absorption  (complete  absorption  of  yolk),  (19)  Head  straight 
(head  free  from  yolk  sac  and  not  deflected  downward),  (20)  Eye 
pigment  (first  uniform  pigmentation  of  retina). 

Presence/absence  characters— Presence  (coded  as  I )  or  absence 
(0)  of  the  following  structures  at  any  time  during  the  larval 
period  was  recorded:  (21)  Jaw  teeth  (teeth  on  the  premaxilla, 
maxilla,  or  dentary),  (22)  Adipose  fin,  (23)  Caudal  spot  (con- 
gregation of  melanophores  at  the  base  of  the  caudal  fin  forming 
a  distinct  spot),  (24)  Lateral  stripe  (melanophores  on  the  mid- 
lateral  myoseptum  forming  a  continuous,  longitudinal  stripe). 

Phylogenetic  results 
The  Wagner  tree  (Fig.  67,  Table  28)  contains  101  steps  for 
the  46  two-state  factors  ("characters").  Members  of  each  major 


Table  29.     Distribution  Statistics  of  Preanal.  Postanal,  and  Total  Myomere  Number  for  Ostariophysan  Larvae.  Values  are  based  on 

means  for  each  species. 


Preanal  myomeres 


Postanal  myomeres 


'  Including  Cyprinidae,  Calostomidae,  and  Cobiloidea. 


Total  myomeres 


Taxon 

species 

Mean 

Extremes 

Mean 

Extremes 

Mean 

Extremes 

Cypriniformes' 

52 

29.4 

18.5-38.8 

12.4 

7.0-18.8 

41.7 

32.8-50.7 

Characiformes 

4 

25.5 

17.8-32.7 

15.2 

8.0-20.0 

42.0 

36.1-50.0 

Siluroidei 

6 

19.7 

15.3-26.3 

25.9 

16.5-38.7 

45.4 

33.0-65.0 

Gymnotoidei 

1 

17.3 

45.9 

63.2 

FUIMAN:  OSTARIOPHYSI 


137 


4-1 


3- 
co 
.2 
'5 

0 

a 

CO 

"?;      2-\ 


E 

3 


r//J  Cyprinidae 
I  I  Cobitoidea 
^B   Siluroidei 


^    ^. 


i 


1 


r 


0.0 


0.4 


0.8 


1.2 


Barbel  Formation 
(onset  as  percentage  of  larval  period) 

Fig.  70.  Frequency  dislnbution  of  relative  timing  of"  barbel  forma- 
tion in  ostanophysan  species.  Cyprinids  are  represented  by  1 0  barbelled 
species,  not  all  of  which  are  discussed  in  the  text. 


taxon  (Cypriniformes.  Characiformes,  Siluroidei)  are  placed  near 
one  another,  but  larval  characters  are  insufficient  to  demonstrate 
the  monophyly  of  characiforms  or  siluroids.  The  largest  number 
of  primitive  characters  is  found  in  Hoplias  (Characiformes),  but 
the  cypriniform  lineage  differs  from  Hoplias  by  only  three  char- 
acter state  changes  (node  3).  As  suggested  by  Fink  and  Fink 
(1981),  the  gymnotoids  are  most  closely  related  to  siluroids 
(node  26). 

The  cyprinifoim  lineage  (node  4)  is  united  by  two  unreversed 
synapomorphies:  an  elongate  yolk  sac  (Figs.  62A  and  68)  and 
the  absence  of  jaw  teeth.  Cypriniforms  and  characiforms  uniquely 
share  large  eyes  at  the  larger  standard  size  (PCI  =  0.6;  Fig.  69). 
This  character  reverses  to  a  plesiomorphous  condition  for  the 
siluriform  lineage.  Synapomorphies  of  siluriforms  include  a  long 
peduncle  at  the  larger  size  (Fig.  69)  (a  unique  state  for  the  group, 
except  for  a  single  reversal  in  Bagarius),  short  head  at  the  larger 
size  (highly  homoplasious),  and  small  eyes  at  the  smaller  size 
(PCI  =  0.0;  Fig.  69)  (unique  except  for  a  reversal  in  Ictalurus). 
The  gymnotoid,  Eigenmannia  (node  28),  expresses  six  auta- 
pomorphies,  two  unique  and  two  occurring  in  only  one  other 
place  on  the  tree.  The  uniquely  derived  conditions  are  a  short 
trunk  at  the  larger  size  (Fig.  69)  and  numerous  postanal  myo- 
meres (Table  29). 

Several  morphometric  characters  make  valuable  contribu- 
tions to  the  phylogenetic  reconstruction.  The  axial  measure- 
ments (head,  trunk,  and  peduncle  lengths)  exhibit  a  clear  trend 
for  increasing  head  and  peduncle  lengths  at  the  expense  of  trunk 
length  through  the  cypriniform  -  characiform  ^  siluroid  - 
gymnotoid  phyletic  sequence.  A  portion  of  the  variation  in  pe- 
duncle size  is  attributable  to  migration  of  the  anus  anteriad  in 


this  phyletic  sequence,  as  evidenced  by  decreasing  preanal  and 
increasing  postanal  myomere  counts  (Table  29).  However,  the 
remaining  peduncle  variation  and  that  of  the  head  length  are 
the  result  of  allometry. 

In  Fink  and  Fink's  (1981)  study,  a  single  character  involving 
the  evolution  of  a  new  structure,  a  pair  of  barbels,  conflicted 
with  their  adult-based  cladogram.  Ontogenetic  evidence  sup- 
ports their  contention  that  the  presence  or  absence  of  barbels  is 
a  poor  indicator  of  relationship  in  ostariophysans.  An  ontoge- 
netic character  for  timing  of  barbel  development  (constructed 
in  the  same  manner  as  described  earlier  for  other  ontogenetic 
characters)  displays  two  distinct  modes  (Fig.  70).  Cyprinids  de- 
velop barbels  during  the  latter  third  of  the  larval  period,  often 
after  finfold  absorption  (i.e.,  as  juveniles).  Siluroids  and  co- 
bitoideans'  do  so  during  the  first  third  of  the  larval  period, 
sometimes  prior  to  hatching.  Although  the  sample  size  of  cob- 
itoideans  is  small,  it  appears  that  they  develop  barbels  somewhat 
later  than  the  siluroids.  Thus,  although  barbels  are  present  in 
adults  of  all  three  groups,  there  is  an  important  difference  in 
these  structures  between  the  groups:  heterochrony.  That  het- 
erochrony is  a  major  cause  of  evolutionary  change  was  amply 
attested  by  Gould  (1977). 

Heterochrony  in  barbels  may  be  an  important  consideration 
for  classification  within  siluroids.  The  number  of  pairs  of  barbels 
(usually  counted  in  the  adult  stage)  is  an  important  character 
for  recognizing  siluroid  families.  At  least  one  pangasiid,  Silonia 
silondia.  has  been  described  in  which  the  larvae  have  three  pairs 
of  barbels  (nasal,  maxillary,  and  mandibular)  that  gradually  be- 
come smaller  until  only  one  pair  of  minute  maxillary  barbels 
are  present  on  the  surface  of  adults  (Karamchandani  and  Mot- 
wani,  1956). 

The  phylogenetic  analysis  presented  here  is  based  on  devel- 
opmental characters.  It  shows  general  congruence  with  the  most 
thoroughly  researched  adult-based  cladogram  (Fink  and  Fink, 
1981);  however,  larval  characters  alone  are  not  as  informative 
as  adult  characters.  Larval  characters  support  the  new  idea  that 
gymnotoids  are  more  closely  related  to  siluroids  than  to  char- 
aciforms. Characiforms  appear  to  be  primitive  ostariophysans 
by  virtue  of  the  basal  location  of  the  relatively  primitive  char- 
aciform Hoplias.  The  apparent  paraphyly  of  characiforms  and 
siluroids  is  due  to  the  lack  of  shared  characters  for  each  of  these 
groups  and  would  be  altered  by  the  reasonable  addition  of  the 
numerous  adult  autapomorphies  discussed  by  Fink  and  Fink 
(1981).  Once  monophyly  is  demonstrated  by  adding  adult  char- 
acters, Hoplias  would  probably  occupy  a  basal  position  (with 
respect  to  the  other  three  characiforms  examined  here)  on  a 
characiform  lineage.  However,  the  position  of  this  lineage  with 
respect  to  that  of  the  cypriniforms  may  or  may  not  agree  with 
Fink  and  Fink's  (1981)  adult-based  cladogram. 

School   of   Natural    Resources,    S.   T.    Dana    Building, 
University  of  Michigan,  Ann  Arbor,  Michigan  48109. 


'  Cobitoideans  included  here  and  in  Fig.  70  were:  Cobitidae— Bo/;a 
.vafir/i;  (Changjiang,  1976);  Cobilis  taenia  (Chyung.  1961;  Koblitskaia, 
1981;  Kokhanova,  1957;  Kryzhanovskii,  1949;  Kryzhanovskii  et  al., 
1951;  Menasse,  1970);  Mtsgurnus  anguillicaudalus  {Chyung,.  1961;  Ko- 
bayasi  and  Moriyana,  1957;  Okada,  1960;  Okada  and  Seiishi,  1938; 
Suzuki,  1955,  1968);  Homalopteridae— A'emac/jei/jis  dorsalis  (Kry- 
zhanovskii, 1949). 


Gonorynchiformes:  Development  and  Relationships 
W.  J.  Richards 


THE  Gonorynchiformes  is  a  small  group  of  fishes  which  have 
been  allied  with  the  clupeiforms  or  salmoniforms  and  most 
recently  have  been  placed  as  a  lineage,  within  the  ostariophysan 
group,  which  includes  also  the  Cypriniformes,  Characiformes, 
and  Siluriformes  (Fink  and  Fink,  1981).  The  group  is  comprised 
of  seven  genera  classified  in  about  four  or  five  families.  The 
most  widely  known  species  is  Chanos  chanos  Forsskil  placed 
in  the  monotypic  family  Chanidae.  The  Gonorynchidae  is  a 
marine  family  of  one  genus  Gonorymchus  and  several  species 
found  in  tropical  waters  of  all  but  the  western  Atlantic  and 
eastern  Pacific.  The  remaining  twelve  or  so  species  are  African 
freshwater  forms  in  the  genera  Kneria.  Parakneria.  Grassei- 
chthys  and  Phractolaemus,  which  may  represent  two  or  three 
families.  The  eariy  life  history  of  Chanos  is  very  well  known 
because  of  the  extensive  culturing;  Gonorymchus  is  poorly  known. 
The  early  life  histories  of  the  freshwater  species  are  unknown. 
Pellegrin  (1935)  notes  that  young  specimens  of  Cromeria  nt- 
lotica  have  a  superficial  resemblance  to  young  Albula.  It  is  ap- 
parent that  this  resemblance  is  to  the  shape  of  juveniles  and  not 


to  a  leptocephalus  stage.  Several  subsequent  papers  have  erro- 
neously reported  that  Pellegrin  said  that  Cromeria  resembled 
larval  Albula. 

Development 

The  early  life  history  of  Chanos  chanos,  the  milkfish,  has  been 
described  by  Delsman  (1926d,  1929b).  Since  Chanos  is  an  im- 
portant aquaculture  organism,  several  recent  papers  have  de- 
scribed various  aspects  of  development,  among  them  the  de- 
scription by  Liao  et  al.  (1979)  is  the  most  complete.  Miller  et 
al.  (1979)  provides  a  good  account  for  separating  them  from 
common  marine  larvae.  To  summarize,  the  eggs  and  larvae 
superficially  resemble  clupeids  and  engraulids  but  differ  in  sev- 
eral trenchant  characters.  The  eggs  as  described  by  Delsman 
( 1 929b)  are  spherical,  1 .2  mm  in  diameter,  lack  oil  droplets  and 
have  a  weakly  segmented  yolk  which  may  be  similar  to  the 
granular  yolks  seen  in  ostariophysans.  Yolk-sac  larvae  have  me- 
lanophores  scattered  over  the  body  and  fin  folds  and  a  myomere 
formula  of  34  -I-  10  (preanal  and  postanal).  As  development 


Fig.  71.  Lateral  and  ventral  views  from  top  to  bottom:  Chanos  chanos.  1 1.7  mm  SL  from  Kumano,  Tanegashima  collected  August  19,  1978, 
drawn  by  J.  C.  Javech;  and  Gonorymchus  abrevialus.  12.8  mm  SL  from  R/V  Shoyo  Maru  station  25,  35°05'N,  144°24.3'E.  collected  on  November 
10,  1963;  drawn  by  J.  C.  Javech. 


138 


RICHARDS:  GONORYNCHIFORMES 


139 


progresses,  the  melanophores  collect  along  the  dorsal  and  ven- 
tral midlines  of  the  trunk.  In  larvae  10-15  mm  SL  (Fig.  71) 
pigmentation  is  variable  with  melanophores  on  the  dorsal  mid- 
line varying  from  one  to  many  and  melanophores  on  the  lateral 
line  varying  from  none  to  many.  The  ventral  midline  has  a 
continuous  streak  of  melanophores  in  sharp  contrast  to  clupeids 
and  engraulids  which  have  melanophores  laterally  on  each  side 
of  the  gut  thus  presenting  two  parallel  streaks  in  ventral  view. 
The  anal  fin  of  Chanos  originates  beneath  the  dorsal  fin  as  in 
engraulids.  In  Hawaiian  waters  meristics  separate  Chanos  from 
Gonorynchiis  and  other  clupeids  and  engraulids.  Chanos  has 
40-46  vertebrae  [44-46  according  to  Miller  et  al.  (1979)  and 
40-45  according  to  Senta  and  Kumagai  (1977)].  Dorsal  rays  are 
14-16,  anal  rays  8-11,  pectoral  rays  17  and  pelvic  rays  10-12 
(Miller  et  al.,  1979). 

Much  less  is  known  about  the  early  life  history  stages  of  Gon- 
orynchus.  Furukawa  (1951)  described  the  larvae  of  G.  ahbrev- 
latus  and  illustrated  18  and  23  mm  specimens.  He  based  his 
identification  on  dorsal  (1 1-1 2)  and  anal  (7-8)  fin  rays,  vertebral 
counts  (55)  and  the  posterior  position  of  the  dorsal  and  anal 
fins.  Hattori  (1964)  illustrated  and  briefly  described  a  series  of 
G.  ahbreviatus  from  8.6  to  90.5  mm.  He  noted  that  the  positions 
of  the  dorsal  and  anal  fins  do  not  shift  during  development. 
Mito  (1966)  illustrates  two  larval  G.  ahbreviatus.  I  examined  a 
series  of  G.  abbreviatus  specimens  and  one  is  illustrated  here 
(Fig.  7 1 ).  The  larvae  resemble  clupeids  with  the  wide  separation 
of  the  dorsal  and  anal  fin.  Pigment  occurs  dorsally  and  ventrally 


on  the  caudal  peduncle  and  extends  posteriad  into  the  bases  of 
the  procurrenl  caudal  rays.  Internal  pigment  occurs  above  the 
hindgut  and  behind  the  brain.  A  few  external  melanophores 
are  present  on  the  top  of  the  head.  Additional  external  mela- 
nophores appear  with  growth.  These  include  a  series  which  de- 
velops as  lateral  spots  increasing  in  number  with  growth.  In  a 
few  specimens  examined  a  15.9  mm  larva  had  one  spot  and 
these  increased  in  number  to  18.  At  23  mm  SL  pigment  also 
appeared  on  the  opercle  and  ventral  rim  of  the  orbit.  The  pelvic 
fin  is  discernible  as  a  bud  in  small  larvae  but  fin  rays  are  not 
defined  until  18  mm  SL.  A  swimbladder  is  not  discernible  on 
any  of  the  specimens  as  it  is  in  clupeids  and  Chanos. 

Relationships 

The  relationships  of  the  Gonorynchiformes  have  been  dis- 
cussed most  recently  by  Fink  and  Fink  (1981).  They  conclude 
that  this  order  is  the  sister  group  of  the  Otophysi  (the  taxon 
which  includes  fishes  with  the  Weberian  apparatus).  Chanos 
and  Gonorynchiis  larvae  more  closely  and  superficially  resemble 
clupeoid  larvae  than  any  other  group.  This  matter  should  be 
thoroughly  investigated  when  early  life  history  aspects  of  the 
freshwater  species  become  better  known.  It  will  be  interesting 
to  see  if  those  larvae  resemble  the  marine  species  or  freshwater 
Otophysi. 

National  Marine  Fisheries  Service,  Southeast  Fisheries 
Center,  75  Virginia  Beach  Drive,  Miami,  Florida  33149. 


Salmoniforms:  Introduction 

W.  L.  Fink 


ORIGINALLY  a  major  portion  of  the  Protacanthopterygii 
of  Greenwood,  et  al.  (1966),  the  order  Salmoniformes  is 
now  the  only  portion  left  in  that  group,  and  the  former  term 
has  ceased  to  have  a  useful  function.  This  erosion  of  the  Pro- 
tacanthoptergyii  has  resulted  from  the  search  for  and  taxonomic 
recognition  of  natural  groups  of  primitive  euteleosts,  a  practice 
that  has  and  is  continuing  to  have  profound  effects  on  fish  clas- 
sification at  all  levels.  This  part  of  the  symposium,  concentrating 
on  the  "salmoniforms,"  places  its  participants  in  the  middle  of 
a  continually  changing  set  of  problems,  some  of  which  have 
been  longstanding.  One  of  the  questions  we  address  here  is 
whether  the  Salmoniformes  as  conceived  by  Greenwood  et  al. 
is  itself  useful  any  more,  and  if  not,  what  are  the  relationships 
of  the  formerly  included  groups.  In  the  years  since  it  was  delin- 
eated, the  Salmoniformes  has  undergone  attrition,  most  notably 
at  the  hands  of  Rosen  (1973).  Of  particular  concern  to  us  is 
whether  there  is  one  large  monophyletic  unit  which  can  be  called 
Salmoniformes,  as  maintained  by  Rosen  (1974),  or  whether 
there  are  several  units,  as  suggested  by  Fink  and  Weitzman 
(1982),  thus  reciuiring  us  to  modify  our  conclusions  and  clas- 
sifications. The  basic  questions  are  these;  (1)  What  are  the  re- 
lationships of  the  Esocoidei  (sensu  Rosen,  1974),  both  to  one 
another  and  to  other  primitive  euteleosts?  (2)  What  are  the 
relationships  of  the  Ostariophysi,  (sensu  Rosen  and  Greenwood, 


1970)?  Do  these  fishes  lie  above  or  below  the  Esocoidei  in  the 
phylogeny?  (3)  What  is  the  pattern  of  relationships  among  the 
traditionally  recognized  "salmoniform"  taxa,  exclusive  of  the 
Esocoidei  and  Ostariophysi?  Is  this  a  natural  division?  (4)  What 
are  the  phylogenetic  relationships  of  and  within  the  Argenti- 
noidei  (sensu  Greenwood  and  Rosen,  1971)?  (5)  What  are  the 
phylogenetic  relationships  of  and  within  the  Osmeroidei?  (6) 
What  are  the  phylogenetic  relationships  of  and  within  the  Sal- 
monidae?  (7)  Where  does  Lepidogalaxias  belong?  (8)  What  are 
the  interrelationships  within  the  stomiiform  fishes?  (9)  What  of 
the  Myclophoidei,  as  recognized  by  Greenwood,  et  al.  (1966)? 
This  "group"  has  been  most  recently  addressed  by  Rosen  (1973) 
in  his  discussion  on  the  Eurypterygii  and  Neoteleostei.  Parts 
of  these  groups  overlap  into  areas  covered  by  this  particular 
part  of  the  symposium,  such  as  placement  of  giganturids,  and 
other  parts  into  non-"salmoniform"  portions  such  as  that  on 
myctophi  forms. 

In  many  ways  this  symposium  is  a  report  on  the  state  of  the 
science  of  fish  classification,  will  summarize  current  ideas  of 
relationships  and,  especially,  will  point  to  where  the  greatest 
need  for  further  research  lies. 

Museum  of  Zoology,  University  of  Michigan,  Ann  Arbor, 
Michigan  48109. 


Esocoidei:  Development  and  Relationships 
F.  D.  Martin 


THE  Esocoidei  consist  of  two  families,  Esocidae  and  Um- 
bridae,  with  one  and  three  genera  respectively  (Nelson, 
1976).  Table  30  lists  all  currently  accepted  species  and  gives 
their  geographic  ranges.  All  recent  classifications  consider  the 
esocoids  as  members  of  the  Salmoniformes  (Greenwood  et  al., 
1966;  Gosline,  1971;  Rosen,  1974;  Nelson,  1976;  and  others). 
All  esocoid  fishes  live  in  freshwater  and  occur  in  temperate  and 
arctic  waters  of  the  Northern  Hemisphere.  All  species  are  pred- 
atory with  Esox  being  primarily  piscivorous.  They  are  distin- 
guished from  other  salmoniform  fishes  by  the  lack  of  the  meso- 
coracoid,  lack  of  pyloric  caeca,  a  single  rudimentary  arch  over 
PUl,  and  a  single  uroneural  (Rosen,  1974).  Table  31  gives  de- 
velopmental features  that  characterize  esocoid  fishes  and  con- 
trasts them  with  Salmonidae  and  Osmeridae. 

Development 

Eggs  are  demersal  and  adhesive  in  most  species  (Breder  and 
Rosen,  1966)  but  Esox  niger  eggs  become  buoyant  at  later  stages 
of  development  and  are  not  adhesive  after  water  hardening  (Jones 
et  al.,  1978).  Eggs  are  of  moderate  size  (1.0  to  2.2  mm  usually) 
(Jones  et  al.,  1978)  and  are  either  scattered  as  by  Esox  or  are 
in  nests  as  with  Umbra  and  Novumhra  (Breder  and  Rosen,  1 966). 


Table  30. 


Genera  and  Species  of  Esocoid  Fishes  and  Geograph- 
ical Ranges. 


Esocidae 
Esox 
E.  lucius 
E.  reicherli 
E.  masquinongy 

E.  niger 

E.  americanus 

Umbridae 

Novumhra 
N.  hiibbsi 

Umbra 
v.  krameri 

V.  linu 

U.  pygmaea 


Datlia 

D.  pectoralis 

D.  asmirabilis 


Holarctic  (Grossman  in  Lee  et  al..  1980). 
Amur  River  region  of  Siberia  (Berg,  1948). 
Eastern  North  America,  primarily  Great 

Lakes  and  Upper  Mississippi  drainage 

(Grossman  m  Lee  et  al.,  1980). 
East  Goast  drainage  of  North  America,  also 

lower  Mississippi  drainage  (Grossman  in 

Leeet  al.,  1980). 
Eastern  half  of  North  America  (Grossman  m 

Leeetal.,  1980). 


Olympic  Peninsula  of  Washington  State 
(Meldnm  m  Lee  el  al.,  1980). 

Middle  and  lower  Danube  System  and  lower 

Dniester  River  (Berg,  1948). 
Southern  Ganada  and  Gentral  United  States 

(Gilbert  m  Leeetal.,  1980). 
Southeastern  New  York  to  Northern  Florida. 

mostly  on  Goastal  Plain  (Gilbert  in  Lee  et 

al.,  1980). 

Arctic  and  sub-Arctic  Alaska  and  eastern  tip 
of  Sibena  (Rohde  in  Lee  et  al.,  1980). 

Amguema  River  basin  of  Siberia  (Gheresh- 
nev  and  Balushkin,  1980). 


Multiple  oil  droplets  occur  with  a  unique  set  of  movements 
producing  alternating  clustering  and  dispersion  as  ontogeny  pro- 
ceeds (Malloy  and  Martin,  1982). 

Larvae  of  nearly  all  species  are  known,  and  developmental 
series  have  been  described  and  illustrated.  Figs.  72  and  73  show 
representative  larvae  oi Esox  and  Umbra.  Those  described  hatch 
relatively  undeveloped,  with  head  flexed  over  and  attached  to 
the  large  yolk  sac;  the  eyes  are  unpigmented.  In  all  species  the 
notochord  is  stout  and  reaches  nearly  to  the  margin  of  the  caudal 
finfold.  During  flexion  the  notochord  extends  well  beyond  the 
developing  hypurals  and  may  form  a  separate  lobe  to  the  de- 
veloping caudal  fin  until  the  hypurals  are  complete.  In  Umbra 
and  Esox  the  pectoral  fin  is  the  first  to  begin  differentiation  (but 
not  form  rays)  with  the  pelvic  fin  the  last  to  develop  fin  rays. 
All  median  fins  differentiate  more  or  less  simultaneously  with 
caudal  starting  ditTerentiation  slightly  ahead  of  the  others. 
Changes  in  body  form  are  gradual  with  no  noticable  point  of 
metamorphosis.  Before  fin  differentiation  is  complete  the  body 


Fig.  72.     Development  of  Esox  niger  from  hatching  to  juvenile. 
Lengths  arc  total  lengths.  (From  Mansueti  and  Hardy,  1967.) 


140 


MARTIN:  ESOCOIDEI 


141 


Common  Cardinal 


Hepatic  Vitelline  Vein 


-'Sublntestinal  Vitelline  Vein 
Common  Cardinal,   /Hepatic  Vitelline  Vein 


Fig.  73.     Early  yolk-sac  and  late  yolk-sac  larvae  of  Umbra  pygmaea. 
(From  Wang  and  Kemehan,  1979.) 


Heart 


5.4  mm  TL 
-'Sublntestinal  Vitelline  Vein 


Fig.  74.  Schematic  representations  of  the  vitelline  venous  systems 
of  Esox  americanus  (upper)  and  Umbra  pygmaea  (lower)— based  in 
part  on  figure  from  Wang  and  Kemehan,  1979. 


Table  31.     A  Comparison  of  Egos  and  Larvae  of  Esocoid,  Salmonoid  and  Osmeroid  Fishes.  Unless  otherwise  noted  information  on 

Umbridae  and  Esocidae  taken  from  Malloy  and  Martin  (1982). 


Egg 
Demersal 
Adhesive 
Oil  droplets 

Size 
In  nests 

Embryo  and  yolk-sac  larva 
Head  deflexed,  adherent  to  yolk-sac 
Eye  pigmented  at  hatching 
Vitelline  circulation 
Common  cardinals 
Hepatic  vitelline  vein 
Sublntestinal  vitelline  vein 
Sublntestinal  v. v.  forming  rete 
Hepatic  v. v.  forming  rete 

Larva 
Vertebrae  (myomeres) 
Adipose  fin 

Dorsal  origin  over  or  behind  anus 
Notochord  forming  a  urostyle 
extending  length  of  hypural 
complex  past  hypurals 

Juvenile  and  adult 
Pyloric  caeca 

Anterior  constriction  of  vertebra 
Pharyngobranchial  1 
Epurals 
Hypurals 

Neural  spine  on  preural  1 
Neural  spine  on  preural  2 


-1- 

-1- 

'-I- 

'  + 

+ 
multiple 

M-2.2mm 

-1-  or  '-± 
multiple 

M.9-'3.4mm 

'±  (mostly  -) 
'•■  '-multiple 
or  *  '-single 

'M. 5-7.0  mm 

-single  or 

"  '^multiple 

-1  mm 

'-I- 

1  _ 

'  + 

1  + 

-I- 
+ 
-t- 
-I- 


'32-""'42 


-I- 
+ 


+ 
+ 
+ 
+ 


'43-67 


-I- 
-I- 


3.4  + 

9.  I0*_ 
9.10  + 

9+  10_ 
9.10_ 

9.10+    _ 


"•46-75 

3.18  + 
4.18_ 


2_ 
2.15  + 

? 
? 
? 

9 
9 


'55-70 


0 

0 

"13-222 

"0-11 

20*«  + 

20***  _     + 

20  _ 

20_ 

- 

- 

+ 

-1- 

0-2 

2 

"2  or  3 

"2  or  3 

5  or  6 

6 

"7 

"6 

-1- 

+ 

"-or  reduced 

"-or  reduced 

fully 

fully 

reduced  or 

"fully 

developed 

developed 

not 

developed 

•  Present  bul  does  not  run  on  surface  of  yolk  sac. 
•"  In  Novumbra  and  Daltia  only  present  in  midabdominal  region  of  juv 
•••  When  present  ther«  is  also  posterior  constriction. 


'  Breder  and  Rosen,  1966. 

'Cooper.  1978 

'  Rajagopal.  1979. 

^  Watling  and  Brown,  1955. 

•  Baugh.  1980 


•  Jones  ctal..  1978. 
'Carbine.  1944. 
"  Uach.  1923. 
'Soin.  1966. 
'"Kunz.  1966. 


'  Bigelow  and  Schroeder,  1963. 

'  Fuiman.  1982b. 

'Nelson.  1972. 

'  Auer,  1981- 

^  Yanagawa.  1 978. 


"  Scott  and  Crossman.  1973. 

'  Hart.  1973. 

■Nagiec,  1979. 

'  Greenwood  and  Rosen.  1 97 1 . 

•Cavender,  1969. 


142 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES -AHLSTROM  SYMPOSIUM 


form  is  basically  that  of  the  adult.  Guts  are  simple  with  no 
elaborations  in  all  species.  At  hatching  Umbra  has  a  shorter  gut 
and  fewer  myomeres  than  Esox  and  this  is  reflected  in  there 
being  5  myomeres  between  the  yolk  sac  and  the  anus  in  newly 
hatching  U.  pygmaea  and  12  in  E.  americanus  (Malloy  and 
Martin,  1982). 

Relationships 

Malloy  and  Martin  (1982)  point  out  three  ontogenetic  char- 
acteristics shared  by  Esox  and  Umbra,  which  indicate  close 
relationship.  The  position  of  the  heart  at  the  time  of  formation 
is  on  the  yolk  sac  anterior  to  and  left  of  the  head.  All  other  fish 
for  which  position  of  the  forming  heart  is  noted  have  it  forming 
under  the  head  in  the  pericardial  cavity  or,  as  in  the  Atherini- 
formes,  near  the  midline  and  anterior  to  the  head.  The  yolk-sac 
circulatory  pattern  consists  of  paired  simple  common  cardinals, 
a  posterior  rete  formed  by  the  subintestinal  vitelline  vein  and 
paired  or  single  hepatic  vitelline  veins  which  enter  the  rete  before 
the  subintestinal  vitelline  vein  joins  the  common  cardinals  at 
the  heart  (see  Fig.  74).  This  differs  from  all  other  salmoniform 
fish  for  which  the  pattern  is  described  (Kunz,  1 964;  Soin,  1 966). 
The  oil  droplets  go  through  a  predictable  series  of  clustering  and 
dispersion.  Oil  droplet  movement  of  this  sort  has  only  been 
documented  previously  by  Ahlslrom  ( 1 968)  for  bathylagid  smelts 
of  the  genera  Bathylagus  and  Leuroglossus. 

McDowall  (1969)  recognized  a  salmonoid-osmeroid-esocoid 
lineage  but  states  "Where  esocoids  fit  into  this  series  of  sub- 
orders and  families  is  not  clear  to  me."  Rosen  (1973)  likewise 


considers  the  esocoids  and  salmonoids  to  probably  be  closely 
related  but  considers  this  alignment  to  be  provisional.  Fink  and 
Weitzman  (1982),  in  contrast,  state  that  they  find  no  evidence 
to  consider  the  esocoids  closely  related  to  the  other  Protacan- 
thopterygii  (sensii  Rosen,  1 974),  which  are  the  Agentinoidei  and 
Salmonoidei  (including  the  Salmonoidea  plus  Osmeroidea). 
Fink  and  Weitzman  list  the  esocoids  as  sedis  mutahilis  at  the 
euteleostean  level  or  as  the  sister  group  to  all  other  euteleosts. 
Soin  (1980),  on  the  basis  of  egg  development  patterns,  feels  that 
the  esocoid  fish  are  incorrectly  placed  as  a  suborder  of  the  Sal- 
moniformes,  however  he  gives  no  guidance  as  to  correct  place- 
ment. While  the  ontogenetic  evidence  presented  in  Table  30  is 
not  conclusive  it  suggests  that  there  is  a  large  difference  between 
the  esocoids  and  the  Salmonoidei  and  this  is  consistent  with  the 
opinions  of  Fink  and  Weitzman. 

The  vertebrae  of  Umbrids  have  a  pronounced  anterior  con- 
striction, giving  them  an  asymmetrical  appearance,  however 
Novumbra  and  Dallia  show  this  characteristic  only  while  young 
and  most  noticeably  in  the  mid-abdominal  region.  In  Esox  the 
vertebrae  are  either  unconstricted  or  are  constricted  both  an- 
teriorly and  posteriorly  so  that  they  appear  symmetrical  (Cav- 
ender,  1969).  Other  differences  between  the  Esocidae  and  the 
Umbridae  are  seen  in  the  Umbridae  having  nine  or  fewer  bran- 
chiostegals,  fewer  infraorbitals,  no  supratemporals  or  intercalars 
and  usually  fewer  than  41  vertebrae  (Wilson  and  Veilleux,  1982). 

Chesapeake    Biological    Laboratory,    University    of 
Maryland,  Box  38,  Solomons,  Maryland  20688. 


Salmonidae:  Development  and  Relationships 
A.  W.  Kendall,  Jr.  and  R.  J.  Behnke 


SALMONIDS  (whitefishes,  ciscoes,  grayling,  trout,  and  salm- 
on) are  highly  important  in  terms  of  aesthetic  appreciation, 
commercial  and  recreational  value,  and  scientific  study.  Studies 
of  the  development  of  salmonids  from  hatching  until  the  time 
of  yolk  depletion,  and  of  the  relationships  among  subfamilies 
and  genera  have  been  largely  neglected  [see  review  of  systematics 
by  Dorofeyeva  et  al.  (1980)]  despite  the  large  body  of  literature 
on  early  embryological  development  and  relationships  among 
species  and  populations.  Salmonids  all  spawn  in  fresh  or  brack- 
ish water,  some  are  anadromous  while  others  are  strictly  fresh- 
water. The  family  is  composed  of  about  10  genera  in  three 
subfamilies:  Coregoninae,  Thymallinae,  and  Salmoninae  (Table 
32)  (Nelson,  1976). 

Along  with  a  precise  homing  ability,  salmonids  tend  to  form 
genetically  isolated  populations.  They  seem  to  be  able  to  occupy 
new  niches  and  habitats  as  these  become  available  in  the  cold 
temperate  parts  of  the  Northern  Hemisphere.  One  result  of  this 
adaptability  is  the  existence  of  taxonomic  problems  mainly  at 
the  species-population  levels  (Utter,  1981). 


Development 

Post-hatching  development  of  salmonids  has  been  little  stud- 
ied (Table  33),  and  only  a  superficial  analysis  of  comparative 
developmental  stages  has  been  attempted  (Soin,  1980).  Thy- 
mallus  and  the  salmonines  share  apparently  advanced  features 
of  development  such  as  large  yolk  sac  with  an  extensive  vitelline 
circulatory  system  and  development  of  rather  uniform  intense 
pigment,  while  coregonines  develop  larvae  that  are  more  typical 
of  other  freshwater  fishes  (Faber,  1970).  Thymallus  seems  inter- 
mediate between  the  coregonines  with  a  "normal"  larval  stage 
and  the  salmonines  in  which  the  larval  stage  is  largely  bypassed 
(the  young  have  fully  formed  fins  by  the  time  the  yolk  is  ab- 
sorbed). Parr  marks  (vertical  blotches  or  bars  of  pigment  over 
the  trunk  of  juveniles)  are  present  in  all  salmonids  except  Cor- 
egomts  a.nd  StenodushuX  are  not  seen  injuveniles  of  other  fishes. 
Norden  (1961)  incorrectly  considered  the  early  stages  of  Core- 
goniis  artedii  as  figured  by  Fish  (1932)  to  be  similar  to  those  of 
Thymallus  arclicus.  He  also  stated  that  "the  development  of 


KENDALL  AND  BEHNKE:  SALMONIDAE 


143 


Table  32.    Characters  that  vary  among  the  Salmonid  Subfamilies. 


Subfamily 

Character 

Coregoninae 

Thymallinae 

Satmoninae 

General 

Genera 

Coregonus,  Prosopium.  Steno- 

Thymallus 

Brachymystax.  Hucho,  Salvelinus. 

dus 

Salmo,  Parasatmo,  Oncorhvnchi 

Species 

30 

4 

32 

Habitat 

freshwater,  few  anadromous 

freshwater 

freshwater  and  anadromous 

Egg  size 

1.8-3.7  mm 

2.5  mm 

3.7-6.8  mm 

Diploid  chromosome  num- 

64-82 

102 

52-92 

bers 

Dorsal  fin  rays 

10-15 

17-25 

8-12 

Dentition' 

Tooth  character 

narrow,  sharp,  2-3  sections 

uniform  in  size 

vary  in  size 

Maxillary 

toothless 

toothed 

toothed 

Dentary 

minute  teeth  restricted  to  ante- 

narrow, teeth  of  uniform  size 

numerous  teeth  of  varying  size  all 

rior  end 

all  along  bone 

along  bone 

Vomer 

small  and  toothless  (except  in 
Stenodus  and  some  Corego- 
nus) 

small,  with  teeth 

long,  with  teeth 

Premaxillary 

small 

large 

large 

Caudal  skeleton- 

Epurals 

3 

3 

2-3^ 

Stegural 

little  developed 

little  developed 

well  developed 

Neural  and  hemal  spine  ex- 

little 

moderate 

large 

pansion 

Urodermal 

present 

absent 

absent 

Neural  spine  on  PU, 

absent 

absent 

present 

Neural  spine  on  PU, 

not  fully  developed 

not  fully  developed 

fully  developed'' 

Cranial  osteology' 

Orbitosphenoid 

present 

absent 

present 

Suprapreopercular 

absent 

absent 

present 

Panetals  meet  at  midline 

yes 

yes 

no 

Hypethmoid 

present 

absent 

usually  absent 

Basisphenoid 

usually  absent 

present 

present 

Uppermost  orbital' 

present 

present 

absent 

'  Vladykov  (li)70). 

■'Cavender  (1970)- 

'  Nordcn  (1961 1, 

*  Some  vanation  withm  Salmonmae  in  these  Iwo  characters.  Those  with  2  cpurais  usually  have  most  extensive  neural  spine  development. 

^  Sometimes  erroneously  termed  dcrmosphenotic;  sometimes  present  in  Salmoninae;  see  Behnke  (1968,  p  9-10). 


the  young  grayling  has  much  in  common  with  that  of  both  the 
coregonines  and  salmonines"  (Norden,  1961:743). 

Among  the  coregonines,  larvae  of  Prosopium  (Faber,  1970; 
Auer,  1982),  L«/c;c7!r/!V5  (Fish,  1932;  Faber,  1970;  Auer,  1982), 
and  Coregonus  (¥'\%\\.  1932;  Faber,  1970;  Auer,  1982)  have  been 
illustrated  and  briefly  described.  All  show  similar  larval  mor- 
phology (Fig.  75).  They  are  rather  slender  with  a  long  preanal 
finfold— the  yolk  being  confined  to  the  anterior  trunk  region. 
The  yolk-sac  length  is  <35%  total  length  (TL),  eye  diameter  is 
<7%  TL,  and  body  depth  at  anus  is  usually  <  10%  TL  (Auer, 
1982).  The  yolk  is  exhausted  before  any  of  the  fins,  except  the 
caudal,  possess  full  complements  of  rays.  Prosopium  eggs  have 
multiple  oil  globules,  while  Leucichthys  and  Coregonus  eggs 
have  a  single  oil  globule  (Auer,  1982).  Pigment  in  preflexion 
and  flexion  larvae  is  mainly  associated  with  the  dorsal  and  ven- 
tral midlines.  Later,  the  body  becomes  more  uniformly  pig- 
mented. Prosopium  develops  parr  marks  during  the  juvenile 
period.  Larvae  oi  Stenodus  are  undescribed  and  they  may  differ 
from  those  described  above,  since  adults  of  this  genus  appear 
quite  divergent  from  the  others  in  this  subfamily. 


Early  development  of  Thymallus  thymallus  has  been  fully 
described  (Penaz,  1975).  They  hatch  with  a  large,  anteriorly 
placed  yolk  sac  that  is  covered  by  a  rather  extensive  vitelline 
circulatory  system,  and  the  preanal  and  postanal  finfolds  are 
about  equal  in  length  (Fig.  75).  The  yolk  sac  is  exhausted  during 
notochord  flexion  and  by  that  time  some  fin  rays  have  developed 
in  all  of  the  fins.  The  larvae  are  rather  heavily  pigmented  during 
this  period.  When  the  fins  have  developed  their  adult  comple- 
ment of  rays,  the  fish  appear  like  juveniles  and  parr  marks  begin 
to  form. 

Early  development  of  all  the  salmonine  genera  and  most  sub- 
genera is  known,  although  several  are  inadequately  described 
(Table  33).  Described  development  of  all  salmonines  is  quite 
similar  (Figs.  76,  77).  Their  eggs  are  among  the  largest  of  all 
teleosts.  They  all  hatch  with  large  yolk  sacs  and  well  developed 
vitelline  circulatory  systems.  The  preanal  finfold  is  shorter  than 
the  postanal  finfold  (except  in  Hucho  where  they  are  about 
equal).  The  preanal  finfold  extends  somewhat  down  the  poste- 
rior of  the  yolk  sac  in  Oncorhynchus.  The  notochord  is  slightly 
flexed  and  some  caudal  rays  are  present.  Yolk-sac  length  is 


144 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  33.     Meristic  Values  and  References  to  Descriptions  of  Larvae  of  Salmonids.  Total  reported  ranges  of  meristic  values  are  given, 

although  the  extremes  of  the  ranges  may  be  rarely  observed. 


Subfamily 
Genus 

Subgenus 


Ranges  of  meristic  values 


References  with 

illustrations  of  flexion 

stage  larvae 


Verte-         Dorsal 
brae'  fin- 


Pec- 

Tolal 

Lateral 

Branch  i 

toral 

Pelvic 

gill 

line 

ostcgal 

fin 

fin 

rakers 

scales 

rays 

Primar\' 
source 


Coregoninae 
Stenodus 

Prosopium 

Coregonus 

Leucichlhys 

Coregonus 

Thymallinae 
Thymallus 

Salmoninae 
Brachymystax 

Hucho 
Hucho 

Parahucho 

Salvelinus 
Sahelinus 

Baione 

Crislivomer 

Salmo 
Salino 

Salmothymus 

Acantholingua 

Plalysalmo 

Parasalmo 

Oncorhvnchus 


64-69     12-19     15-18     16-17        11         19-24       90-110 
Faber  (1970),  Auer(  1982)      50-65     10-15     10-14     13-18       9-12     11-44 


9-12     Scott  and  Cross- 
man  (1973) 
50-108       6-10     Scott  and  Cross- 
man  (1973) 


Fish  (1932);  Faber  (1970).      50-67       8-15       9-16     13-18       8-13     21-64       58-110       7-10     Scott  and  Cross- 

Auer(1982)  man  (1973) 

Fish  (1932),  Faber  (1970),      55-64     10-13       9-14     14-17     11-12     15-78       70-102       6-10     Scott  and  Cross- 


Auer(1982) 
Penaz(1975) 

Smol'yanov  (1961) 
Balon(1956) 

Balon  (1980) 


58-62     17-25     11-15     14-16     10-11      16-33       81-103       7-9 


man  (1973) 

Scott  and  Cross- 
man  (1973) 


58-62     12-15     11-14     15-18 

64-71      12-14     11-13     15-18 
57-62     12-14     12-14     14-17 


9-10     20-30     120-150     10-13     Behnke(1968) 

and  original 

10        10-17     120-150       9-12     Behnke(1968) 

and  original 
9        14-20     110-120       9-12     Behnke(1968) 

and  original 


57-71      10-12       8-10     14-16       9-11      11-51      105-152     10-15     Scott  and  Cross- 
man  (1973) 

Balon  (1980),  Auer(  1982).     57-62     10-14       9-13     11-14       8-9       14-22     110-130       9-13     Scott  and  Cross- 
Martinez  (1983)  man  (1973) 
Fish  (1932),  Balon  (1980).      61-69       8-10       8-10     12-17       9-10     16-26     116-138     10-14     Scott  and  Cross- 


Auer(1982) 

Auer(1982).  Martinez 
(1983) 


Auer(1982),  Martinez 

(1983) 
Auer(1982) 


man  (1973) 

54-62     10-15       8-13     12-16       9-10     14-25     100-130     10-12     Behnke(1968) 

and  original 

56-60     13-15     11-13     12-14       9-10     25-32     100-115     10-12     Behnke(1968) 

and  original 

52-59     11-13     10-12     11-13       9-10     18-22       95-110       9-11     Behnke(1968) 

and  original 

57-59        13  11  14  9        23-24     109-110     10-11     Behnke(1968) 

and  original 

55-67       8-12       8-12     11-17       9-10     14-28     100-150       9-13     Scott  and  Cross- 
man  (1973) 

61-75       9-16     12-19     11-21       9-11      18-43     120-160     11-19     Scott  and  Cross- 
man  (1973) 


Overall  ranges 


50-75   8-25 


-19  11-21   8-13  10-78   50-160   6-19 


'  Vanalions  exist  in  the  literature  in  how  many  of  last  3  upturned  vertebrae  are  counted;  some  authors  omit  the  last  3  upturned  vertebrae. 

'  Includes  rudiments  where  specified.  A  variation  of  2-3  rays  may  result  from  different  methods  of  counting  (whether  unbranched  or  rudimentary  rays  are  included). 


>35%  TL,  eye  diameter  >7%  TL,  and  body  depth  at  anus 
usually  >  10%  TL  (Auer,  1982).  Pigmentation  is  unifoimly  heavy 
at  hatching  or  later  in  the  yolk-sac  stage.  The  median  fins  de- 
velop rays  before  the  paired  fins.  By  the  time  the  yolk  is  absorbed 
the  finrays  have  completed  foimation  and  the  fish  takes  on  a 
juvenile  appearance.  Thus,  the  yolk  remains  a  source  of  nutri- 
tion throughout  the  larval  stage. 

Relationships 

Although  salmonids  are  considered  to  be  living  representa- 
tives of  the  basal  stock  from  which  euteleostean  evolution  pro- 
ceeded, there  is  no  clear  consensus  on  their  relationships  to  other 
fishes.  Since  there  are  differing  opinions  on  the  relationships 


between  the  major  teleostean  lineages  (i.e.,  the  divisions  of 
Greenwood  et  al.;  1966),  it  is  difficult  to  select  representatives 
of  outgroups  to  compare  with  the  salmonids.  Recent  studies 
(Rosen,  1974;  Fink  and  Weitzman,  1982;  Fink,  this  volume) 
have  pointed  out  that  the  Protacanthopterygii  and  even  the 
Salmoniformes  are  probably  not  monophyletic  taxa.  The  sal- 
monids along  with  the  galaxioids,  osmeroids,  and  argentinoids, 
may  form  a  group  (Salmonae)  that  is  the  primitive  sister  group 
of  the  neoteleostei.  However,  the  relationships  among  these 
groups  is  not  clear,  and  the  salmonids  may  be  closer  to  the 
neoteleostei  than  to  these  other  groups  with  which  they  have 
frequently  been  aligned  (Fink  and  Weitzman,  1982;  Lauder  and 
Liem,  1983;  Fink,  this  volume).  Some  primitive  teleost  traits 


KENDALL  AND  BEHNKE:  SALMONIDAE 


145 


Fig.  75.     Flexion  stage  larvae  of:  (A)  Coregonus  (Leucichthys)  artedii  (17.5  mm);  (B)  Coregonus  (Coregonus)  clupeaformis  (18.5  mm);  (C) 
Thymatlus  ihymallus  {\6.0  mm).  A  and  B  from  Fish  (1932),  C  from  Penaz  (1975). 


Table  34.    Characters  that  vary  among  the  Coregonine  Genera  and  Subgenera  (sg)  mainly  from  Norden  (1961)  and  Cavender 

(1970). 


Coregonu. 

Prosopium 

Character 

Coregonus  (sg) 

Leucichthys  (sg) 

Stenodus 

Species 

8 

17 

1 

1 

Habitat 

Some  occasionally  anad- 

Several  anadromous 

Freshwater 

Anadromous 

romous 

Basibranchial  plate 

Absent 

Absent 

Present 

.Absent 

Parietal  bones  meet  along 

Yes 

Yes 

Yes 

No:  narrowly  separated 

midline 

Postorbitals  in  contact  with 

Yes 

Yes 

Yes 

No 

preopercle 

Parr  marks 

Absent 

Absent 

Present  in  some 

Absent 

Flaps  between  nostnls 

2 

2 

1 

2 

Mouth  size 

Small 

Moderately  large 

Small 

Large 

Teeth 

Weak  or  none 

Weak  or  none 

Weak  or  none 

Many,  small 

Mouth  position 

Subterminal 

Supenor  or  terminal 

Subterminal 

Terminal 

Vomer 

Small,  toothed  in 

some 

Small,  toothed  in 
some 

Small,  toothless 

Large,  toothed 

First  supraorbital 

Moderate 

Moderate 

Short 

Long 

Supraethmoid 

Short 

Short 

Long 

Short 

146 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  76.  Flexion  stage  larvae  of:  (A)  Brachymystax  lenox(\1.2  mm);  (B)  Hucho  (Hucho)  hucho  (20.8  mm);  (C)  Salvelinus  (Sahelinus)  alpinus 
(19.8  mm);  (D)  Sahelinus  (Cnslivomer)  namaycush  (approx.  20.4  mm).  A  from  Smoryanov  (1961),  B  from  Balon  (1956),  C  and  D  from  Balon 
(1980). 


Fig.  77.     Flexion  stage  larvae  of;  (A)  Sahelinus  (Batone)  fontmalis  ( 1 4.0  mm);  (B)  Parasahno  gairdnert  ( 1 4.0  mm);  (C)  Parasalmo  darki  (14.2 
mm);  (D)  Salmo  trutta  (14.0  mm);  (E)  Oncorhynchus  tshawytscha  (25.0  mm).  A-D  from  Martinez  (1983),  E  original. 


KENDALL  AND  BEHNKE:  SALMONIDAE 


147 


148 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


large  mouth    w  7i3 
many  small  teeth  n  753 
^large  toothed  vomer 

^  N  720 


no  teeth  on  vomer 

N  720 


/ 

mostly: 
small  mouth  n  7S3 
small  teeth  on  vomer 
of  young  only  n=720 


parr  marks  absent    n  743 
two  flaps  between  nostril    n  71 
enlarged  first  supraorbital  n  71 
Joss  of  basibranchial  plate  n  71 


/        \\ 


N  727  slightly  notched  ethmoid 
cartilage 

light  spots  1 
Palatine  vomer    strong  ascending  pre- 

teeth  form  z^  maxillary  process  J  n 

j"  shaped  "C      gap  between  palatine  I 

band    n  753^v  vomer  teeth  1 


Thymallinae 


N  72S  notched  ethmoid 

\  cartilage 
N  732  palatine  vomer 
teeth  form  a  "T" 


/ 


Coregoninae 

N  679  no  teeth  on  maxillary 

c  9  one  urodermal 

c  11  small  neural  spine  on  PU^ 

N-753  <  16  dorsal  rays 

N-750  general  loss  of  teeth 


c  11     small  neural  spine  on  PU-, 
N-679  >  16  dorsal  rays 
N-e79  no  orbitosphenoid 


increase  in  size  and  amount  of  yolk  in  egg 
bypass  larval  stage 


Salmonidae 


100) 


•  N  739 

•  •N  752 
♦IM  743 
*N  752 

•  N  752 

•  *C  27 

•  •  N  752 

•  •rj  752 

•  *N  752 

•  "N  752 

•  •C27 
«»C  11 

•  *c  n 

*?C9 


Tetraploid  karyotype  (  2n  chromosomes  -   ' 

Axillary  pelvic  process 

Three  upturned  caudal  vertebrae  (  two  ural  centra) 

Parr  marks  in  juveniles 

Three  post  cleithra 

Mesopterygoid  toothless 

Last  four  hemal  spines  and  parhypural  fit  together  (  peg  and  socket) 

Adipose  fin  present 

Oviducts  incomplete  or  absent 

Mesocoracoid  present 

Opisthotic  present 

Principal  caudal  rays  =  19 

Three  epurals 

Full  neural  spine  on  PU, 

Two  hypurais  (ventral)  on  U^,  4  hypurals  (dorsal)  on  U,  . 

1  long,  2  short  uroneurals 


N  726  blunt  pointed  ethmoid  cartilage 
N  732   gap  between  palatine  vomer  teeth 
M  '53  no  ascending  premaxillary  process 
N  753  postorbitals  contact  preopercular 
N  753  opisthotic  touches  prootic 
N  728   reduced  dorsal  fontanelles  in  adult 


well  developed  stegural 

expanded  caudal  neural  and 

hemal  spines 

neural  spine  on  PU. 

large  neural  spine  on  PU, 
N-679  parietals  separated  by 

supraoccipital 
N  739  small  scales  (">  100  in 

lateral  Ime) 
N-679  suprapreopercular  present 
N  731   curved  preopercular 
N  736   dorsal  rays  <  16 
e-9       reduction  or  loss 

of  hypethmoid 


B  -  Behnke  1968 
N  -  Norden  1961 
C  -  Cavender.  1970 
H  =  Hol£ik,1982 

(number  refers  to  page 

in  above  references) 
*  Salmonidae  (synapomorph  for  family) 
""  Salmonoidei  (synapomorph  for  suborder) 
■••  Shared  primitive  (plesiomorph)  character 

with  other  "primitive"  teleosts 


Fig.  78.     Hypothesis  of  relationships  among  extant  saimonid  genera.  Groupings  and  branching  points  are  based  largely  on  a  consensus  of  recent 
literature  and  are  not  the  result  of  a  strict  cladistic  analysis. 


possessed  by  salmonids  include  lack  of  oviducts,  presence  of 
abdominal  pores,  and  three  upturned  caudal  vertebrae  sup- 
porting the  hypurals.  Salmonids  are  autapomorphic  with  about 
twice  the  DNA  content  of  other  '^salmoniform"  families,  ap- 
parently the  result  of  having  a  common  tetraploid  ancestor.  The 
salmonids  possess  an  adipose  fin,  a  mesocoracoid,  pyloric  caeca, 
and  the  vestige  of  a  spiral  valve  intestine.  The  gill  membranes 
extend  far  forward  free  from  the  isthmus  and  there  is  a  pelvic 
axillary  process.  Two  shared  derived  features  of  the  salmonids 
and  neoteleostei  are:  1)  the  articulation  of  both  the  basioccipital 
and  exoccipital  with  the  first  vertebra,  and  2)  the  presence  of  a 
medial  cartilage  between  the  ethmoid  and  premaxilla  (Fink  and 
Weilzman,  1982). 

Although  it  is  not  possible  at  present  to  perform  a  meaningful 
cladistic  analysis  of  the  salmonids,  some  evidence  is  available 
in  the  literature  which  can  contribute  to  such  an  analysis  (Fig. 
78).  Cavender  (1970)  compared  the  osteology  of  leptolepids. 
extinct  fish  thought  to  represent  the  basal  teleost  condition,  with 
that  of  the  salmonids.  He  found  several  characters  that  indicated 
1)  that  the  salmonids  are  monophyletic,  and  2)  how  the  three 
subfamilies  of  salmonids  are  interrelated.  The  coregonines  ap- 


peared to  be  most  similar  to  the  leptolepids,  the  thymallines 
more  derived  than  the  coregonines,  and  the  salmonines  more 
derived  than  the  thymallines.  Reshelnikov  (1975).  on  the  basis 
of  several  types  of  characters,  suggested  elevating  the  subfamilies 
to  familial  status. 

Coregoninae  contains  about  30  species  in  three  genera.  They 
are  mainly  freshwater,  and  produce  rather  small  eggs,  compared 
to  those  of  the  other  two  subfamilies.  They  share  several  ad- 
vanced characters  with  the  other  subfamilies,  indicating  that 
salmonids  are  monophyletic,  but  lack  a  number  of  advanced 
character  states  possessed  by  the  other  two  subfamilies,  as  these 
branched  oflTafter  the  coregonines.  Within  the  coregonines,  Pro- 
sopium  seems  least  diverged  (Table  34).  Sienodus  shows  several, 
possibly  secondarily  derived  character  states  concordant  with 
feeding  on  large  active  prey  (expanded  dentition,  large  mouth). 
Coregoniis,  which  seems  to  be  a  sister  group  to  Stenodus.  is 
separated  into  two  subgenera:  Leucichthys  with  adaptations  for 
plankton  feeding,  and  C orego nus  y^h\ch  are  mainly  benthic  feed- 
ers. 

Thymallinae  contains  one  genus,  Thymallus,  with  about  four 
species  in  freshwater  of  the  colder  parts  of  the  Northern  Hemi- 


KENDALL  AND  BEHNKE:  SALMONIDAE 


149 


Table  35.    Characters  that  vary  among  the  Salmonine  Genera. 


Characters 

Brachvmyslax 

Hucho 

Salvetinus 

Salmo' 

Parasalmo^ 

Oncorhynchus^ 

Subgenera 

Hucho.  Para- 
hucho 

Sahelinus, 
Baione.  Crisli- 
vomer 

Salmo,  Salmo- 
ihy/mts.  Acan- 
iholingua. 
Ptatvsahno 

Species 

2 

3-5 

8 

8 

5 

6 

Habitat 

freshwater 

freshwater  and 

freshwater  and 

freshwater  and 

freshwater  and 

usually  anadro- 

anadromous 

anadromous 

anadromous 

anadromous 

mous 

Mouth  size 

small 

large 

large 

large 

large 

large 

Teeth  on  shaft  of 

no 

no 

no 

yes 

yes 

yes 

vomer 

Palatine-vomer- 

U-shaped  band 

U-shaped  band 

teeth  narrowly 

teeth  narrowly 

teeth  narrowly 

teeth  widely  sepa- 

ine teeth 

separated 

separated 

separated 

rated 

Postorbitals  con- 

no 

no 

no 

no 

no 

yes 

tact  preopercle 

Supraethmoid 

long,  with  nu- 

broad, with  nu- 

long, with  nu- 

notched poste- 

notched poste- 

deeply notched 

shape 

merous  poste- 
rior projec- 
tions 

merous  short 
posterior  pro- 
jections 

merous  poste- 
rior projec- 
tions 

riorly 

riorly 

posteriorly 

Ascending  pre- 

intermediate 

intermediate 

extended  and 

intermediate 

intermediate 

none 

maxillary  pro- 

sized 

sized 

well  developed 

sized 

sized 

cess 
Opisthotic  touch- 

no 

no 

no 

no 

no 

yes 

es  prootic 

Dorsal  fonta- 

persistent 

persistent 

persistent 

persistent 

persistent 

reduced  in  adult- 

nelles 

Egg  size 

4-5  mm 

large 

4-5  mm 

5-7  mm 

large 

large 

Diploid  chromo- 

92 

84 

78-84 

56,  80-82' 

56-70 

52-74 

somes 

Dark  spots-light 

yes 

yes 

no 

yes" 

yes 

yes 

background 

'  There  is  lack  of  agreement  on  the  relationships  between  these  laxa;  e.g.,  some  consider  Parasalmo  a  subgenus  in  Salmo.  while  others  would  also  consider  Oncorhynchus  a  subgenus  of  Salmo. 

*  Retained  in  O.  ma.sou 

*  Salmo  salar  has  56-60  diploid  chromosomes- 

*  Salmo  marmoratus  and  S  platycephalus  have  no  dark  spots. 


sphere.  They  have  several  character  states  that  seem  advanced 
over  those  seen  in  coregonines.  They  are  moderate-sized,  gen- 
eralized insectivores  (Table  32). 

Salmoninae  contains  four  to  six  genera,  depending  on  opin- 
ions over  the  relationships  among  the  species  in  Salmo,  Par- 
asalmo, and  Oncorhynchus  (Table  35).  These  seem  to  be  the 
most  advanced  of  the  salmonids,  and  share  several  character 
states  that  are  derived  compared  to  the  other  two  subfamilies 
(Table  35).  Holcik  (1982)  presented  evidence  which  suggests 
that  the  genera  Hucho,  Brachymysla.x,  and  Salvelinus  form  one 
lineage;  Parasalmo  and  Salmo  another;  and  Oncorhynchus  a 
third.  Salmonines  are  mainly  active  predators  and  most  tend 
toward  an  anadromous  life  histoi7. 

Early  life  history  and  developmental  information  should  con- 
tribute to  the  rigorous  analysis  of  characters  that  will  be  required 


to  validate  the  foregoing  hypotheses  about  relationships.  Such 
information  is  not  presently  available  in  the  literature,  but  should 
be  readily  obtainable,  since  so  many  of  these  fishes  are  routinely 
reared  in  laboratories  and  hatcheines.  Developmental  infor- 
mation seems  particularly  promising  in  this  family,  since  a  wide 
range  of  the  life  history  patterns  are  present  and  larvae  can  be 
superficially  grouped  according  to  their  representative  subfam- 
ilies. 


(A.W.K.)  Northwest  and  Alaska  Fisheries  Center,  2725 
MoNTLAKE  Blvd.  E.,  Seattle,  Washington  98112  and 
(R.J.B.)  Department  of  Fishery  and  Wildlife  Biology, 
Colorado  State  University,  Fort  Collins,  Colorado 
80523. 


Southern  Hemisphere  Freshwater  Salmoniforms:  Development 

and  Relationships 

R.  M.  McDowALL 


SEVERAL  family-level  groups  of  diadromous  salmoniform 
fishes  are  found  in  cool-temperate  southern  hemisphere 
fresh  waters,  forming  an  obvious  ecological  counterpart  to  the 
northern  cool-temperate  Salmonidae,  Osmeridae,  Plecoglossi- 
dae,  Salangidae,  etc.  With  the  exception  of  a  single  species,  in 
a  high  elevation  lake  in  New  Caledonia,  they  are  all  south  of 
about  latitude  28°S.  They  occupy  all  of  the  main  land  masses 
(Australia,  New  Zealand,  South  America,  South  Africa)  and 
some  of  the  more  distant  southern  islands  (Lord  Howe.  Chat- 
hams,  Aucklands,  Campbell,  Falklands).  Diagnostic  familial  and 
generic  characters  are  listed  in  Table  36. 

Familial  arrangement  of  these  fish  varies  from  including  all 
in  a  single  purportedly  monophyletic  family  Galaxiidae  (Nelson, 
1972),  through  two  families  in  separate  sub-orders  (Rosen,  1974) 
to  four  families  in  one  or  two  suborders.  There  are  two  obvious 
and  widely  accepted  familial  groupings:  Galaxiidae— Aplochi- 
tonidae  and  Retropinnidae  — Prototroctidae  (McDowall,  1969). 
The  most  recent  view  (Fink  and  Weitzman,  1982)  suggests  that 
these  four  family  level  taxa  are  possibly  all  of  osmeroid  deri- 
vation agreeing  with  my  own  evaluation  (McDowall,  1 969),  and 
in  contrast  with  Rosen  (1974  — he  links  galaxiids  and  aplochi- 
tonids  with  salmonoids;  retropinnids  and  prototroctids  with 
osmeroids).  The  southern  taxa  are  all  clearly  primitive  prota- 
canthopterygians  of  salmoniform  type.  Beyond  that  little  can 
be  said  other  than  that  a  further  search  of  additional  character 
complexes  is  needed  to  clarify  relationships. 

Within-family  relationships  are  little  studied.  Three  of  the 
southern  families  (Retropinnidae,  Prototroctidae,  Aplochiton- 
idae)  can  be  dealt  with  more  simply  than  the  fourth  (Galaxiidae). 

Retropinnidae  (Australia  and  New  Zealand— see  McDowall, 
1979).  — four  species  in  two  genera:  Present  state  of  knowledge 
does  not  permit  explicit  recognition  of  affinities.  Elongation  of 
the  alveolar  process  in  the  premaxilla  of  Stokellia  anisodon  is 
an  advanced  character  which  leaves  three  species  of  Retropinna 
with  the  primitive  condition  (alveolar  process  short,  maxilla 
sometimes  toothed).  Stokellia  also  has  unossified  gill  rakers  (an 
"advanced"  but  "loss"  condition)  and  high  scale  count  (100 
compared  with  70  or  less  in  Retropinna— which  is  the  derived 
condition?) 

Prototroctidae  (Australia  and  New  Zealand— see  McDowall. 
1976).— Two  species  in  one  genus.  Two  congeneric  species  pose 
no  phylogenetic  problems.  The  only  significant  question  to  ask 
is  "How  do  these  species  relate  to  the  Retropinnidae?"  Answers 
to  this  question  have  not  yet  been  sought. 

Aplochitonidae  (Tasmania  and  South  America— see  McDowall. 
197 la).— Three  (perhaps  four)  species  in  two  genera.  Mono- 
phyly  of  the  Aplochitonidae  (Aplochiton  and  Lovettia)  should 
not  be  assumed.  Inclusion  of  Lovettia  in  the  Gala.xias-Aplo- 
chiton  assemblage  is  supported  by  characters  in  Table  36  but 
Lovettia  has  such  reduced  osteology  that  a  search  for  characters 


in  other  structural  systems  is  needed  before  its  relationships  can 
be  clarified.  Inclusion  oi Lovettia  in  the  Aplochitonidae  is  based, 
in  part,  on  history  (it  has  always  been  there!)  and  in  part,  on 
the  fact  that  it  is  a  "galaxioid"  with  the  dorsal  fin  over  the  pelvics 
and  an  adipose  fin  present  (like  Aplochiton  and  unlike  Gala.xias). 

Galaxiidae  (.Australia.  New  Caledonia,  Lord  Howe.  New  Zea- 
land. South  America.  South  Africa).  S'w  genera  with  37  species 
distributed  as  follows:  Gala.xias,  24— all  areas  but  New  Cale- 
donia; Paragala.xias,  4— Tasmania;  Neochanna,  3  — New  Zea- 
land; Gala.xiella,  3  — Australia;  Brachygala.xias,  2  — South 
America;  NesogalcLxias.  1  —New  Caledonia.  This  larger  and  more 
complex  family  offers  scope  for  phylogenetic  analysis  that  has 
had  little  attention. 

Relationships 

Previous  studies  of  within-family  relationships  have  been 
based  on  morphological  similarity  (McDowall,  1 970),  phenetics 
based  on  muscle  myogens  (Mitchell  and  Scott,  1979),  or  den- 
drograms derived  from  cluster  analysis  of  morphometric  or  me- 
ristic  data  (Campos,  1979).  Johnson  et  al.  (1981,  1983)  have 
sought  to  establish  relationship  on  the  basis  of  karyotypes  and 
multivariate  analysis  of  morphometric  and  meristic  characters 
in  the  diverse  Tasmanian  fauna. 

The  only  attempt  at  a  "strictly  phylogenetic"  interpretation 
of  within-family  relationships,  by  Rosen  (1978),  was  based  on 
misinterpretation  of  character  states  and  a  limited  perception 
of  variation  in  the  family,  and  achieved  nothing  (McDowall, 
1980).  A  broad  and  strictly  phylogenetic  analysis  of  galaxiid 
inter-relationships  is  not  yet  available  and  probably  depends  on 
examination  of  additional  character  complexes. 

On  the  basis  of  out-group  comparisons  (all  salmonoid— os- 
meroid—galaxioid  families  have  members  that  are  diadromous) 
it  is  my  view  that  diadromy  in  the  Galaxiidae  is  a  primitive 
character.  It  is  represented  in  at  least  six  species. 

Diadromous  species  tend  to  be  large  and  generalised  in  char- 
acter, but  with  specific  adaptations  to  habitats  occupied  during 
freshwater  life.  Vertebral  numbers  are  high  (>  60)  and  ray  counts 
in  pelvic  (7)  and  caudal  (16)  fins  very  stable. 

There  are  indications  of  close  relationships  with  diadromous 
stocks,  e.g.,  Gala.xias  maculatus  seems  likely  to  be  a  neotenous 
derivative  of  some  other  diadromous  galaxiid;  distinctive  ju- 
venile colour  patterns  may  relate  G.  argenteus  to  G.  fasciatus 
and  perhaps  G.  truttaceus. 

There  are  numerous  landlocked  populations  of  diadromous 
species,  and  present  interpretations  are  that  several  species  are 
derived  by  isolation  following  landlocking,  e.g.,  G.  auratus 
(landlocked)  derived  from  G.  truttaceus  (diadromous)  in  Tas- 
mania; G.  gracilis  from  G.  maculatus  in  New  Zealand. 

Wholly  freshwater  species  tend  to  be  the  more  specialised 
members,  in  which  there  is  often  dwarfing,  reduced  vertebral 
counts,  greater  meristic  instability,  as  well  as  the  loss  of  the 
distinctive  marine  juvenile  stage.  Some  freshwater  groups  have 


150 


McDOWALL:  SALMONIFORMS 


151 


not  yet  recognised  origins  within  the  diadromous  stocks  and 
there  is  identifiable  speciation  related  to  known  geo-tectonic 
events.  The  relationships  of  some  of  the  more  distinctive  species 
groups— Neochanna  (New  Zealand),  Galaxiella  (Australia),  and 
including  geographical  outliers  like  Gala.xias  zebratus  (South 
Africa)  and  Nesogalaxias  neocaledonicus  (New  Caledonia)— re- 
main obscure.  Previous  inclusion  of  Australian  and  South 
American  species  in  Brachygalaxias  is  ill-founded,  on  present 
data,  and  confuses  the  understanding  of  relationships. 

An  interesting  phylogenetic  problem  in  the  Galaxiidae  in- 
volves the  diminutive  Tasmanian  Paragalaxias.  with  four  species 
in  high  elevation  lakes  that  probably  pre-date  Pleistocene  gla- 
ciations.  Paragalaxias  is  distinctive  in  having  the  dorsal  fin 
origin  only  a  little  behind  the  pelvic  bases.  In  this  regard  it 
resembles  aplochitonids  differing  from  all  other  galaxiids  in  which 
the  dorsal  origin  is  close  to  the  level  of  the  vent/anal  origin. 
Thus  is  Paragalaxias  a  galaxiid  in  which  the  dorsal  fin  has 
migrated  forwards,  the  resemblance  to  Aplochiton  being  con- 
vergent or  is  it  an  aplochitonid  in  which  the  anterior  dorsal  fin 
position  is  primitive  but  in  which  the  adipose  fin  has  been  lost? 
Examination  of  additional  character  complexes  in  which  gal- 
axiids and  aplochitonids  differ  is  needed  to  clarify  this  question. 

The  preceding  discussion  makes  it  evident  that  relationships 
between  and  within  the  southern  diadromous  salmoniforms  re- 
main in  need  of  clarification.  Only  the  Galaxiidae  is  large  and 
diverse  enough  to  provide  fertile  ground  for  a  study  of  within- 
family  phylogeny.  In  all  the  families,  species  and  characters  are 
conservative  in  nature  and  lack  distinctive  or  extreme  speci- 
alisation. Inter-specific  differences  tend  to  be  expressed  as  changes 
in  meristic  characters  (like  vertebral  and  fin  ray  counts),  often 
to  presence/absence  character  states  (pyloric  caeca,  canine  teeth) 
and  sometimes  to  distinctive  and  stable  differences  in  colour 
patterning.  There  are  few  readily  evident  characters  that  are 
indicative  of  major  phyletic  lineages.  Possibly  investigation  of 
laterosensory  papillary  rows  will  be  informative.  At  present, 
establishment  of  phylogenies  appears  difficult.  A  study  of  re- 
lationships using  DNA  hybridisation  techniques  (Sibley  and 
Ahlquist,  1981)  is  at  present  in  early  planning  stages. 

Life  History  Patterns  and  Reproduction 

In  general  life  history  patterns  are  understood  although  details 
are  sparse.  There  are  broad  similarities  in  patterns. 

Retropinnidae.— Aspects  of  early  life  history  have  been  de- 
scribed by  Milward  {1966  — Retropinna  sewon/— Australia), 
Jolly  (1967-«.  retropinna— N.Z.)  and  McMillan  (1961— Sto- 
kellia  anisodon—N.Z.).  The  eggs  are  tiny— 0.5  to  0.6  mm  in 
lacustrine  R.  retropinna,  0.95  mm  in  R.  semom.  They  are  de- 
mersal and  adhesive,  spherical,  without  distinctive  features.  They 
are  a  pale  straw  colour.  They  are  deposited  on  sandy  bottoms 
in  lower  river  reaches  or  estuaries  (around  lake  shores  in  land- 
locked populations),  where  development  occurs;  development 
is  relatively  slow  ( 10-20  days)  and  description  of  development 
shows  nothing  distinctive  (Fig.  79).  Newly  hatched  larvae  in 
some  species  go  to  sea.  In  others  they  are  lacustrine  or  riverine. 
Larvae  at  hatching  are  small  (2-5  mm),  very  slender  and  elon- 
gated, the  yolk  sac  with  a  single  oil  globule,  and  situated  ante- 
riorly beneath  the  opercular  openings/pectoral  fins.  The  gut  is 
long,  the  vent  at  about  70%  of  length.  A  continuous  finfold 
encompasses  the  trunk.  Pectoral  fin  buds  are  present.  Newly 
hatched  larvae  are  positively  phototropic.  Pigmentation  and 
later  development  are  undescribed.  Juveniles  from  a  summer- 


Table  36.  Character  States  in  Principal  Genera  of  Southern 
Freshwater  Salmoniforms.  (*  except  Paragalaxias;  +  present,  -  ab- 
sent; u  uniserial;  m  muUiserial;  1  parhypural  +  hypurals;  2  tubercles  in 
Lmetlia  may  not  be  comparable  with  others).  Figures  are  "usual"  al- 
though variants  are  known.  The  divergent  galaxiid  genera  are  excluded 
(Paragalaxias,  Galaxiella,  Neochanna,  etc.). 


Proto- 

Retro- 

Galaxi- 

Irocli- 

pinni- 

idae 

Aplochilonidae 

dae 

dac 

Ga- 

Lovel- 

Prolo- 

Retro- 

Characters 

laxtas 

Aplochiton 

tia 

Iroctes 

pinna 

Dorsal  fin 

Over  pelvics 

x 

X 

X 

X 

Over  anal 

X* 

X 

Adipose 

- 

+ 

+ 

+ 

+ 

Scales 

- 

- 

- 

+ 

+ 

Homy  keel 

- 

- 

- 

+ 

+ 

Cucumber  odour 

- 

- 

- 

+ 

+ 

Pyloric  caeca 

4- 

-1- 

+ 

- 

- 

Vomerine  shaft 

long 

- 

long 

short 

short 

Vomerine  teeth 

- 

- 

- 

+ 

+ 

Basi  branchial 

- 

- 

- 

-1- 

+ 

teeth 

Palatine  teeth 

- 

- 

- 

+ 

+ 

Mesopterygoidal 

u 

u 

u 

m 

m 

teeth 

Extrascapular 

- 

- 

- 

+ 

-1- 

Ectopterygoid 

~ 

slender 
splint 

~ 

+ 

+ 

Coracoid-cleithrum 

- 

- 

- 

+ 

+ 

process 

Posterior  pubic 

- 

- 

- 

+ 

-1- 

symphysis 

Pubic  foramen 

- 

- 

- 

+ 

+ 

Caudal  skeleton 

1  +5 

1  +  5 

1  +5 

1  +6 

1  +6 

Branched  caudal 

14 

14 

14 

16 

16 

rays 

Nuptial  tubercles 

- 

- 

+ 

+ 

+ 

Ovaries 

both 

both 

both 

left 

left 

autumn  spawning  may  return  to  fresh  water  the  following  spring 
and  are  transparent  and  elongate;  mostly  mature  adults  return 
one  year  later  (age  about  2  years)  to  spawn  and  die  (see  Jolly, 
1967;  McMillan,  1961;  Milward,  1966). 

Protolroclidae.-lAXXXe  is  known  of  this  family,  with  one  species 
extinct  the  other  rare.  McDowall  (1976)  and  Berra  (1982)  have 
described  what  is  known  of  life  histories.  The  eggs  are  small 
(~  1  mm)  round  and  demersal,  and  are  probably  deposited  in 
upstream  fresh  waters.  The  larvae  are  not  known  but  believed 
to  be  carried  to  estuaries  or  the  sea  to  develop,  probably  for 
about  six  months,  and  return  to  freshwater  in  spring  as  slender 
transparent  juveniles  (Fig.  80). 

Aplochilonidae.  — \n  Lovettia,  mature  adults  migrate  from  the 
sea  in  spring  to  spawn  in  fresh  water,  and  are  strongly  dimorphic. 
The  male's  reproductive  opening  migrates  forward  to  the  isth- 
mus and  the  opercular  flaps  become  elongated  and  papillated. 
Fecundity  is  very  low  (=1 50-200).  The  tiny  eggs  (=  I  mm)  are 
demersal  and  spherical,  and  are  attached  in  clusters  to  hard 
surfaces  (logs,  stones,  etc.)  taking  up  to  23  days  to  hatch,  and 
the  larvae  drift  downstream  to  sea.  The  post  spawning  adults 
die.  The  life  cycle  is  essentially  annual.  Larvae  at  hatching  are 


152 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  79.     Young  of  Relropinna  retropmna.  35  mm  (above);  and  Aplochilon  sp.,  24  mm  (below). 


5-7  mm  long  with  little  yolk  anteriorly  below  pectoral  fins.  They 
are  very  elongate,  the  vent  posterior  at  more  than  75%  body 
length,  the  trunk  encompassed  by  a  low  finfold  from  head  around 
tail  to  yolk  sac.  Small  pedunculate  pectoral  fins  occur.  Pigmen- 
tation is  confined  to  the  eyes  and  a  narrow  band  in  mid-ventral 
between  head  and  vent.  Newly  hatched  larvae  disperse  to  sea 
and  are  not  fiirther  studied  (see  Blackburn,  1950). 

Aplochilon  taeniatus  is  recorded  spawning  in  ft-esh  water  dur- 
ing winter,  the  small  (1.5  mm),  spherical  eggs  being  demersal 
and  attached  to  firm  benthic  objects,  fecundity  2,500-3,000  and 
development  about  20  days.  The  larvae  are  very  elongate  and 


slender  with  a  yolk  sac  beneath  the  pectoral  fins.  The  vent  is  at 
about  75%  of  body  length.  A  finfold  encompasses  the  trunk  and 
tail.  Campos  (1969)  shows  a  single  large  melanophore  just  in 
front  of  the  vent.  His  figure  of  a  larva  presumably  8  mm  long 
(he  states  80  mm)  shows  a  series  of  melanophores  along  the 
abdomen  and  a  few  on  the  lower  caudal  peduncle.  Recent  col- 
lections of  larval  Aplochiton  from  Fiordo  Aisen  in  southern 
Chile  show  that  some  movement  to  sea  occurs.  At  a  length  of 
24  mm  the  late  larva  has  well  differentiated  rays  in  the  dorsal, 
caudal,  anal  and  pectoral  fins  and  distinct  pelvic  fin  buds  are 
evident  (Fig.  79).  An  adipose  fin  is  also  differentiated.  Pigmen- 


Fig.  80.     Young  of  Galaxias  maculatus.  14.5  mm  (above);  and  Prolotrocles  maraena  35  mm  (below). 


McDOWALL:  SALMONIFORMS 


153 


tation  is  sparse,  limited  to  spaced  melanophores  along  the  ab- 
domen. The  larva  remams  very  elongate,  the  vent  at  about  85% 
of  total  length.  Eigenmann  ( 1 928)  reported  that  A.  manmis  (=A. 
taenialus)  spawns  in  the  sea  but  this  has  never  been  corroborated 
(see  Campos.  1969). 

Galaxiidae.  —  Diadromous  species:  Spawning  is  usually  in  fresh- 
water. Eggs  of  Galaxiasfasciatus  are  deposited  in  autumn-win- 
ter on  stream-side  forest  debris  during  floods  and  develop  out 
of  water,  hatching  when  re-immersed  during  a  subsequent  flood. 
The  larvae  go  to  sea  on  hatching,  returning  in  spring  as  elongate, 
transparent  juveniles  about  45  mm  long.  A  minor  metamor- 
phosis involves  shrinkage  at  freshwater  entry.  The  eggs  are  of 
moderate  size  (~2  mm)  and  number  many  thousands;  devel- 
opment takes  about  30  days.  Most  other  diadromous  species 
have  unobserved  habits.  G.  maciilatus  spawns  in  tidal  estuaries 
where  streamside  vegetation  is  inundated  at  high  spring  tides 
and  development  takes  place  between  successive  series  of  spring 
tides.  Most  adults  die  after  spawning  and  larval  life  is  marine. 
The  eggs  are  simple,  spherical,  demersal  and  adhesive,  varying 
from  1-2  mm  diameter  and  more  or  less  colourless.  Benzie 
(1968a)  described  eggs  of  G.  maculatus  as  "finely  etched."  Lar- 
vae at  hatching  have  a  well  developed  yolk  sac.  with  a  single 
oil  globule,  the  sac  below  and  behind  the  pectoral  fins.  The  larvae 
are  slender  and  elongate  at  hatching,  7-8  mm  long,  and  have 
the  finfold  continuous  from  about  mid  dorsal  around  tail  to  yolk 
sac.  The  vent  is  posterior,  at  about  75%  of  total  length. 

Non-diadromous  species:  Most  species  in  the  family  are  non- 
diadromous  (31  of  37  species).  Those  known  spawn  on  sub- 
strates near  adult  habitats  and  the  pelagic  "whitebait"  juvenile 
stage  is  omitted.  Eggs  are  laid  in  aggregations  (G.  vulgaris).  Lar- 
vae on  hatching,  where  described,  resemble  those  of  G.  mac- 
ulatus. 

Galaxiella  pusilla  is  distinctive  in  being  sexually  dimorphic, 
spawning  in  pairs,  the  females  laying  eggs  individually  on  stream 
vegetation.  Individual  placement  of  eggs  is  also  reported  for 


Brachygalaxias  bullocki.  The  ability  to  aestivate  is  recorded  for 
some  species  (Neochanna.  New  Zealand)  and  spawning  follows 
restoration  of  water.  It  is  suspected  in  others  (Galaxiella.  Aus- 
tralia; Brachygalaxias.  Chile)  and  may  involve  drought  survival 
of  eggs  (see  Benzie,  1968a,  b;  Backhouse  and  Vanner,  1978; 
Cadwallader,  1976;  Campos,  1972;  McDowall  1968b,  1978; 
McDowall  et  al.,  1975;  Mitchell  and  Penlington,  1982). 

Little  IS  known  about  the  marine  larval/juvenile  life  of  any 
of  these  southern  salmoniforms.  Small  numbers  of  Galaxias 
larvae  (Fig.  80)  have  been  collected  at  sea  (McDowall  et  al., 
1975),  as  have  a  few,  usually  pre-migratory  Retropinna.  The 
presence  of  a  pelagic-living,  transparent,  elongate,  migratory 
juvenile  seems  to  be  common  to  most  species  that  are  marine 
or  lacustrine  at  some  slage— Galaxias.  Retropinna.  Prototroctes. 
Aplochiton.  This  is  likely  to  have  little  phylogenetic  significance 
but  to  relate  more  to  their  pelagic,  oceanic  habits.  These  small 
fish  resemble  many  other  unrelated  fish  with  pelagic  juveniles. 
The  marine,  pelagic  phase  is  followed  in  all  instances  by  a  minor 
metamorphosis  on  entry  to  fresh  water.  Principally  this  involves 
rapid  assumption  of  pigmentation  and  in  some  species  a  distinct 
change  in  body  form.  Shrinkage  is  recorded  in  a  few  species. 

Identification  of  oceanic  larvae  and  juveniles  to  family  is 
assisted  by  dorsal  fin  position  and  the  early  development  of  an 
adipose  fin  in  all  but  galaxiids.  The  elongate  form  with  the  vent 
at  about  75%  of  total  length  is  helpful.  Differences  have  been 
recorded  in  pigment  patterns  between  some  of  the  diadromous 
galaxiid  juveniles  although  insufficiently  to  use  as  diagnostic 
differences  (McDowall  and  Eldon.  1980).  Meristic  differences 
between  species  are  of  little  value  for  specific  identification  ow- 
ing to  their  wide  ranges  and  latitudinal  variability.  Identification 
remains  a  difficulty  and  improvement  will  depend  on  the  capture 
and  examination  of  additional  material. 

Fisheries  Research  Division,  Ministry  of  Agriculture  and 
Fisheries,  Christchurch,  New  Zealand. 


Osmeridae:  Development  and  Relationships 
M.  E.  Hearne 


OSMERIDAE,  the  true  smelts,  are  a  small  family  of  northern 
hemisphere  salmoniform  fishes.  The  family  includes  2 
subfamilies,  6  genera,  10  species,  and  13  forms  (monotypic  and 
subspecies).  They  have  marine,  anadromous  or  landlocked  and 
freshwater  life  histories  in  the  Pacific,  Arctic  and  Atlantic  oceans 
and  their  drainages  (McAllister,  1963).  These  silvery  tasty  little 
fishes  are  captured  by  both  recreational  and  commercial  pur- 
suits along  the  open  coast  beaches  and  rivers  during  their  spawn- 
ing runs. 

Development 
The  smelts  are  highly  selective  spawners,  choosing  to  spawn 
on  very  specific  sub-tidal  areas,  beaches  and  rivers.  Some  species 


spawn  in  the  daytime,  and  some  spawn  at  night.  The  eggs  of 
osmerids  possess  an  adhesive  membrane  that  attaches  to  sand 
grains  and  plant  material.  This  anchor  membrane  results  from 
the  ruptunng  of  an  outer  "chorion"  during  spawning,  which 
turns  out  and  onto  the  substrate.  This  adaptation  for  demersal 
spawning  is  observed  in  all  10  species  of  osmerids  (Hamanda, 
1961;  Thompson  et  al.,  1936;  Morris,  1951;  McAllister,  1963; 
Simonsen,  1978;  DeLacy  and  Batts,  1963;  Hearne,  1983). 

The  first  description  of  smelt  development  was  made  by  Eh- 
renbaum  (1894)  for  the  Elbe  River  smelt,  Osmerus  eperlans 
illustrating  embryological  stages,  yolk-sac  larva,  transforming 
larva,  and  the  juvenile.  Up  to  now,  the  yolk-sac  stage  of  many 
of  these  species  has  been  at  least  illustrated  or  photographed. 


154 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  81.  (A)  Yolk-sac  larvae  of  Spirinchus  starksi.  Osmeridae,  7.4  mm,  from  Morris  (1951);  (B)  Yolk-sac  larvae  of  Plecoglossus  altivelis. 
Plecoglossidae,  ca.  6.0  mm,  from  Okada  (1960);  (C)  Post  yolk-sac  larva  of  Salangichlhys  inicrodon  (Salangidae),  ca.  7.0  mm,  drawn  from  two 
specimens  in  CAS  504 1 5. 


Ahlstrom  (pers.  comm.)  deteimined  that,  in  general,  osmerid 
larvae  were  unique  from  other  elongate  larvae  in  the  California 
Current  system  by  having  a  single  mid-ventral  row  of  mela- 
nophores  below  the  gut.  Based  on  all  the  available  larval  de- 
scriptions for  osmerids,  including  the  Atlantic  forms,  this  single 
row  of  melanophores  appears  to  be  a  hallmark  of  the  family. 
Listed  in  Table  37  are  sources  of  larval  and  juvenile  descriptions 
for  the  ten  species  of  smelts. 

These  descriptions  use  various  characteristics  for  each  species 
and  are  not  comparative  in  design.  Melanophore  counts  are 
referred  to  by  Yapchionges  (1949),  Follett  (1952),  Simonsen 
(1978),  Morris  (1951),  Dryfoos  (1965)  and  Moulton  (1970). 
Myomere  counts  were  used  by  Delacy  and  Batts  ( 1 963).  Cooper 
(1978)  and  Morris  (1951)  used  both  myomere  and  melanophore 
counts. 

Larval  osmerids  have  the  following  external  features  in  com- 
mon: elongate  body  shape;  gut  about  75%  body  length;  mouth 
sub-terminal;  head  dorso-ventrally  flattened;  lower  jaw  not  well- 
developed  in  early  larvae;  conspicuous  choroid  fissure  in  ventral 
third  of  eye  with  ventral  rim  of  clear  choroid  tissue;  stalked 
pectorals,  stalk  becoming  more  pronounced  in  late  larvae;  yolk 
sac  positioned  6-12  myomeres  posterior  to  the  pectoral  base; 
finfold  extending  from  midbrain  area  to  tail,  from  mid-yolk  sac 
to  anus,  and  from  anus  to  tail;  no  dorsal  melanophores;  scattered 
melanophores  (20-50)  on  ventral  half  of  yolk  sac;  0-2  mela- 
nophores on  posterior  ventral  half  of  yolk  sac;  single  row  of 
melanophores  along  ventral  midline  of  gut,  sometimes  extend- 
ing into  finfold;  1-3  melanophores  on  dorsal  surface  of  gut  at 
the  anal  bend;  single  row  of  melanophores  on  ventral  midline 
of  tail;  conspicuous  opaque  liver  ventral  to  foregut  (Ehrenbaum, 
1894;  Yapchionges,  1949;  Morris,  1 95  l;DeLacy  and  Batts  1963; 


Dryfoos,  1965;  Eldridge,  1970;  Blackburn,  1973;  Cooper,  1978; 
Heame,  1983). 

A  comparative  study  of  four  of  the  species  oflTOregon  (Heame, 
1983)  used  ventral  melanophore  counts  and  myomere  counts 
in  an  attempt  to  characterize  the  larvae  of  these  species.  Ten- 
dencies in  these  counts  showed  Hypomesus  pretiosus  and  Spi- 
rinchus starksi  to  have  high  ventral  melanophore  counts  while 
Spirinchus  thaleichthys  and  Thaleichthys  pacificus  have  lower 
melanophore  counts.  Myomere  counts  showed  tendencies  that 
further  separated  each  similarly  pigmented  pair. 


Table  37,    Sources  of  Larval  and  Juvenile  Descriptions  of  Smelts,  (x 
no  description  found.) 

Taxon  Larvae  Juveniles 


Hypomesus 

pretiosus 
Hypomesus 

transpacificus 
Spirmchus 

lanceolatus 
Spirinchus 

starksi 
Spirinchus 

thaleichthys 
Thaleichthys 

pacificus 
Alios  merus 

elongatus 
Mallotus  villosus 
Osmerus  mordax 
Osmerus  eperlanus 


Yapchionges,  1949 

X 

Hikita,  1958 
Morris,  1951 

Dryfoos,  1965; 

Moulton,  1970 
DeLacy  and  Batts, 

1963 

X 

Schmidt,  1906c 
Cooper,  1978 
Ehrenbaum,  1894 


Follett,  1952 

Simonsen,  1978 

Hikita,  1958 

Heame,  1983 

Simonsen,  1978 

Baraclough,  1964 

Heame,  1983 

Templeman,  1948 
Cooper,  1978 
Ehrenbaum,  1894 


HEARNE:  OSMERIDAE 


155 


The  transformational  stages  of  osmerids  are  not  fully  known, 
since  complete  developmental  series  have  not  been  reported  for 
all  of  the  species.  However,  it  is  apparent  from  rearing  studies 
(Morris,  1951;  Cooper,  1978)  that  caudal  flexion  occurs  after 
yolk  absorption  and  along  with  median  fin  formation.  The  pelvic 
fins  arise  from  the  ventral  body  musculature  as  prominent  buds 
after  the  median  fin  rays  have  formed,  and  appear  stalked,  be- 
coming inserted  as  the  ventral  musculature  joins  ventrally.  The 
pectoral  fins  are  present  at  hatching  and  remain  pedunculate 
until  postflexion  stages  acquire  adult-like  pigmentation. 

During  flexion  an  additional  series  of  melanophores  forms 
along  the  ventro-lateral  edge  of  the  body  musculature  and  ap- 
pears as  a  double  row  of  spots  from  ventral  view.  There  are  also 
count  differences  between  the  species  in  these  secondary  me- 
lanophores (Heame,  1983),  and  they  may  aid  in  identification 
of  flexion  and  postflexion  stages. 

The  postflexion  stages  of  two  species  of  osmerids  have  been 
erroneously  described  as  new  species  belonging  to  other  families 
by  Chapman  ( 1 939).  Hubbs  (1951)  has  shown  that  one  of  these 
smelts,  placed  in  the  family  Paralepididae  as  Lestidium  parn. 
is  actually  a  late  postflexion  stage  of  Thaleichthys  pacificus.  and 
the  other  one,  placed  in  the  family  Sudidae  as  Sudis  squamosa, 
is  a  postflexion  Mallotiis  villosus.  The  blackened  gut  cavities  of 
the  postflexion  stages  of  these  two  species,  lend  a  distinct  re- 
semblance to  the  midwater-inhabiting  sudids  and  paralepidids, 
and  also  suggest  a  unique  departure  from  the  developmental 
trend  of  the  other  species  that  may  warrant  the  use  of  the  term 
"pre-juvenile"  as  defined  by  Hubbs  (1943). 

Relationships 

In  a  recent  statement  on  classification,  Rosen  ( 1974)  proposed 
an  infraorder  Salmonae  to  include  two  suborders,  the  Argen- 
tinoidei  and  Salmonoidei,  the  Osmeridae  being  placed  in  the 
latter  under  the  superfamily  Osmeroidea  (with  the  Plecoglos- 
sidae,  Retropinnidae,  and  Salangidae).  On  the  basis  of  embry- 
ological  and  larval  features,  Soin  (1980)  characterized  different 
types  of  salmoniform  fishes.  He  placed  the  Piecoglossidae  and 
Osmeridae  in  the  same  category  based  on  similar  egg  mor- 
phology (presence  of  an  anchor  membrane),  degree  of  devel- 
opment at  time  of  hatching  and  at  time  of  yolk  absorption.  In 
a  study  of  stomiiform  fishes  using  adult  characters.  Fink  and 


Weitzman  (1982)  placed  the  families  Osmeridae,  Salangidae, 
Piecoglossidae,  Retropinnidae,  and  Galaxiidae  all  together  as 
"unresolved  sister  taxa." 

The  larvae  of  osmehds  (Spin nchus  slarksi.  Fig.  8 1  A)  are  strik- 
ingly similar  to  larval  plecoglossids  (Plecoglossus  alttvelis.  Fig. 
8 1 B).  The  yolk  sac  of  these  two  families  is  positioned  such  that 
its  posterior  edge  is  near  myomere  11-12.  The  plecoglossids 
also  have  a  single  median  ventral  row  of  melanophores  and,  as 
development  proceeds,  another  latero-ventral  row  of  spots  ap- 
pears along  the  ventral  edge  of  the  body  musculature,  just  as  in 
osmerid  development. 

Photographs  of  the  yolk-sac  stage  of  Salangichthys  microdon, 
Salangidae,  (Okada,  1960:  pi.  17)  show  that  the  yolk-sac  mor- 
phology is  different  than  in  the  Osmeridae  and  Piecoglossidae. 
The  yolk  sac  of  Salangtchthys  microdon  is  co-extensive  with  the 
undersurface  of  the  gut  and  is  more  oblong  shaped  (pyriform) 
than  the  more  rounded,  anteriorly  placed  yolk  sac  of  the  os- 
merids and  plecoglossids.  The  post  yolk-sac  larvae  of  salangids 
(Fig.  81C)  are  nearly  identical  to  those  of  osmerids  and  pleco- 
glossids, exhibiting  the  single  median  ventral  row  of  melano- 
phores. Also,  the  eggs  of  salangids  are  different  than  the  osmerid- 
plecoglossid  type  by  having,  instead  of  an  anchor  membrane, 
an  anchoring  structure  that  is  composed  of  various  kinds  of 
filaments  that  turn  out  and  onto  the  substrate  (Wakiya  and 
Takahashi,  1913).  Larval  development  is  not  yet  documented 
for  the  Sundasalangidae,  however  adults  of  this  minute  family 
of  salangoid  fishes  have  ventral  pigment  patterns  (Roberts  1981: 
fig.  1)  that  are  strikingly  similar  to  the  postflexion  pigment  pat- 
terns of  osmerids.  The  same  ventral  pigment  patterns  (single 
ventral  midline,  paired  latero-ventral  melanophores)  can  also 
be  seen  in  adults  of  Salangidae  (Okada,  1960). 

One  interpretation  may  be  that  the  similarities  in  ventral 
pigment  patterns  and  egg  morphology  may  be  the  retention  of 
a  trait  of  an  ancestor  common  to  the  Osmeridae,  Piecoglossidae, 
and  Salangidae,  and  give  support  to  theories  arising  from  sys- 
tematic observations  of  adult  salmonoids  that  these  families  are 
closely  related  to  each  other  and  not  to  the  other  salmoniform 
families. 

184  Day  Street,  San  Francisco,  California  94131. 


Argentinoidei:  Development  and  Relationships 


E.  H.  Ahlstrom,  H.  G.  Moser  and  D.  M.  Cohen 


THE  argentinoid  fishes  as  here  discussed  have  been  consid- 
ered a  suborder  by  Cohen  ( 1 964b)  and  many  other  authors 
and  a  super-family  of  an  expanded  suborder  that  also  includes 
the  alepocephaloids  by  Greenwood  and  Rosen  (1971).  The  latter 
group  is  not  treated  at  length  in  this  book,  because  little  infor- 
mation on  alepocephaloid  ELH  stages  has  appeared  since  Beebe's 
(1933a)  survey  which  showed  they  hatch  from  large  eggs  and 
have  direct  development.  The  argentinoids  sensu  strictu  appear 
to  be  monophyletic  on  the  basis  of  four  derived  characters.  One 
character  concerns  the  development  of  rays  in  the  finfold  of  the 


larva  and  is  described  later  in  this  paper.  A  second  character  is 
the  development  of  pustules  on  the  inner  surface  of  the  chorion 
(not  known  for  opisthoproctids).  A  third  character  relates  to  the 
swimbladder,  which,  when  present,  is  served  by  a  unique  kind 
of  rete  mirabile,  first  described  by  Fange  (1958)  and  further 
investigated  by  Marshall  (1960)  who  named  these  structures 
micro-retia  mirabilia.  A  fourth  unique  character,  and  one  which 
never  has  been  adequately  studied  and  documented,  is  the  ten- 
dency in  the  group  for  the  vomer  and  palatines  to  assume  the 
functions  of  the  premaxillary  and  maxillary. 


156 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  38.    Literature  References  for  Ontogenetic  Stages  of  Argentinoids. 


Species 


Egg 


Transformation  stage 


Argentinidae 
Argentina  elongata 
A.  silus 

A.  sphyraena 
Glossanodon  leioglossus 
G.  polli 

G.  semifascialus 
Microstoma  microstoma 

Nansenia  groenlandica 
N.  oblita 
Xenopthahnichthys  danae 

Bathylagidae 
Balhylagus  antarcticus 

B.  euryops 

B.  longirostris 


B.  nigrigenys 

B.  ochotensis 
B.  schmidti 
B.  stilbius 
B.  wesethi 

Opisthoproctidae 

Balhylychnops  exitis 
Dolichopteryx  spp. 
Dolichopteryx  longipes 
Macropinna  microstoma 
Opisthoproctus  grimaldii 
Rhynchohyalus  natalensis 
H'interia  telescopa 


Robertson,  1975a 
Schmidt,  1906c 
Sanzo,  193  Id 
Sanzo,  193  Id 

Nishimura,  1966 
Sanzo,  193  Id 


Sanzo,  193  Id 


Yefremenko,  1982 


Pertseva-Ostroumova 
and  Rass,  1973 

Ahlstrom,  1969 
Ahlstrom,  1969 
Ahlstrom,  1969 


Holt,  1898;  Schmidt,  1906c 
Schmidt,  1906c,  Sanzo,  193 Id 
Schmidt,  1918,  Sanzo.  193 Id 

Nishimura,  1966 

Lo  Bianco,  1903;  Schmidt,  19 If 

Sanzo,  193  Id 
Schmidt,  1918 
Schmidt,  1918;  Sanzo,  193 Id 


Yefremenko,  1979a,  1983 

Brauer,  1906;  TSmng,  1931 

Ehrenbaum,  1905-09;  Murray 
and  Hjort.  1912;  Roule  and 
Angel,  1930;  Beebe,  1933b 

Pertseva-Ostroumova  and  Rass, 
1973 

Ahlstrom,  1972b 

Dunn,  1983a 

Ahlstrom,  1965,  1972b 

Ahlstrom,  1965,  1972b 


Roule  and  Angel,  1930 
Beebe,  1933a 
Chapman,  1939 
Schmidt,  1918 
Bertelsen  et  al.,  1965 
Belyanina,  1982b 


Schmidt.  1906 
Sanzo,  1931d 
Poll,  1953 
Nishimura,  1966 
Schmidt,  1918 

Schmidt,  1918 
Schmidt,  1918 
Bertelsen,  1958 


TSning,  1931 
Beebe,  1933b 


Ahlstrom,  1972b 
Dunn,  1983a 


Cohen,  1960 


Although  now  there  seems  to  be  general  agreement  as  to  the 
genera  to  be  included  in  the  group,  their  internal  arrangement 
is  an  unsettled  matter.  Opinions  range  from  those  of  C.  L.  Hubbs 
(1953),  who  relegated  all  to  a  single  family,  to  those  of  Chapman 
(1948  and  papers  cited  therein),  who  advocated  eight  different 
families.  Subsequently  Cohen  (1964b)  classified  the  group  in 
three  families  using  inadequately  evaluated  characters. 

Family  Argentinidae  (most  genera  are  probably  worldwide): 
Subfamily  Argentininae  (benthopelagic.  outer  shelf  to  slope): 

Argentina  (12  species)  and 

Glossanodon  (seven  or  more  species). 
Subfamily  Microstomatinae  (mesopelagic)': 

Microstoma  (one  or  two  species), 

Nansenia  ( 1 3  species)  and 

Xenophthalmichthys  (one  or  two  species). 
Family  Bathylagidae  (meso-to  bathypelagic): 

Bathylagns  (including  Leiiroglossiis  and  Therobromus;  about 

a  dozen  to  1 5  species;  several  species  in  the  Arctic  and 

Antarctic). 
Family  Opisthoproctidae  (mesopelagic): 

Group  1: 

Macropinna  (one  species;  restricted  to  N.  Pacific  and  east- 
em  S.  Pacific), 


'  Herein  considered  a  distinct  family. 


Opisthoproctus  (two  species), 

Rhynchohyalus  (one  species;  Atlantic  and  Indian  Oceans) 
and  H'interia  (one  species). 

Group  II: 

Balhylychnops  (one  or  more  species),  and 

Dolichopteryx  (perhaps  half  a  dozen  species). 
An  alternate  arrangement  presented  by  Greenwood  and  Ro- 
sen (1971)  and  essentially  based  on  inadequately  evaluated  char- 
acters in  the  branchial  arches  and  caudal  fin  skeleton  proposed 
two  families  within  a  superfamily  Argentinoidea:  Family  Ar- 
gentinidae and  Family  Bathylagidae  with  Subfamily  Bathyla- 
ginae  (including  Microstomatidae)  and  Subfamily  Opistho- 
proctinae. 

Unanswered  questions  concerning  the  systematics  of  the  group 
are  numerous  and  exist  at  all  levels.  Following  is  a  summary. 
( 1 )  What  are  the  external  relationships  of  the  argentinoids?  (2) 
How  many  distinct  lineages  exist  within  the  group,  how  should 
they  best  be  arranged  with  respect  to  each  other,  and  how  many 
families  should  be  recognized?  (3)  Do  Argentina  and  Glossan- 
odon constitute  a  monophyletic  group?  If  not,  where  does  each 
belong?  (4)  How  many  genera  should  be  recognized  among  the 
bathylagids?  (5)  Within  the  opisthoproct  group  do  the  elongate 
species  in  the  Bathylychnops-Dolichopieryx group  and  the  short- 
bodied  species  in  the  Opisthoproctus  group  constitute  mono- 
phyletic lineages  and  if  so  should  they  be  named?  (6)  Since 
species  complements  of  genera  are  inadequately  known,  espe- 


AHLSTROM  ET  AL.:  ARGENTINOIDEI 


157 


Table  39.    Characters  of  the  Eggs  of  Argentinoidei. 


Number  of 

Distribution  of 

Diameter  of 

Species 

Diameter 

oil  globules 

oil  globules 

oil  globules 

Source 

Argentina  stalls 

1.31-1.66 

vegetal  pole 

0.27-0.46 

Original 

Argentina  siliis 

3.0-3.5 

vegetal  pole 

0.95-1.16 

Schmidt,  1906c 

Argentina  sphyraena 

(Mediterranean) 

1.60-1.68 

vegetal  pole 

0.44 

Sanzo,  193  Id 

(North  Sea) 

1.70-1.85 

vegetal  pole 

0.37-0.47 

Schmidt,  1906c 

Argentina  elongata 

1.67-1.80 

vegetal  pole 

0.35-0.45 

Robertson,  1975a 

Glossanodon  leioglossus 

1.44-1.52 

vegetal  pole 

0.36 

Sanzo,  193  Id 

Glossanodon  semifasciatus 

1.5-1.6 

vegetal  pole 

0.36 

Nishimura,  1966 

Microstoma  microstoma 

(Atlantic) 

1.60-1.72 

vegetal  pole 

0.48-0.52 

Sanzo,  193  Id 

(Pacific) 

2.05-2.38 

vegetal  pole 

0.49-0.82 

Original 

Nansenia  Candida 

1.39-1.56 

vegetal  pole 

0.41-0.49 

Original 

Nansenia  crassa 

1.05-1.30 

vegetal  pole 

0.30-0.35 

Original 

Nansenia  ohlita 

1.39-1.56 

vegetal  pole 

0.40-0.53 

Sanzo,  193  Id 

Bathylagiis  antarclicus 

1.8-2.2 

3-8 

* 

0.2-0.3 

Yefremenko,  1982 

Bathylagus  schmidti 

1.65-1.90 

up  to  9 

* 

Ahlstrom,  1969 

Bathylagiis  slilhiiis 

1.01-1.21 

15-25 

*■ 

Ahlstrom,  1969 

Bathylagus  urotranus 

1.03-1.21 

15-25 

* 

Pertseva-Ostroumova  and  Rass, 
1973,  and  original 

Bathylagus  ochotensis 

0.92-1.1 

many  to 
two  clumps 

** 

Original 

Balhylagits  wesclhi 

0.90-1.10 

12-20 

** 

Ahlstrom,  1969 

Bathylagus  nigrigenys 

0.83-1.09 

12-20 

** 

Pertseva-Ostroumova  and  Rass, 
1973,  and  original 

'  First  grouped  at  vegetal  pole,  then  move  to  beneath  embryo,  then  coalesce  to  one  at  each  equatonal  pole. 
•  Numerous  globules  at  vegetal  pole  then  coalesce  to  one  clump  at  each  equatonal  pole. 


cially  the  mesopelagic  ones,  do  presently  available  early  life 
history  specimens  help  define  the  species  composition  of  argen- 
tinoid  genera? 

Development 

Eggs  are  known  for  1 3  species  of  argentinoids  and  larvae  for 
22  species  (Table  38).  We  present  in  this  paper  eggs  of  5  ad- 
ditional argentinoid  species  and  larvae  of  8  additional  species. 
These  are:  eggs  and  larvae  oi  Argentina  sialis.  Microstoma  sp., 
Nansenia  Candida  and  N.  crassa;  larvae  only  for  Bathylagus 
argyrogaster.  B.  bencoides.  B.  pacificus  and  Balhylychnops  ex- 
ilis:  eggs  only  for  Bathylagus  ochotensis. 

Eggs 

The  eggs  of  argentinoids  are  pelagic,  round,  have  a  moderate 
to  narrow  perivitelline  space,  segmented  yolk  and  a  chorion 
with  distinctive  pustules  on  the  inner  surface  (Table  39,  Fig. 
82).  Egg  diameters  and  oil  globule  characters  are  given  in  Table 
39. 

Argentinoid  larvae  hatch  as  relatively  undifferentiated  yolk- 
sac  larvae,  regardless  of  egg  size.  That  is,  yolk-sac  larvae  of  A. 
silus  at  7.5  mm,  newly  hatched  from  eggs  3.0-3.5  mm  diameter, 
are  at  about  the  same  stage  of  development  as  3  mm  bathylagid 
yolk-sac  larvae  which  hatch  from  1  mm  eggs.  In  most  marine 
fishes  larger  eggs  produce  more  highly  differentiated  hatchlings. 

Larvae 

Body  form.  —  Argentinid  and  bathylagid  larvae  are  slender,  those 
of  microstomatids  are  deeper-bodied,  and  opisthoproctids  have 
a  wide  variety  of  body  shapes  ranging  from  the  slender  larvae 
of  Balhylychnops  to  the  deep-bodied  Opisthoproctus  (Table  40, 
Figs.  83-87). 


The  gut  is  elongate  and  straight  in  argentinids  and  bathylagids, 
with  the  exception  of  B.  milleri  where  the  gut  is  straight  but 
only  about  half  the  body  length.  In  argentinids  the  gut  is  lined 
with  transverse  rugae  for  almost  the  entire  length.  In  most  bath- 
ylagids the  gut  has  two  distinct  sections:  an  anterior  section  with 
longitudinal  internal  ridges,  separated  by  a  valve  from  a  shorter 
posterior  section  with  transverse  rugae.  The  anterior  section  in 
B.  hericoides  and  B.  longirosths  is  markedly  smaller  in  diameter 
compared  with  other  species.  Larvae  of  j5.  wesethi.  B.  nigrigenys 
and  B.  argyrogasterhave  transverse  rugae  along  the  entire  length 
of  the  gut  and  the  anterior  section  is  relatively  larger  in  diameter 
and  thin-walled.  Also  the  posterior  section  is  subdivided  by  a 
second  valve.  B.  ochotensis  larvae  develop  a  similar  structure. 

The  gut  in  microstomatid  larvae  is  long,  but  anteriorly  has 
an  elongate  S-shaped  fold  that  lies  flat  on  the  left  side  (Fig.  84). 
The  lumen  of  the  anterior  folded  section  is  characterized  by 
longitudinal  ridges  whereas  the  posterior  straight  section  has 
transverse  rugae.  The  short  pyloric  section  has  longitudinal  ridges. 
Schmidt  (1918)  shows  the  gut  extended  beyond  the  finfold  mar- 
gin in  Nansenia  ohlita  and  trailing  in  early  stage  Microstoma 
microstoma  larvae  but  we  have  not  seen  this  in  any  specimens 
of  these  genera. 

In  opisthoproctids  the  gut  is  elongate  in  Balhylychnops  and 
Dolichoptery.x  and  relatively  shorter  in  the  deeper-bodied  gen- 
era, Macropinna.  Rhyncholyalus  and  Opisthoproctus.  In  all  gen- 
era there  is  a  sac-like  stomach,  which  exits  through  a  constricted 
pyloric  section  to  the  intestine.  In  Balhylychnops  and  Doll- 
chopteryx  the  sac  is  elongate  and  pointed  at  its  tip  whereas  in 
the  other  genera  it  is  more  rounded  in  form.  The  sac  lies  on  the 
left  side,  except  in  Balhylychnops  where  it  lies  on  the  right.  In 
the  latter  genus  the  pyloric  constriction  leads  into  a  short  but 
prominent  bulbous  section.  DoHchopteryx  is  similar  but  lacks 


158 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig,  82.  Eggs  of  argentinoids.  (A)  Argentina  sialus.  1.5  mm,  CalCOFI  5103,  Sta.  1 17.35;  (B)  Microstoma  sp.,  2.2  mm,  CalCOFI  751 1,  Sta. 
87.90;  (C)  Nansenia  Candida.  1.4  mm,  CalCOFI  Sta.  60.90;  (D)  N.  crassa.  1.5  mm,  CalCOFI;  (E)  Bathylagus  stilhius.  1.1  mm,  from  Ahlstrom 
(1969);  (F)  B.  schmidli.  1.8  mm.  from  Ahlstrom  (1969);  (G)  B.  ochotensis.  1.1  mm,  CalCOFI  5002  Sta.  60.90;  (H)  B.  weselhi.  1.0  mm,  Ahlstrom 
(1969);  (I)  B.  nigrigenys.  0.96  mm.  CalCOFI  5106  Sta.  157.20. 


the  post-pyloric  bulb.  In  Macropinna  and  Opisthoproctus  there 
is  a  straight  section  leading  posteriorly  from  the  pylorus,  which 
ends  in  an  S-shaped  fold  and  an  enlarged  rectal  bulb,  the  latter 
described  by  Bertelsen  and  Munk  (1964).  The  anterior  section 
including  the  sac  and  pyloiojs  have  longitudinal  internal  ridges 
while  sections  posterior  to  this  have  transverse  rugae.  In  late 
larval  stages  the  entire  section  posterior  to  the  pylorus  becomes 
part  of  the  S-shaped  coil. 

The  head  is  relatively  small  in  argentinids  and  has  a  rounded 
blunted  anterior  profile  (Fig.  83,  Table  40).  It  is  slightly  larger 


in  most  microstomatids,  with  the  exception  oi  Microstoma  sp. 
(Pacific  form)  which  has  a  small  head.  In  most  microstomatids 
the  head  has  a  rounded,  blunted  anterior  profile  and  is  bent 
slightly  downward  from  the  longitudinal  axis.  In  both  families 
the  eye  is  either  round  or  slightly  ellipsoidal.  In  bathylagids  the 
head  is  moderate  in  size  but  highly  various  in  shape  (Figs.  85, 
86;  Table  40).  The  snout  is  generally  longer  than  in  Argentinidae 
and  Microstomatidae. 

Eye  shape  and  structure  vary  greatly  within  the  bathylagids. 
Bathylagus  milleri  has  a  large,  nearly  round  eye  in  contrast  to 


AHLSTROM  ET  AL.:  ARGENTINOIDEI 


159 


v^^&^  t&.        **J      isf      *tii        "jv       tn^        'zr 
•  •^^^I'''         ."»•       *.        *S        ^^       S        -S^* 

»    ^i^^-^j***  «  #»♦  "^  »?  PIt^i 


Fig.  83.  Larvae  of  Argentinidae.  (A)  Argentina  stalls.  7.0  mm.  CalCOFI  5103  Sta.  1  17.35;  (B)  A.  stalls.  9.0  mm,  CalCOH  5104  Sta.  97.40; 
(C)  .-1.  stalls.  17.5  mm,  CalCOFI  5103  Sta.  120.35;  (D)  A.  stalls.  21.0  mm,  CalCOFI  5105  Sta.  123.40;  (E)  A.  silus.  32.5  mm.  redrawn  from 
Schmidt  (1906c);  (F)  A.  sphyraena.  19.2  mm,  ibid;  (G)  Glossanodon  semifasciatus,  12.5  mm,  from  Nishimura  (1966). 


AHLSTROM  ET  AL.:  ARGENTINOIDEI 


161 


Table  40.    Comparative  Morphometry  of  Aroentinoid  Larvae.  Mean  values  (%)  of  body  proportions  for  three  ontogenetic  stages  (preflexion- 

flexion-postflexion)  are  listed. 


Snoul-anus 

Eye  stalk 

Snout-anal 

Snout-dorsal 

Snout-pelvic 

distance 

Head  length 

Head  width 

Eye  length 

length 

Body  depth 

fin  distance 

fin  distance 

fin  distance 

Species 

Body  length 

Body  length 

Head  length 

Head  length 

Head  length 

Body  length 

Body  length 

Body  length 

Body  length 

A  rgentma  sialis 

76-78-84 

17-21-22 

54-44-41 

28-24-24 

— 

9-10-10 

0-78-81 

0-46-47 

0-0-49 

Microstoma  microstoma 

?-?-80 

7-7-23 

7-7-45 

7-7-27 

— 

7-7-13 

7-7-80 

7-7-68 

0-7-64 

Microstoma  sp.  (Pacific) 

76-79-80 

17-19-19 

53-49-44 

31-29-27 

_ 

8-10-10 

0-78-81 

0-70-72 

0-64-67 

Nansenia  Candida 

74-77-82 

21-25-26 

60-50-44 

36-28-28 

— 

12-14-16 

0-75-82 

0-54-58 

0-56-61 

Nansenia  crassa 

74-78-80 

22-25-28 

58-50-44 

36-29-24 

_ 

10-12-15 

0-76-80 

0-52-57 

0-56-60 

Nansenia  groenlandica 

7-78-80 

7-27-25 

7-50-42 

7-21-23 

— 

7-15-15 

7-77-80 

7-52-52 

7-54-57 

Xenophthatmichthys  danae 

?-?-82 

7-7-24 

7-7-48 

7-7-21 

— 

7-7-12 

7-7-86 

7-7-74 

''-7-52 

Bathylagus  milleri 

59-57-61 

20-19-26 

56-54-52 

31-27-26 

_ 

9-9-15 

0-0-71 

0-0-50 

0-0-45 

Bathylagus  schmidli 

72-76-78 

16-19-22 

50-52-46 

39-26-25 

.04-0-0 

7-8-10 

0-0-79 

0-0-57 

0-0-55 

Bathylagus  slilbius 

74-77-80 

20-22-24 

54-53-47 

32-25-20 

.03-0-0 

8-10-13 

0-0-79 

0-0-57 

0-0-55 

Bathylagus  urotranus 

78-82-81 

20-24-28 

56-53-46 

27-18-21 

.03-0-0 

10-10-12 

0-0-81 

0-0-61 

0-0-59 

Bathylagtis  pacificus 

76-85-81 

22-24-25 

39-42-44 

29-22-18 

28-29-20 

8-10-13 

0-81-80 

0-49-48 

0-51-51 

Bathylagus  curyops 

78-80-82 

18-20-20 

46-50-50 

31-26-25 

10-7-3 

10-11-12 

0-78-80 

0-45-48 

0-0-47 

Bathylagus  bericoides 

84-85-89 

25-26-26 

34-38-36 

27-25-22 

60-64-36 

8-8-9 

0-83-88 

0-0-52 

0-0-53 

Bathylagus  longiroslris 

85-88-92 

26-27-25 

34-34-34 

24-20-19 

54-48-27 

8-10-10 

0-88-90 

0-0-53 

0-0-57 

Bathylagus  ochotensis 

81-85-90 

20-23-23 

44-44-44 

32-21-21 

17-15-15 

8-10-11 

0-83-87 

0-53-54 

0-56-56 

Bathylagus  wesethi 

79-89-94 

13-26-27 

59-53-50 

27-16-13 

— 

9-14-16 

0-85-90 

0-58-60 

0-57-59 

Bathylagus  nigrigenys 

80-86-93 

20-29-28 

78-60-53 

30-18-14 

— 

12-16-18 

0-86-90 

0-57-60 

0-0-60 

Bathylychnops  exilis 

7-80-82 

7-21-22 

7-42-38 

7-22-18 

— 

7-8-7 

7-82-84 

7-71-73 

7-66-67 

Dolichopleryx  longipes 

7-74-75 

7-24-26 

7-44-34 

7-22-16 

— 

7-8-10 

7-0-77 

7-0-71 

7-62-62 

Macropinna  microstoma 

7-64-59 

7-26-35 

7-52-47 

7-22-21 

— 

7-15-21 

7-0-70 

7-0-66 

7-43-48 

Opisthoproctus  soleatus 

7-7-80 

7-7-37 

7-7-46 

7-7-18 

- 

7-7-18 

7-7-83 

7-7-63 

7-7-40 

other  species  which  have  relatively  smaller,  more  elliptical  eyes. 
Eyes  are  sessile  in  B.  milleri  and  in  the  B.  wesetht  group  but  are 
stalked  to  some  degree  in  all  other  species  known.  In  B.  slilbius 
and  relatives  (B.  urotranus,  and  B.  schmidli')  the  stalks  are  short 
and  found  only  in  early  larvae.  Stalks  are  longer  and  persist  into 
later  larval  stages  in  other  species,  reaching  a  ma.ximum  of  65% 
of  the  head  length  in  B.  bericoides. 

In  opisthoproctids  the  head  is  moderate  in  size  in  the  slender 
forms,  Bathylychnops  and  Dolichopleryx,  and  longer  and  more 
massive,  with  a  pronounced  hump  or  bend  at  the  nape,  in  the 
deep-bodied  genera.  All  genera  have  an  elongate  snout  and  Bath- 
ylychnops has  a  unique  triangular  flap  at  its  tip.  Bathylychnops 
has  round  eyes  that  are  rotated  slightly  dorsoanteriad.  In  the 
other  genera,  the  eyes  are  tubular  and  directed  dorsally,  even 
in  the  smallest  larvae  available.  Eye  diverticulae  with  associated 
accessory  retinae,  characteristic  of  opisthoproctid  adults,  begin 
to  form  at  the  end  of  the  larval  period. 

Fins —A  major  feature  of  all  argentinoid  larvae  is  the  devel- 
opment of  a  prominent  median  finfold  in  which  the  dorsal  and 
anal  fins  develop,  connected  to  the  trunk  by  a  series  of  hyaline 
strands  (Figs.  83-87).  The  first  fins  to  form  are  the  pectorals.  In 
argentinids  and  bathylagids  they  are  relatively  small  and  de- 
velop rays  late  in  the  larval  period.  Microstomatid  and  opis- 
thoproctid pectoral  fins  are  generally  larger;  however,  there  is  a 
wide  size  range,  from  relatively  small  fins  in  Microstoma  to 
large,  fan-like  fins  in  some  species  of  Nansenia  (e.g.,  N.  groen- 
landica) to  very  elongate  pectorals  in  Dolichopleryx  binocularis. 


Ossification  of  rays  begins  earlier  in  these  groups,  usually  before 
notochord  flexion. 

After  the  pectorals,  the  caudal  fin  is  usually  the  next  to  form. 
In  argentinids  notochord  flexion  and  development  of  principal 
caudal  rays  occurs  at  a  size  about  midway  in  larval  growth 
whereas  in  opisthoproctids  this  occurs  earlier  in  the  larval  pe- 
riod. In  bathylagids  the  process  is  somewhat  delayed  and  in 
some  species  (e.g.,  B.  euryops.  B.  milleri)  notochord  flexion  may 
not  be  completed  until  near  the  end  of  the  larval  period. 

The  dorsal  and  anal  fins  begin  to  form  at  about  the  stage  of 
notochord  flexion  in  all  argentinoids  except  opisthoproctids, 
where  notochord  flexion  slightly  precedes  the  appearance  of 
dorsal  and  anal  fins.  The  anal  fin  begins  forming  far  posteriad 
in  argentinoids,  just  posterior  to  the  anus  or  the  point  of  de- 
flection of  the  free  terminal  gut  section.  In  B.  milleri  and  in  the 
deep-bodied  opisthoproctids  with  coiled  guts  there  is  a  space 
between  the  anus  and  the  anal  fin  origin. 

The  position  of  the  dorsal  fin  is  varied  among  argentinoids 
and  forms  in  the  larvae  in  approximately  the  same  position  that 
it  will  occupy  in  the  adult.  The  fin  has  its  most  anteriad  location 
in  Argentina  where  its  origin  is  well  forward  of  the  midpoint  of 
the  body  (Fig.  83).  The  extreme  case  is  found  in  A.  silus  where 
snout  to  dorsal  origin  is  about  38%  of  the  body  length  in  larvae 
and  about  43%  in  adults.  In  most  bathylagids  the  dorsal  origin 
is  slightly  anterior  to  mid-body.  The  exceptions  are  B.  slilbius 
and  relatives,  where  the  dorsal  origin  is  slightly  posterior  to 
mid-body,  and  B.  wesethi  and  relatives  where  it  is  located  still 
further  posteriad. 


Fig.  84.  Larvae  of  Microstomatidae.  (A)  Microstoma  microstoma.  1 1.0  mm,  from  Schmidt  (1918);  (B)  Microstoma  sp.,  12.0  mm,  CalCOFl 
5 104  Sta.  90.52;  (C)  Nansenia  Candida.  8.4  mm,  CalCOFl  5007  Sta.  1 00.70;  (D)  N.  crassa.  8.5  mm,  CalCGR  5 103  Sta.  1 37.50;  (E)  N.  groenlandica, 
10.0  mm,  from  Schmidt  (1918);  (F)  N.  oblita,  9.0  mm,  ibid;  (G)  Xenopthalmichthys  danae.  16.5  mm,  from  Bertelsen  (1958). 


162 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  85.  Larvae  of  Balhylagus.  (A)  B.  milleri.  27.5  mm,  CalCOFl  5106  Sta.  70.60,  dorsal  view  of  9.5  mm  specimen  at  left;  (B)  B.  schmidli. 
31.5  mm,  CalCOFI  Northern  Holiday  Exped.  Sta.  31;  (C)  B.  snlbiits,  23.2  mm,  CalCOH  4905  Sta.  1 1 1.38,  dorsal  view  of  8.5  mm  specimen  at 
left;(D)  B.  pacificus.  21.4  mm,  CalCOFI  7905  Sta.  63.60;  (E)  B.  euryops  24.0  mm,  dorsal  view  of  14.0  mm  specimen  at  left,  from  Tuning  (1931); 
(F)  B.  antarcticus.  26.5  mm,  from  Yefremenko  (1983). 


AHLSTROM  ET  AL.:  ARGENTINOIDEI 

Table  41.    Meristics  of  Argentinoid  Fishes. 


163 


Branchiostegal 

Dorsal 

Anal 

Pectoral 

Pelvic 

Procurrenl 

Species 

Venebrae 

rays 

fin  rays 

fin  rays 

lin  rays 

fin  rays 

caudal  fin  rays 

Argentina 

altceae 

43-46 

5 

11-13 

13-15 

16-18 

10-12 

australiae 

50-53 

5 

10-12 

12-13 

13-14 

11-13 

hrucei 

45-47 

5 

10-12 

11-13 

18-20 

13-14 

elongata 

52-55 

5 

10-12 

11-14 

13-16 

11-12 

euchus 

48-49 

5 

12 

13-15 

16-18 

10-11 

georgei 

48-51 

5 

10-12 

10-13 

16-19 

12-14 

kagoshimae 

51-52 

5 

10-12 

11-13 

15-17 

11-12 

sialis 

47-51 

5 

10-13 

12-15 

11-18 

10-12 

12-Hl 

silus 

65-70 

6 

11-13 

11-17 

15-18 

12-13 

sphyraena 

46-55 

6 

10-12 

11-15 

12-15 

10-12 

stewarti 

53-54 

5 

10-12 

12-13 

18-21 

13-15 

striata 

48-52 

5 

10-12 

11-14 

18-21 

11-15 

10-1-9 

Glossanodon 

leiglossus 

49-51 

5 

12-14 

10-13 

19-22 

11-12 

tmeatus 

4 

11-13 

15 

18-21 

11-13 

mildredae 

50-52 

5 

13 

13 

23 

12-13 

polli 

5 

12-14 

11-14 

19-22 

12-13 

pygmaeus 

43-44 

5 

10-12 

11-13 

12-14 

10-12 

semifasciatus 

49 

5 

11-13 

11-13 

18-21 

10-12 

struhsakeri 

51-53 

12-14 

12-13 

23-25 

13-15 

Microstoma 

microstoma 

45-47 

3-4 

11-12 

8-9 

8 

9-11 

11  +  11 

sp.  (Pacific) 

49-50 

4 

9-11 

7-8 

11 

9 

10- 

-lH-10 

Xenophthalmichthys 

danae 

3 

10-12 

9-10 

7 

8-9 

10-1-9 

Nansenia 

atlanlica 

41-42 

4 

9-10 

8-9 

12-13 

10-11 

ardesiaca 

46-48 

4 

9-10 

9-10 

11-14 

10-12 

Candida 

44-47 

3 

9-10 

8-9 

9-11 

9-11 

11-1-14 

crassa 

43-46 

4 

9-10 

8-9 

11-13 

10-11 

groenlandica 

42-45 

3 

9-10 

8-10 

11-13 

10-12 

ohiita 

42-45 

4 

10-11 

9-10 

10-11 

10-11 

Bathylagus 

amarcticus 

2 

9-11 

21-25 

9-10 

argyrogasler 

2 

12 

14-15 

8 

bericoides 

48-53 

-) 

10-11 

18-22 

10-12 

9-10 

euryops 

44-46 

2 

9-11 

16-19 

7-12 

7-9 

greyae 

2 

11-13 

13 

12-13 

10-11 

longirostris 

48-51 

2 

10-12 

19-21 

9-12 

9-10 

mtlleri 

51-55 

2 

6-9 

20-28 

11-16 

6-8 

16- 

-18-1-15-17 

nigrigenys 

41 

2 

11-12 

14-17 

10 

8-10 

ochotensis 

47-49 

2 

9-12 

12-15 

9-11 

9-10 

13- 

-14+15-16 

pacificus 

45-49 

2 

8-9 

15-22 

7-11 

7-10 

13+13-14 

schniidti 

47-52 

2 

10-11 

11-14 

8-9 

8-9 

16- 

-17+16 

stilhius 

38-42 

2 

9-11 

11-14 

8-11 

8-10 

12- 

-16+13-15 

urolranus 

39-42 

2 

9-10 

10-11 

9-11 

7-8 

12- 

-14+12-13 

weselhi 

43-46 

2 

12-13 

14-16 

10-11 

9-11 

14-15-1-14-15 

Dolichopteryx 

anascopa 

2 

10 

12 

14 

12 

bmocularis 

2 

15 

11 

14 

9 

hrachyrhynchus 

2 

13 

12 

13 

8 

longipes 

41-44 

2 

10-11 

8-9 

13 

8-9 

Bathylychnops 

exilis 

81-84 

2 

14-16 

13-14 

12-13 

7 

RhynJichyalus 

natalensis 

4 

10-12 

10 

19-20 

11-12 

Macropinna 

microstoma 

36 

3 

11-12 

14 

17-19 

10 

Winteria 

telescopa 

8 

8 

12-14 

9 

164 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  41. 

Continued. 

Species 

Vertebrae 

Branchios 
rays 

legal 

Dorsal 
fin  rays 

Anal 
tin  rays 

Pectoral 
fin  rays 

Pelvic 
fin  rays 

Procurrent 
caudal  fin  rays 

Opisthoproctus 
grimaldii 
soleatus 

31 

2 
2 

12-14 
10-12 

8 
13-14 

11 

9-11 

10 

The  dorsal  fin  forms  in  a  variety  of  positions  among  micro- 
stomatids.  In  most  species  of  Nansenia.  the  dorsal  fin  originates 
slightly  posterior  to  mid-body,  although  in  some  species  (e.g., 
A',  groenlandica),  its  origin  is  slightly  anterior  to  mid-body.  The 
dorsal  origin  is  further  posteriad  in  Microstoma.  In  M.  microsto- 
ma predorsal  length  is  about  67-68%  of  the  body  length  and 
assumes  a  more  anterior  position  in  adults  (ca.  63%).  In  larvae 
of  the  Pacific  species  predorsal  length  is  about  75%  of  the  body 
length,  and  is  slightly  more  posteriad  in  adults.  In  adult  Xen- 
ophthalijuchthys  the  dorsal  origin  is  at  mid-body;  however,  in 
the  16.5  mm  specimen  from  the  Atlantic  (Bertelsen,  1958)  pre- 
dorsal length  is  62%  of  the  body  length.  In  our  single  larva  (12.2 
mm)  from  the  Pacific  predorsal  length  is  75%  of  body  length, 
indicating  a  marked  anteriad  migration  during  ontogeny  or  strong 
allometric  growth  posterior  to  the  dorsal  fin.  Alternatively,  the 
Pacific  form  may  prove  to  be  distinct  when  adult  specimens  are 
captured. 

The  dorsal  fin  in  opisthoproctids  is  located  posteriad  on  the 
body.  This  is  most  marked  in  the  slender  forms,  Bathylychnops 
and  Dolichopteryx.  and  reaches  an  extreme  in  D.  hinocularis 
where  predorsal  length  is  greater  than  Vj  of  the  body  length.  In 
the  deep-bodied  genera  the  dorsal  origin  is  posterior  to  mid- 
body,  but  less  so  than  in  the  slender-bodied  forms. 

The  pelvic  fins  are  the  last  fins  to  form  in  most  argentinoids, 
usually  late  in  the  larval  period.  The  exception  is  opisthoproctids 
where  the  pelvic  fins  form  early  in  the  larval  period.  In  argen- 
tinids,  bathylagids  and  microstomatids  the  pelvic  fins  form  at 
about  mid-body,  below  the  dorsal  fin.  In  the  slender  opistho- 
proctid  genera  the  pelvics  form  well  back  on  the  body,  but 
anterior  to  the  dorsal  fin.  Among  the  deep-bodied  genera,  Op- 
isthoproctus forms  the  pelvics  far  back  on  the  body,  beneath  the 
dorsal  fin.  In  Rhynchohyalus  and  Macropinna  the  pelvics  de- 
velop just  posterior  to  mid-body  and  anterior  to  the  dorsal  fin. 
In  the  larvae  the  fins  are  elevated  to  the  sides  of  the  body.  This 
position  persists  in  juvenile  and  adult  Macropinna  where  the 
fins  are  located  just  behind  and  below  the  pectoral  fin  bases. 
The  pelvic  fins  become  elongate  in  Dolichopteryx  and  the  deep- 
bodied  genera.  The  pelvic  fin  base  is  pedunculate  in  opistho- 
proctid  larvae,  a  condition  that  persists  into  the  adults  of  some 
genera,  notably  Dolichopteryx.  Argentinoids,  except  Microsto- 
ma. Xenophthalmichthys  and  some  species  of  Dolichopteryx. 
develop  adipose  fins  late  in  the  larval  period. 

A  summary  of  meristics  of  argentinoids  is  given  in  Table  41. 
The  sequence  of  ossification  of  fins  and  other  skeletal  elements 
o(  Bathylagus  schmidti  is  described  by  Dunn  (1983a). 


Pigmentation.  — \n  argentinids,  pigmentation  consists  of  a  series 
of  6-8  ventral  trunk  blotches  that  extend  from  the  pectoral  fin 
base  to  the  end  of  the  gut  (Fig.  83).  The  series  is  continued 
posteriorly  as  I  or  2  median  ventral  blotches  and  ends  as  a  large 
blotch  at  the  caudal  region.  The  number  of  blotches  is  constant 
for  each  species,  as  is  the  sequence  of  formation.  In  Argentina 
sialis  and  Glossanodon  the  ventral  blotches  expand  dorsally  as 
lateral  bars,  but  this  does  not  occur  in  A.  silus  and  A.  sphyraena. 
These  latter  species  differ  additionally  in  lacking  the  internal 
head  pigment  which  develops  in  A.  sialis  and  Glossanodon  lar- 
vae. 

A  feature  common  to  most  microstomatid  larvae  is  a  heavy 
line  of  embedded  pigment  above  the  gut  (Fig.  84).  In  Micro- 
stoma this  pigment  continues  forward  to  the  gill  arches  and 
within  the  head  anteriorly  to  the  snout.  In  Nansenia,  head  pig- 
mentation is  superficial,  or  concentrated  ventrally  on  the  head. 
In  Microstoma,  an  embedded  dorsal  line  of  pigment  is  located 
posterior  to  the  dorsal  fin.  Dorsal  pigmentation  in  Nansenia 
may  take  the  form  of  a  series  of  embedded  blotches  (e.g.,  N. 
crassa)  or  an  embedded  line  of  melanophores  running  the  length 
of  the  body  (e.g.,  N.  ohlita).  Most  microstomatids  have  con- 
spicuous melanistic  pigment  associated  with  the  caudal  fin  re- 
gion. A  notable  feature  oi Microstoma  and  some  Nansenia  (e.g., 
N.  crassa)  is  the  presence  of  heavy  melanistic  pigment  at  the 
curve  of  the  gut  loop.  Our  single  damaged  specimen  of  Xenoph- 
thalmichthys (12.2  mm)  has  pigmentation  similar  to  Micro- 
stoma but  lacks  the  posterior  dorsal  body  pigment  and  has  a 
series  of  slanted  melanophores  along  the  hypaxial  myosepta. 

Pigment  patterns  in  bathylagids  may  be  grouped  into  two 
categories— those  species  with  large  isolated  melanophores  (Fig. 
85)  and  those  with  linear  series  of  smaller  melanophores  (Fig. 
86).  Bathylagus  milleri  has  a  unique  pattern  of  opposing  dorsal 
and  ventral  midline  melanophores,  large  melanophores  on  the 
head  and  pectoral  fin  base  and  a  large  lateral  blotch  on  the 
notochord  tip. 

Bathylagus  stilbius  and  B.  urotranus  develop  a  series  of  5-6 
melanophores  on  each  side  of  the  posterior  section  of  the  gut. 
A  single  large  melanophore,  is  found  on  the  lower  trunk  midway 
between  the  pectoral  fin  and  the  anus  and  the  head  has  mela- 
nophores, chiefly  on  the  upper  and  lower  jaws  and  opercle  (Fig. 
85).  B.  schmidti  differs  in  having  a  series  of  lower  trunk  blotches 
and  1  or  2  postanal  lateral  blotches. 

Bathylagus  euryops  has  a  series  of  3-6  melanophores  on  the 
lateral  surface  of  the  gut  and  3-5  large  melanophores  on  the 
lateral  surface  of  the  trunk  (Fig.  85).  Other  pigmentation  consists 


Fig.  86.  Larvae  of  Bathylagus.  (A)  B.  hericoides.  \1 .1  mm,  Dana  Sta.  4007,  dorsal  view  ofl  1.8  mm  specimen  at  left;  (B)  B.  longirostris.  20.1 
mm,  SIO/STOW  XIII  Exped.,  dorsal  view  of  12.4  mm  specimen  at  left;  (C)  B.  ocholensis.  21.5  mm,  CalCOFl  5106  Sta.  77.65,  dorsal  view  of 
8.5  mm  specimen  at  left;  (D)  B.  wesetlu.  1  1.3  mm.  from  Ahlstrom  (1972b),  dorsal  view  of  8.5  mm  specimen  at  left;  (E)  B.  mgrigenys.  21.8  mm, 
SIO  Shellback  Exped.  Sta.  92,  dorsal  view  of  8.7  mm  specimen  at  left;  (F)  B.  argyrogaster.  17.1  mm,  Dana  Sta.  4003. 


AHLSTROM  ET  AL.:  ARGENTINOIDEI 


165 


166 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


AHLSTROM  ET  AL.:  ARGENTINOIDEI 

Table  42.    Characters  used  in  Analysis  of  Four  Argentinoid  Groups. 


167 


Character 
number 


Dervied  character  state 


Outgroup 


1  Accessory  cartilage  at  posterior  tip  ceratobr.  5 

2  PU,  +  U,  fused 

3  Light  organs  present 

4  Frontals  fused 

5  Epibr.  4  with  one  post.  art.  surface 

6  Larval  gut  with  stomach 

7  Pelvic  fins  form  early  and  large 

8  Swimbladder  absent 

9  Urodermal  absent 

10  LL  scales  extend  onto  caudal  fin 

1 1  Larval  gut  folded 

12  Extrascapular  attached  to  pterotic 

13  Uncinate  process  lacking  on  epibr.  4 

14  Pectoral  fin  forms  early  and  large 


Osmerids 

Greenwood  and  Rosen,  1971 

Teleosts  in  general 

Goody,  1969 

Teleosts  in  general 

Bertelsen  and  Munk,  1964 

Teleosts  in  general 

Cohen,  1964b 

Osmerids 

Greenwood  and  Rosen,  1971 

Osmerids 

This  paper,  Heame  (this  volume) 

Osmerids 

This  paper,  Heame  (this  volume) 

Teleosts  in  general 

Cohen,  1964b 

Teleosts  in  general 

Greenwood  and  Rosen,  1971 

Teleosts  in  general 

Osmerids 

Heame  (this  volume) 

Teleosts  in  general 

Chapman,  1942 

Osmerids 

Greenwood  and  Rosen,  1971 

Osmerids 

This  paper,  Heame  (this  volume) 

of  a  line  of  small  melanophores  above  and  below  the  notochord 
tip,  a  patch  of  melanophores  on  the  opercle  and  groups  of  small 
melanophores  on  the  upper  and  lower  jaws.  Balhylagns  anlarcti- 
cus  has  3  lateral  gut  spots,  a  large  lateral  trunk  melanophore  at 
the  I0lh-I2th  myomere,  and  head  and  notochord  pigment  sim- 
ilar to  that  of  B.  euryops.  Early  larvae  of  B.  pacificus  have  a 
large  lateral  blotch  at  mid-body  and  another  one  posteriad  on 
the  body.  Initially  these  melanophores  are  located  at  the  junction 
of  the  gut  and  body  but  in  later  larvae  are  located  on  the  trunk. 
Later  a  3rd  blotch  forms  midway  between  these  two.  A  4th 
lateral  trunk  blotch  forms  in  some  late  larval  specimens  between 
the  pectoral  fin  and  the  large  mid-body  blotch  and  melanophores 
form  lateral  to  the  liver  and  at  the  free  terminal  section  of  the 
gut.  Head  and  notochord  pigment  is  similar  to  B.  euryops  and 
B.  antarcticus. 

Bathylagiis  hericoides  is  unusual  in  having  only  a  series  of  as 
many  as  18  lateral  gut  melanophores  (Fig.  86).  Late  postflexion 
larvae  develop  pigment  on  the  lower  jaw,  isthmus,  opercle,  pec- 
toral fin  base  and  lateral  caudal  peduncle.  Bathylagus  longiros- 
tris  develops  a  heavier  pattern  of  pigmentation,  beginning  with 
a  series  of  small  melanophores  on  the  posterior  section  of  the 
gut  in  early  larvae.  Also  in  preflexion  larvae  a  series  of  rect- 
angular-shaped melanophores  develops  on  the  hypaxial  myo- 
meres. Later  in  the  larval  period  the  lateral  gut  series  is  extended 
forward  along  the  entire  gut,  although  with  wider  spacing  than 
on  the  posterior  gut  section.  Also,  the  epaxial  myomeres  develop 
rectangular-shaped  melanophores,  beginning  posteriorly  and  ac- 
cruing anteriorly.  The  head  develops  pigmentation  from  the 
opercle  to  the  jaws  (Fig.  86).  Bathylagus  ochotensis  develops  a 
similar  pigment  pattern  except  that  the  melanophores  on  the 
posterior  gut  section  are  comparatively  larger  and  fewer,  the 
anterior  region  of  the  gut  lacks  melanophores  and  the  epaxial 
myomere  series  is  limited  to  the  posterior  region. 

Larvae  of  B.  wesethi.  B.  nigrigenys  and  B.  argyrogaster  have 
a  similar  pigment  pattern  that  differs  markedly  from  that  of 
other  Bathylagus  (Fig.  86).  Initially  there  is  a  series  of  paired 


melanophores  dorsolateral  to  the  gut,  extending  from  the  pec- 
toral fin  base  to  the  terminal  section.  These  remain  throughout 
the  larval  period  but  become  embedded  and  obscured  in  late 
larvae.  Bathylagus  nigrigenys  begins  with  about  8  pairs,  which 
increase  to  10,  whereas  B.  wesethi  begins  with  6  pairs  and  has 
7-8  during  most  of  the  larval  period.  Both  species  develop  pig- 
ment at  the  notochord  tip;  B.  wesethi  has  a  dorsal  and  ventral 
spot,  while  B.  nigrigenys  has  only  a  ventral  spot.  At  notochord 
flexion  a  series  of  melanophores  appears  along  the  hypaxial 
region  of  the  body  and,  soon  after,  a  series  develops  along  the 
epaxial  myomeres.  More  lateral  series  are  added  and  in  late 
larvae  the  entire  body  is  covered.  Melanophores  also  form  in 
the  median  finfold  of  advanced  larvae.  Initially  head  pigmen- 
tation consists  of  melanophores  on  the  opercle  and  jaws  but  in 
later  larvae  the  entire  head  is  covered. 

Opisthoproctid  larvae  have  distinctive  and,  in  some  genera, 
heavy  pigment  patterns  (Fig.  87).  Bathylychnops  has  a  dorsal 
series  of  6  large  paired  blotches  that  permeate  the  musculature, 
bridge  across  the  longitudinal  septum  and  expand  onto  the  fin- 
fold.  A  series  of  8  large  ventrolateral  blotches  alternate  with 
those  of  the  dorsal  series,  with  the  exception  that  the  postanal 
blotch  lies  opposite  the  dorsal  blotch  and  expands  to  form  a 
band.  A  large  blotch  covers  the  base  of  the  caudal  fin.  The  head 
IS  heavily  pigmented  with  superficial  melanophores  on  the  bran- 
chiostegals,  urohyal  and  lateral  brain  and  deeply  embedded  me- 
lanophores in  the  snout,  jaws,  cheek  and  ventral  brain  region. 
The  lower  limbs  of  the  gill  arches  and  their  filaments  are  heavily 
pigmented  as  are  both  the  pectoral  and  pelvic  fin  bases. 

The  species  of  Dolichopteryx  have  lateral  series  of  melano- 
phores above  the  gut  and  some  species  develop  serial  melano- 
phores on  the  hypaxial  myomeres  (Fig.  87).  Head  pigment  con- 
sists of  melanophores  on  the  jaws,  gill  arches  and,  in  most  species, 
the  internal  snout  region.  Macropmna  develops  a  series  of  slant- 
ed melanophores,  one  on  each  hypaxial  myomere,  and  a  heavy 
embedded  blotch  at  the  pelvic  fin  base,  that  expands  both  dorsad 
and  ventrad  as  a  band.  The  caudal  fin  base  has  a  large  blotch 


Fig.  87.  Larvae  of  Opisthoproctidae.  (A)  Bathylychnops  exilis.  1  5.6  mm.  CalCOFI  7203  Sta.  67.80;  (B)  Ventral  view  of  above;  (C)  Dolichopteryx 
hinoculans.  58.0  mm,  redrawn  from  Roule  and  Angel  (1930);  (D)  Afacropinna  microstoma.  1  1.7  mm,  CalCOFI  7412  Sta.  120.50;  (E)  Ventral 
view  of  above;  (F)  Rhynchohyalus  natalensis.  23.0  mm,  from  Bertelsen  et  al.  (1965);  (G)  Opisthoproctus  grimaldii.  14.0  mm  from  Schmidt  (1918). 


168 
ARGENTINIDAE 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 

MICROSTOMATIDAE  BATHYLACIDAE  OPISTHOPROCTIDAE 


3,4,5,6,7, 
(13),  (14) 


Fig.  88.     Cladogram  showing  the  distribution  of  character  states  in  four  nominal  families  of  argentinoid  fishes.  Numbers  refer  to  characters  in 
Tables  42  and  43.  Parentheses  mdicate  character  reversals. 


and  in  the  gut  region  there  is  pigment  above  the  terminal  section 
and  ventral  to  the  liver.  Head  pigment  is  confined  to  the  lower 
jaw.  The  pigment  pattern  of  Rhynchohyalus  as  described  by 
Bertelsen  et  al.  (1965)  consists  of  a  series  of  four  dusky  bars 
beginning  at  the  pelvic  fin  and  ending  at  the  caudal  fin  base. 
Embedded  beneath  these  is  a  layer  of  diffuse  melanophores 
which  becomes  denser  toward  the  caudal.  The  pectoral  fin  bases 
are  pigmented  and  in  the  ventral  region  there  are  melanophores 
on  the  isthmus  and  gut.  The  anal  light  organ  is  covered  with  a 
melanistic  sheath.  The  late  larval  specimen  of  Opisthoproctus 
grimaldii  illustrated  by  Schmidt  (1918)  shows  a  diffuse  covering 
of  melanophores  over  the  body  and  a  dusky  bar  extending  down 
from  the  dorsal  fin.  A  10  mm  larva  of  O.  so/eat  us  in  our  col- 
lection has  a  pigment  pattern  similar  to  Macropuma,  with  a 
series  of  slanted  melanophores  on  the  hypaxial  myomeres, 
embedded  blotches  at  the  pelvic  and  caudal  fin  bases,  pigment 
at  the  liver  and  ventrally  at  the  angles  of  the  lower  jaw. 

Transformation  stage 

In  argentinids  transformation  from  larva  to  demersal  juvenile 
is  a  prolonged  process  and  pelagic  juveniles  with  the  retained 
larval  pigment  blotches  or  bars  have  been  reported  many  times 
(see  Cohen,  1958;  Nishimura,  1966).  Morphological  changes 
(e.g.,  deepening  of  the  body,  prolongation  of  the  snout,  eye 
enlargement)  and  the  masking  of  the  larval  pigment  occur  grad- 
ually. The  beginning  of  this  stage  may  be  defined  by  the  folding 
of  the  anterior  gut  region  to  form  a  stomach.  This  occurs  at  25- 


30  mm  in  Argentina  sialis,  but  has  not  been  documented  for 
other  species.  Pelagic  juveniles  of  Glossanodon  and  A.  sialis 
develop  a  silvery  stripe  at  the  lateral  line  region.  This  has  not 
been  reported  for  pelagic  juveniles  of  A.  silus  and  A.  sphyraena 
and  may  afford  an  additional  character  for  separating  Argentina 
into  two  groups.  The  end  of  the  pelagic  juvenile  stage,  marked 
by  the  development  of  scales  and  silvery  integument,  is  attained 


Table  43.    Distribution  of  Char.acter  States  in  Folir  Nominal 
F.MHILIES  OF  Argentinoid  Fishes.  Direction  of  transformation  A    *  B. 


Character 

Opistho- 

number 

Argenlinidae 

Micrstomatidae 

Bathy[ag]dae 

proctidae 

1 

B 

B 

B 

A 

2 

B 

B 

B 

A 

3 

A 

A 

A 

B 

4 

A 

A 

A 

B 

5 

A 

A 

A 

B 

6 

A 

A 

A 

B 

7 

A 

A 

A 

B 

8 

A 

A 

B 

A 

9 

A 

A 

B 

A 

10 

A 

B 

A 

A 

II 

A 

B 

A 

A 

12 

B 

A 

A 

A 

13 

A 

B 

B 

B 

14 

A 

B 

A 

B 

AHLSTROM  ET  AL.:  ARGENTINOIDEI 


169 


at  various  lengths  by  different  species.  Schmidt  (1906c)  reports 
complete  transformation  at  about  50  mm  in  A.  sphyraena  and 
at  a  much  larger  size  in  A.  silus.  Size  at  completion  of  trans- 
formation in  Glossanodon  species  is  also  in  the  50-100  mm  size 
range  (Nishimura,  1966). 

Microstomatids  develop  a  lustrous  guanine  layer  on  the  in- 
tegument in  late  larvae  and  some  species  develop  distinct  ju- 
venile pigmentation.  In  Mil  rosloma  juvenWcs  the  region  of  the 
body  from  the  dorsal  fin  origin  posteriad  is  more  darkly  pig- 
mented than  the  rest  of  the  body,  and  grades  to  a  solid  black 
pigment  at  the  caudal  fin  base.  Juveniles  of  some  Nansenia 
species  develop  heavy  melanistic  pigment  at  the  base  of  the 
caudal  fin  and  often  at  the  base  of  the  adipose  fin  (Schmidt, 
1918;  Kawaguchi  and  Butler,  in  press). 

Bathylagids  have  a  direct  transformation  and  undergo  a  marked 
morphological  change  from  the  slender  larval  form  to  the  ju- 
venile form,  characterized  by  a  large  head  and  eyes  and  deeper 
body.  The  gut  becomes  coiled  and  covered  by  a  black  peritoneal 
sheath.  The  head  becomes  heavily  pigmented  but  the  body  is 
slower  to  develop  the  black  pigment  characteristic  of  all  Bath- 
ylagiis  species  (other  than  the  B.  stilbnis  group)  and,  in  species 
such  as  B.  euryops  and  B.  nulleri.  the  large  larval  melanophores 
are  visible  in  specimens  up  to  30  mm  and  50  mm  respectively. 

In  the  deep-bodied  opisthoproctid  genera  transformation  to 
the  juvenile  stage  is  marked  by  deepening  of  the  body  and  at- 
tainment of  melanistic  integument  and  large  scales.  Cohen  ( 1 960) 
described  the  large  (up  to  124  mm)  transitional  specimens  of 
Bathylychnops  which  are  semi-transparent  and  retain  the  large 
larval  pigment  blotches.  Sexually  mature  specimens  of  Doli- 
chopteryx  are  semi-transparent,  have  a  membranous  body  en- 
velope, poorly  developed  musculature,  an  exposed  gut  covered 
only  by  peritoneum,  weakly  attached  fins  and  melanistic  pig- 
ment of  the  type  usually  associated  with  larvae  (Cohen,  1960). 

Relationships 

Our  survey  of  argentinoid  ontogenetic  characters  provides 
insight  into  some  of  the  systematic  questions  posed  at  the  be- 
ginning of  the  paper.  A  close  relationship  between  argentinoids 
and  alepocephaloids  is  not  supported  since  the  latter  hatch  from 
large  eggs  (estimated  at  3-4  mm  based  on  size  of  yolk-sac  lar- 
vae), have  direct  development,  and  share  no  specialized  onto- 
genetic characters  with  argentinoids.  Four  major  argentinoid 
lineages  can  be  defined  by  specializations  of  the  eggs  and  larvae 
and  thus  four  families  recognized:  Argentinidae,  Microstoma- 
tidae,  Bathylagidae,  and  Opisthoproctidae.  Argentina  and  Glos- 


sanodon have  generalized  larvae  except  that  all  known  species 
have  distinct  lateral  series  of  melanistic  blotches  or  bands,  not 
found  elsewhere  among  argentinoids.  The  pattern  of  banding 
does  not  separate  the  two  genera. 

All  known  bathylagid  eggs  have  multiple  oil  globules.  A  num- 
ber of  bathylagid  groups  are  apparent  from  larval  characters:  1) 
niillcri,  2)  slilhms-schmidti-iirotranus,  3)  euryops- pad ficus-ant- 
arcticus,  4)  hericoides-longirostris,  5)  wesethi-argyrogaster-ni- 
grigenys.  Of  these  groups,  stilbius-schmidti-urotranus  has  the 
most  generalized  morphology  and  pigmentation,  lending  no 
support  for  its  separation  as  a  distinct  genus. 

Opisthoproctid  larvae  share  a  number  of  neotenic  features, 
including  a  saccular  stomach.  Except  for  body  shape,  Dolichop- 
teryx  shares  more  derived  larval  characters  with  the  deep-bodied 
genera  than  with  Bathylychnops.  and  the  latter  has  a  number  of 
characters  unique  to  opisthoproctids.  Division  of  the  family 
based  on  body  shape  is  not  supported  by  ontogenetic  evidence. 

Ontogeny  offers  little  information  on  species  composition  of 
genera,  because  only  a  fraction  of  argentinoid  eggs  and  larvae 
are  known.  However,  egg  and  larval  characters  clearly  separate 
Atlantic  and  Pacific  Microstoma  as  distinct  species.  Bathylagits 
hericoides  larvae  from  the  Atlantic  and  Pacific  are  indistinguish- 
able. The  same  is  true  for  B.  longirostris  from  all  oceans.  Bath- 
ylagus  nigrigenys  and  B.  argyrogaster  larvae  are  indistinguish- 
able, lending  support  for  a  single  circumtropical  species. 
Bathylagus  stilbiiis  eggs  and  larvae  are  indistinguishable  from 
those  of  B.  urotranus. 

We  have  attempted  to  analyze  the  distribution  among  four 
nominal  groups  of  argentinoids,  of  14  characters,  four  of  which 
are  taken  from  developmental  stages  and  10  from  the  adult 
(Table  42).  We  have  used  teleosts  in  general  and  osmerids  as 
our  outgroup  following  Fink  and  Weitzman  ( 1 982).  Distribution 
of  character  states  are  presented  in  Table  43. 

A  possible  arrangement  of  groups  based  on  the  fewest  number 
of  character  reversals  is  presented  in  Figure  88.  Opisthoproc- 
tidae appears  to  be  a  well-founded  family.  More  precise  inter- 
pretation of  the  inter-relationships  and  nomenclatural  ranking 
for  argentinids,  microstomatids,  and  bathylagids  requires  ad- 
ditional data. 

(H.G.M.)  Southwest  Fisheries  Center,  P.O.  Box  271,  La 
JoLLA,  California  92038;  (D.M.C.)  Natural  History 
Museum  Los  Angeles  County  900  E.xposition  Boule- 
vard, Los  Angeles,  California  90036. 


Stomiatoidea:  Development 
K.  Kawaguchi  and  H.  G.  Moser 


FISHES  of  this  group  of  midwater  predators  are  characterized 
by  their  dark  coloration,  serial  photophores,  large  jaws, 
fang-like  teeth,  and  chin  barbels.  Traditionally  they  have  been 
grouped  in  six  families  allied  to  the  lightfishes  and  hatchetfishes 
(Weitzman,  1974),  and  together  are  now  considered  monophy- 


Ictic  and  given  ordinal  status  (Rosen,  1 973;  Fink  and  Weitzman, 
1982).  Fink  (this  volume)  gives  evidence  for  reducing  the  six 
stomiatoid  families  to  one.  Because  knowledge  of  stomiatoid 
ontogeny  lags  far  behind  that  of  the  adults,  for  convenience  of 
discussion  we  use  Weitzman's  (1974)  grouping  of  the  families 


170 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  89.  Larvae  of  Slomias  and  Chauliodus.  (A)  5.  alnventer.  4.6  mm,  CalCOFI  7501  Sta.  97.60;  (B)  5.  atriventer,  10.0  mm,  CalCOH  6604 
Sta.  107.65;  (C)  5.  atmenler.  22.2  mm;  CalCOH  6604  Sta.  107.65;  (D)  S.  ferox.  30  mm,  from  Ege,  (1918);  (E)  C.  sham.  6.0  mm;  from  Mito 
(1961a);  (F)  C  macouni.  15.0  mm,  CalCOH  6204  Sta.  60.60;  (G)  C.  macouni.  45.2  mm,  CalCOH  5707  Sta.  67.60. 


Astronesthidae,  Stomiatidae,  Chauliodontidae.  Melanostomia- 
tidae,  Maiacosteidae,  and  Idiacanthidae,  in  the  Superfamily  Sto- 
miatoidea. 

Eggs 

Eggs  are  known  for  Chauliodus,  Stomias,  and  Tactostoma 
and  have  in  common  a  round  shape,  smooth  chorion,  and  seg- 
mented yolk,  Chauliodus  eggs  have  a  wide  perivitelline  space 
and  lack  an  oil  globule.  Egg  diameters  are:  C  sloani.  2.2-2.5 


mm  (Sanzo,  1931d);  C.  barbatus,  3.1-3.6  mm  (Pertseva-Os- 
troumova  and  Rass,  1973);  C  macouni.  2.7-3.1  mm,  with  an 
initial  yolk  diameter  of  1.3-1.5  mm  (original  data).  Mito  (196  la) 
described  an  egg,  referred  to  C.  sloani.  l.\l  mm  in  diameter 
with  no  oil  globule  but  with  a  second  membrane.  Stomias  eggs 
have  a  second  membrane,  a  single  oil  globule  and  the  following 
diameters:  S.  colubrinus.  1.3-1.5  mm,  with  inner  membrane 
1.05-1.1  mm  (Pertseva-Ostroumova  and  Rass,  1973);  S.  atri- 
venter. 0.88-0.92  mm,  inner  membrane  diameter  is  0.82-0.84 


KAWAGUCHI  AND  MOSER:  STOMIATOIDEA 


171 


Table  44.    Meristic  Counts  of  Stomiatoid  Genera.  Most  frequent  count  or  range  is  followed  by  overall  range  or  infrequent  count  in  parentheses. 
Data  from  Gibbs  (1964a,b),  Gibbs  et  al.  (1983),  Morrow  (1964a,  b,  c).  Morrow  and  Gibbs  (1964),  Bolin  (1939a),  Imai  (1941).  onginal  counts. 


Vertebrae 

Fin  rays 

Family  and  genus 

Dorsal 

Anal 

Pectoral 

Pelvic 

Stomiatidae 

Macrostomias 

164 

13,14 

16(15-18) 

7(6) 

4 

Stomias 

64-83 

17-20(16-22) 

19-21  (18-25) 

6-7  (6-9) 

5(4) 

Chauliodontidae 

Chautiodus 

51-62 

6,7  (5-7) 

10-12(10-13) 

12,13(11-14) 

7  (6-8) 

Astronesthidae 

Astronesthes 

46-58 

15(10-21) 

12-22 

8  (5-9) 

7  (6-8) 

Borostomias 

53-55 

13(10-14) 

13-16(10-19) 

7  (6-9) 

7 

Hetempholus 

66 

11  (13) 

12-15(17) 

7 

7 

Neoneslhes 

53 

9-11  (12) 

25-27  (22-28) 

8(7) 

7  (6-8) 

Rhadineslhes 

67 

11  (12.13) 

18(19-21) 

7  (6-8) 

7 

Melanostomiatidae 

Bathophilus 

38-45  (33-50) 

13-16(9-18) 

15-16(9-18) 

1-37 

11-16(4-26) 

Chiroslomias 

54-55 

18-20 

22-26 

6 

7 

Echwstoma 

57-59 

11-14(11-16) 

13-18(13-19) 

1  +  3 

8 

Eiistomias 

56-69 

21-25(20-30) 

32-46 

0-13 

7 (6-8) 

Ftagellostomias 

65 

16(14-17) 

23-25(21-26) 

I  +  8-9  +  I1 

7 

Grainmalostomias 

50-56 

18-21 

21-23(20-24) 

4-11 

7-8 

Leptostomias 

77-80  (75-83) 

16-22 

20-29 

10(9-11) 

7(8) 

Melanostomias 

50-57 

12-17 

16-20 

5(4-6) 

7(8) 

Opostoimas 

60 

21 

24 

1+4 

8 

Pachystomias 

48 

22(21-24) 

27  (25-29) 

5-6 

8-9  (7) 

Photonectes 

49-64 

15-24 

17-24 

0-3 

7(6) 

Tactosloma 

80-82 

14-16 

19-22 

0 

8-10 

Thysanaclis 

61 

17-18 

21-25 

1+10,11 

7 

Trigonolampa 

61-62 

19-20(18) 

18(19) 

5 

7 

Malacosteidae 

Anslostomias 

44-56 

18-26 

24-32 

6-10(3-17) 

6 

Malacosteus 

49 

14-19(20) 

17-21 (23) 

3-4  (5) 

6 

Photostomias 

52-58 

22-28 

25-32 

0 

6 

Idiacanthidae 

Idiacamhus 

79-85 

54-74 

34 (33-39) 

0 

6 

mm,  oil  globule  diameter  of  0.20-0.25  mm,  initial  yolk  diameter 
of  0.70  mm  (original  data).  Tactosloma  macropus  eggs  have  a 
single  membrane,  1.44-1.54  mm  in  diameter,  an  oil  globule 
0.30-0.40  mm  in  diameter  and  an  initial  yolk  diameter  of  0.78- 
0.80  mm  (original  data).  Eggs  of  C.  macouni  and  5.  athventer 
are  illustrated  in  Matarese  and  Sandknop  (this  volume). 

Larvae 

Larvae  of  Stomiatoidea  occur  in  the  upper  water  column, 
some  at  the  surface.  In  most  groups  the  larvae  are  elongate,  have 
a  large  head,  elliptical  eyes  that  protrude  slightly  from  the  dorsal 
head  profile,  an  elongate,  straight  gut  (trailing  from  the  body  in 
some  species),  a  well  developed  finfold,  large  paddle-shaped 
pectoral  fins  that  lack  rays  until  transformation,  and  late-form- 
ing pelvic  fins.  Melanophore  patterns  provide  a  useful  set  of 
characters  and  genera  usually  have  a  distinct  pattern.  The  larval 
melanophores  are  retained  in  a  subcutaneous  position  in  trans- 
forming specimens  and  provide  a  means  for  identifying  larvae. 
During  transformation,  photophores  form  simultaneously  and 
initially  are  unpigmented.  Counts  of  fin  rays,  vertebrae,  and 
photophores  are  summarized  in  Tables  44  and  45. 

Stomiatidae  (Fig.  89).  —  Larvae  of  five  species  are  known  (Table 


46).  Larvae  are  3-4  mm  at  hatching  and  have  an  elongate  yolk 
sac.  The  slender  body  is  round  in  cross-section,  but  becomes 
slightly  deeper  by  late  postflexion.  The  head  is  relatively  small 
with  a  slightly  flattened  snout.  The  eyes  are  elliptical.  The  elon- 
gate gut  extends  almost  the  entire  length  of  the  body  and  has  a 
slightly  enlarged  terminal  section  that  reaches  the  anal  fin  origin. 
The  median  finfold  is  small  and  best  developed  posteriorly.  The 
opposing  dorsal  and  anal  fins  develop  far  posteriad  on  the  body 
in  early  postflexion  larvae,  but  the  pelvic  fins  do  not  appear 
until  just  before  transformation. 

Late-stage  embryos  oi  Stomias  have  melanophores  along  the 
dorsum,  which  migrate  ventrad  and  form  a  distinct  series  be- 
tween the  body  and  gut.  This  series  extends  to  the  tip  of  the 
notochord.  The  series  is  lost  before  notochord  flexion  but,  in 
most  species,  another  sparser  series  develops  along  the  ventral 
midline  of  the  gut,  from  the  isthmus  to  the  anus.  5".  boa  and  S. 
fero.x  develop  a  mid-lateral  series  of  melanophores  along  the 
body  and  S.  colubrimis  has  scattered  melanophores  along  the 
entire  hypaxial  body  region.  These  species  also  develop  exten- 
sive dorsal  and  lateral  head  pigment.  All  species  form  scattered 
pigment  on  the  dorsal,  anal,  and  caudal  fins. 

A  75-mm  specimen  (MCZ  Cat.  No.  59858)  with  an  extremely 
slender  body  form  (body  depth  1.3%  of  body  length)  has  fin  and 


172 


ONTOGEIVY  AND  SYSTEMATICS  OF  FISHES -AHLSTROM  SYMPOSIUM 


Table  45.     Photophore  Counts  of  Stomiatoid  Genera.  Most  frequent  count  or  range  is  followed  by  overall  range  or  infrequent  count  in 
parentheses.  Data  sources  as  in  Table  1.  Photophore  groups  as  defined  by  Morrow  (1964a). 


Photophore 

groups 

Family  and  genus 

IP 

PV 

vav 

AC 

ov 

VAL 

Stomiatidae 

Macrostomias 

11(12) 

80-86 

58-67 

19-22 

79-85 

58-68 

Stomias 

9-13 

32-51 

5-16 

14-20 

32-50 

4-17 

Chauliodontidae 

Chauliodus 

8-11 

17-23 

22-30 

8-13 

17-21 

22-29 

Astronesthidae 

Astronesthes 

5-12 

6-20 

7-27 

7-13 

5-19 

7-26 

Borostomias 

10-13 

20-31 

15-25 

9-15 

21-29 

16-25 

Heterophotus 

10-11 

32-35 

13-14 

12-15 

33-36 

16-20 

Neonesthes 

9-12 

14-17 

16-21 

13-18 

13-15 

13-21 

Rhadinesthes 

10(6) 

25(26) 

20-23 

16 

22-24 

27 

Melanostomiatidae 

Bathophiius 

5  (4-6) 

12-18 

11-13(11- 

17) 

5-7 (5-9) 

13-14(10-16) 

9-11  (8-17) 

Chirostomias 

9(8) 

25-27  (28) 

19-20(16) 

9(10) 

23  (24-25) 

19-20(16) 

Echiostoma 

8  +  2 

25-28 

14-18 

12-13(11) 

24-31 

13-17(18) 

Eustomias 

7-8  (9) 

27-33 (24-36) 

13-17(11- 

21) 

17-23(15-25) 

26-33  (24-37) 

13-18(12-22) 

Flagellostomias 

9-10(8) 

31-34 

14-16 

16-18(15) 

31-32(30) 

14-15(12-17) 

Grammatostomias 

7(6) 

15-18 

19-22 

10-13 

15-18 

19-22 

Leptostomias 

10(11) 

42-45  (39-48) 

20-23  (24) 

11-13(14) 

40-43  (39-48) 

20-22  (23-24) 

Metanostomias 

8  +  2  or  3 

23-30 

12-15 

9-11 

22-28 

11-15 

Opostomias 

4  +  4 

27 

17 

16 

27 

17 

Pachysiomias 

8-9 

14-16(17) 

13-14 

8-9 

17-18 

14-15 

Photonecies 

8-11 

19-24,  34-38 

11-15(16- 

18) 

10-13(9) 

19-24(17),  30-36 

11-14 (15-17) 

Tactostoma 

8 

46 

19 

12 

43 

18 

Thysanactis 

20 

31-32 

14-16 

11-12 

30-32 

14-16 

Trigonolampa 

11 

23-24 (22) 

22  (24) 

10-11 

22-24 

23-24  (26) 

Malacosteidae 

Aristoslomias 

5  +  3 

15-17(14-19) 

15-18 

9-11  (12) 

16-19(14-20) 

15-17(14-18) 

Malacosieus 

(Serial  photophores  absent  or  uncountable) 

Photostomias 

5  +  2 

13-16 

21-25 

12-15 

12-17 

20-23 

Idiacanthidae 

Idiacanlhus 

1P  +  PV  =  31-36 

16-18(15) 

13-18 

22-25 

31-35(30-36) 

vertebral  counts  that  match  Macrostomias  longibarbatus.  Its 
morphology  is  that  of  a  highly  attenuate  Stomias  larva.  Pig- 
mentation is  restiicted  to  a  series  of  small  melanophores  along 
the  ventral  midline  of  the  gut.  The  ventral  photophore  rows  are 
beginning  to  form. 

Chauliodontidae  (Fig.  59). —  Larvae  of  five  species  are  known 
(Table  46).  Larvae  are  6-7  mm  long  at  hatching,  with  an  elongate 
yolk  sac.  The  body  is  slender  with  a  circular  cross-section,  and 
remains  so  throughout  development.  The  head  is  relatively  small, 
with  elliptical  eyes  and  a  short,  acute  snout.  The  gut  has  a  smaller 
diameter  than  in  Stomias  but  is  relatively  longer.  The  short 
terminal  section  extends  beyond  the  anal  fin  origin.  The  median 
finfold  is  small  and  best  developed  rearward  on  the  body.  The 
dorsal,  anal,  and  pelvic  fins  form  in  late  postflexion  larvae  in 
the  adult  position.  A  fan-shaped  array  of  melanophores  occurs 
in  the  caudal  region  of  yolk-sac  larvae  but  is  soon  lost.  No  other 
pigment  develops.  Larvae  of  some  species  reach  46  mm  SL  and 
there  appears  to  be  marked  shrinkage  at  transformation. 


Astronesthidae  (Fig.  90).  — Astronesthid  larvae  have  been  illus- 
trated and  described  briefly  by  Roule  and  Angel  ( 1930),  Whitley 
(1941),  Pertseva-Ostroumova  and  Rass  (1973),  and  Belyanina 
(1982b);  however  only  two  of  these  were  identified  to  genus 
(Table  46).  We  have  examined  more  than  10  types  of  astro- 
nesthid  larvae,  7  of  which  are  listed  in  Table  46.  Astronesthid 
larvae  display  a  great  variety  of  structure  and  pigmentation,  but 
hold  in  common  the  advanced  position  of  the  dorsal  fin,  in 
contrast  to  other  Stomiatoidea,  except  Chauliodus.  The  types 
differ  fundamentally  in  gut  shape  and  body  form:  Types  I  and 
II  are  laterally  compressed,  relatively  deep-bodied,  and  have  a 
non-trailing  or  slightly  trailing  gut  with  terminal  section  as  in 
melanostomialids;  Types  III-VIl  have  a  slender  body  and  a 
trailing  gut;  in  Types  III-V  the  gut  is  deflected  ventrad  from 
the  body  just  anterior  to  the  anal  fin  base  and  in  Type  VI  and 
VII  at  midbody,  anterior  to  the  dorsal  fin  (Figure  90). 

Type  I  (Fig.  90A).  — larvae  up  to  26.5  mm;  laterally  compressed; 
head  shallow  with  acute  snout;  eyes  relatively  large,  slightly 


Fig.  90.     Larvae  of  Astronesthidae.  (A)  Type  I,  23.7  mm,  ORl  A105;  (B)  Type  II.  SIO  Tasaday  I  A3;  (C)  Type  IV.  33.0  mm.  MCZ  Cat.  No. 
59855;  (D)  Type  V,  22.0  mm,  Dana  Sta.  3931;  (E)  Type  VII,  28  mm,  MCZ  Cat.  No.  59856. 


KAWAGUCHI  AND  MOSER:  STOMIATOIDEA 


173 


174 


ONTOGE^Pr'  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  46.    Pigment  Characters  and  Gut  Structure  in  Larvae  and  Transforming  Specimens  of  Stomiatoidea.  (NT  =  not  trailing,  ST  = 

slightly  trailing,  T  =  trailing  freely). 


Hypaxial 

myoseptum 

Length 

Length  of 

Dorsal  myomere 

Epaxial  myoseptum 

melanophores 

of  larvae 

ti^nsfonning 

melanophores 

melanophores 

(no./myo- 

Gut 

Species 

(mm) 

specimens  (mm) 

(no./myomere) 

(no. /myoseptum) 

seplum) 

stnicture 

Source 

Stomiatidae 

Stomias  boa 

_ 

38 

0 

0 

0 

NT 

Sanzo,  1912a 

Stomias  boa 

10.4-30.4 

41.5 

0 

0 

0 

NT 

Sanzo,  193  Id 

Stomias  boa 

9.0-32 

— 

0 

0 

0 

NT 

Ege,  1918 

Stomias  ferox 

9.0-44 

— 

0 

0 

0 

NT 

Ege,  1918 

Stomias  colubrinus 

3.3-16 

— 

0 

0 

0 

NT 

Pertseva-Ostroumova 
and  Rass,  1973 

Stomias  alriventer 

4.6-32 

— 

0 

0 

0 

NT 

original 

Macrostomias  longibarbatus 

- 

75 

0 

0 

0 

NT 

original 

Chauliodontidae 

Chauliodus  sloani 

33.6 

41.6 

0 

0 

0 

NT 

Sanzo,  1915a 

Chauliodus  sloani 

5.7-41.6 

27.1 

0 

0 

0 

NT 

Sanzo,  193  Id 

Chauliodus  sloani 

2.1 

— 

0 

0 

0 

NT 

Mito,  1961a 

Chauliodus  danae 

22.5-25 

— 

0 

0 

0 

NT 

Belyanina.  1977 

Chauliodus  macouni 

38.0-49 

35-44 

0 

0 

0 

NT 

Belyanina,  1977 

Chauliodus  mmimus 

23.5-35 

— 

0 

0 

0 

NT 

Belyanina,  1977 

Chauliodus  pammelas 

10.6-40 

_ 

0 

0 

0 

NT 

Belyanina,  1977 

Chauliodus  sloani 

7.4-35 

27-34.2 

0 

0 

0 

NT 

Belyanina,  1977 

Chauliodus  macouni 

5.6-46 

- 

0 

0 

0 

NT 

original 

Astronesthidae 

Unidentified 

14.0-23 

— 

0 

0 

0 

T,  NT 

Roule  and  Angel,  1930 

Astronesthes  lupina 

20 

— 

0 

0 

0 

T 

Whitley,  1941 

Boroslomias  panamense 

5.0-17 

— 

0 

0 

0 

T 

Pertseva-Ostroumova 
and  Rass,  1973 

Unidentified 

16 

— 

7 

+ 

+ 

NT 

Belyanina,  1982b 

Unidentified 

17.7 

— 

2  total 

0 

0 

T 

Belyanina,  1982b 

Type  I 

12.3-26.5 

— 

several 

several 
to  many 

several 
to  many 

NT 

original 

Type  II 

14.9-26 

29,40 

0 

0 

0 

ST 

original 

Type  III 

16.2 

20.5,22.5 

0 

0 

0 

T 

original 

Type  IV 

14.4-34.5 

40.5 

0 

0 

0 

T 

original 

TypeV 

17.4-19.4 

20,22 

0 

0 

0 

T 

original 

Type  VI 

— 

28 

0 

0 

0 

T 

original 

Type  VII 

— 

28 

0 

0 

0 

T 

original 

Melanostomiatidae 

Tactostoma  macropus 

5.0-44 

49 

0-1 

0 

1-3 

NT 

original 

Melanostomias  spilorhynchus 

17 

21-32 

0 

ca.  3 

NT 

Beebeand  Crane,  1939 

Melanostomias  biseriatus 

— 

21-25 

0 

ca.  3 

NT 

Beebe  and  Crane,  1939 

Melanostomias  valdiviae 

— 

25 

0 

2-3 

NT 

original 

Melanostomias  sp. 

13.4-17.2 

16.4-22 

0 

2-4 

NT 

original 

Echiosloma  tanneri 

20,25 

— 

0 

2-5 

NT 

Beebe  and  Crane,  1939 

Echiostoma  sp.? 

13.8 

_ 

0 

2-4 

NT 

Belyanina,  1982b 

Echiostoma  barbatum 

— 

34 

0 

1-2 

NT 

original 

Photonectes  dinema 

— 

24  and  > 

1  (?) 

0 

3-4 

NT 

Beebe  and  Crane,  1939 

Photonectes  leucospilus 

— 

25  and  > 

1(?) 

0 

3-4 

NT 

Beebe  and  Crane,  1939 

Photonectes  albipinnis 

— 

16-22 

0 

2-3 

NT 

original 

Photonectes  sp. 

11.0-12.5 

— 

0 

4-5 

NT 

original 

Photonectes  parvimanus 

12.0-26 

25 

3-6 

0 

3-4 

NT 

Beebe  and  Crane,  1 939 

Photonectes  parvimanus 

27 

— 

3-4 

0 

2-4 

NT 

original 

Photonectes  parvimanus 

— 

28 

1-2 

0 

2-4 

NT 

original 

Photonectes  sp. 

5.4-22.2 

— 

ca.  7 

0 

5-7 

NT 

original 

Opostomias  mitsuii 

15.0-21 

1 

0-1  (2-3 
posteri- 
orly) 

1-2  (3-5 
post.) 

NT 

original 

Flagellostomias  boureei 

20.0-21 

34,39 

1 

0 

1 

NT 

Beebeand  Crane,  1939 

Flagellostomias  boureei 

10.8-36.4 

— 

1 

0 

1-2 

NT 

original 

Odontostomias  micropogon 

— 

42 

1 

1-? 

2-4 

NT 

Beebe  and  Crane,  1939 

Leptostomias  gladiator 

12.0-30 

38-45 

1  +  several 

1-5 

2-4 

NT 

Beebe  and  Crane,  1939 

Lepiostomias  gracilis 

— 

37.8 

1  +  1-5 

5-7 

6-9 

NT 

original 

Leptostomias  sp. 

25 

— 

1  +  1-3 

4-5 

4-6 

NT 

original 

Bathophilus  nigerrimus 

11.6 

21.7 

1  or  > 

0 

0 

NT 

Sanzo,  1915a 

Bathophilus  nigerrimus 

5.9,  14.0 

19.2-21.7 

1  or  > 

0 

0 

NT 

Sanzo,  193  Id 

KAWAGUCHI  AND  MOSER:  STOMIATOIDEA 


175 


Table  46.    Continued. 


Hypaxial 

myoseptum 

Length 

Length  of 

DoPial  myomere 

Epaxial  myoseptum 

melanophores 

of  larvae 

transforming 

melanophores 

melanophores 

(no./myo- 

Gut 

Species 

(mm) 

specimens  (mm) 

(no,  myomere) 

(no./myoseptum) 

septum) 

structure 

Source 

Bathophilus  metallicus 



25 

3  or  > 

0 

0 

NT 

Beebe  and  Crane,  1939 

Bathophilus  sp. 

11,  12 



1  or  > 

0 

0 

NT 

Beebe  and  Crane,  1939 

Bathophilus  sp. 

7 



1  or  > 

0 

0 

NT 

Beebe  and  Crane,  1939 

Bathophilus  sp. 

15 

_ 

(?) 

0 

0 

NT 

Rouleand  Angel.  1930 

Bathophilus  sp. 

18.2 

— 

1  or  > 

0 

0 

NT 

de  Sylva  and  Scotten, 
1972 

Bathophilus  filifer 

4-10 

- 

1  or  > 

0 

0 

NT 

Pertseva-Ostroumova 
and  Rass  1973 

Bathophilus  brevis 

15.7 

— 

1  or  > 

0 

0 

NT 

original 

Bathophilus  Jlemingi 

2.9-23.8 

— 

1  to  several 

0 

0 

NT 

original 

Euslomias  sp. 

33 

_ 

7  total 

0 

0 

T 

Regan,  1916 

Eustomias  sp. 

13 

— 

7  total 

0 

0 

T 

Beebe  and  Crane,  1939 

Eustomias  spp.  (4  types) 

6.0-45 

- 

5-1 1  total 

0 

0 

T 

original 

Malacosteidae 

Aristostomias  scintillans 

4.3-47 

45 

14  total 
to  many 

0 

0 

T 

original 

Photostomias  guernei 

20.0-27.5 

30,31 

8  pairs 
total 

0 

0 

T 

original 

Unidentified 

12 

_ 

1 2  total 

0 

0 

T 

Beebe  and  Crane,  1939 

Unidentified 

34.5 

- 

0 

0 

0 

T 

original 

Idiacanthidae 

Idiacanthus  fasciola 

16.0-28 

35-48 

0 

0 

1 

T 

Beebe,  1934 

Idiacanthus  sp. 

7.0-39 

— 

0 

0 

1 

T 

Pertseva-Ostroumova 
and  Rass,  1973 

Idiacanthus  antrostomus 

4.5-71 

67-> 

0 

0 

1 

T 

original 

elliptical;  gut  moderately  slender,  thin-walled;  finfold  moderate; 
pigment  pattern  consists  entirely  of  minute  melanophores,  in- 
creasing in  number  with  development,  principally  in  the  ex- 
paxial  and  hypaxial  myosepta;  other  pigment  above  brain,  paired 
internal  streaks  in  snout,  melanophores  in  dorsal  and  ventral 
finfold,  dorsal  fin  base,  and  on  posterior  half  of  gut. 

Type  II  (Fig.  90B).  — larvae  reach  at  least  26  mm;  deep-bodied 
and  laterally  compressed  in  late-stage  larvae;  head  deep;  eyes 
small,  slightly  elliptical;  gut  slightly  trailing  and  with  larger  di- 
ameter than  in  Type  I;  dorsal  finfold  relatively  deep;  pigment 
above  brain,  along  lower  jaw  and  at  angular  and  gular  region; 
blotch  at  posterior  margin  of  superior  hypural  complex  and  one 
midway  out  on  inferior  group  of  caudal  rays;  fin  ray  and  ver- 
tebral counts  and  photophore  counts  match  Astronesthes  gem- 
mifer. 

Type  III.  — larvae  reach  at  least  16.2  mm;  body  slender;  head 
and  eyes  moderate  in  size;  eyes  elliptical;  slender  gut  trails  free 
from  body  at  anal  fin  origin;  finfold  moderately  developed,  ex- 
cept posterior  to  dorsal  fin  the  finfold  appears  as  an  enlarged 
adipose  fin;  pigment  restricted  to  a  series  of  melanophores  along 
lower  jaw  and  between  upper  and  lower  hypural  complexes; 
counts  match  Astronesthes  richardsoni. 

Type  IV  (Fig.  90C).— lai^ae  reach  40  mm;  morphology  similar 
to  Type  III,  except  head  relatively  longer  and  eyes  almost  round; 
gut  with  leaf-like  appendages  on  trailing  section;  pigment  re- 
stricted to  postorbital  blotch  and  interorbital  band;  fin  and  ver- 
tebral counts  and  photophore  arrangement  match  Heterophotus. 


Type  V  (Fig.  90D).  — larvae  reach  about  20  mm;  morphology 
as  in  Types  III  and  IV;  eyes  slightly  elliptical;  pigment  heavy; 
melanophores  on  head,  lateral  to  posterior  brain  region,  on 
snout  and  lower  jaw  symphysis;  lateral  surface  of  body  covered 
with  an  irregular  pattern  of  large  melanophores;  melanophores 
on  trailing  gut.  Pertseva-Ostroumova  and  Rass  (1973)  identified 
larvae  of  this  type  as  Borostomias  panamense. 

Type  VI.  — specimen  transfoiming  at  28  mm;  morphology  sim- 
ilar to  Types  II-V,  except  trailing  gut  deflected  from  body  far 
in  advance  of  anal  fin  origin;  eyes  elliptical;  dorsal  finfold  highly 
developed  and  ventral  finfold  anterior  to  anal  fin  is  rudder-like; 
pigment  lacking;  meristics  indicate  it  is  in  the  genus  Astro- 
nesthes. 

Type  VII  (Fig.  90E).  — specimen  transforming  at  28  mm;  mor- 
phology similar  to  Type  VI;  dorsal  and  anal  fins  supported  on 
cartilaginous  pedestals;  a  series  of  4  melanophores  along  hori- 
zontal septum;  some  melanophores  on  anterior  region  of  dorsal 
and  anal  fin  bases  and  on  preanal  finfold.  Whitley  (1941)  de- 
scribed a  larva  similar  to  this  as  Astronesthes  lupina. 

Melanostomiatidae  (Figs.  91-92).  — Larvae  have  been  identified 
for  10  of  the  15  genera  (Table  46).  Bathophilus  was  the  first  to 
be  identified  (Sanzo,  1915a).  The  only  comprehensive  work  on 
melanostomiatid  ontogeny  is  that  of  Beebe  and  Crane  (1939) 
who  identified  larvae  of  8  genera  and  5  species  by  the  use  of 
transforming  series.  Since  then,  the  only  other  melanostomiatid 
larvae  that  have  been  described  are  Bathophilus  filifer  {Pertseva- 
Ostroumova  and  Rass,  1973),  Bathophilus  sp.  (de  Sylva  and 
Scotten,  1972),  and  Echiosloma  (?)  sp.  (Belyanina,  1982b).  De- 


176 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


scriptions  of  Opostomias  and  Tactostoma  are  included  in  this 
paper.  Larvae  of  Tactostoma  were  initially  identified  by  E.  H. 
Ahlstrom. 

Larval  representatives  of  the  10  genera  are  highly  various  in 
form  and  pigmentation,  however,  with  the  exception  of  Euslo- 
mias,  they  share  the  following  structural  features:  body  elliptical 
in  cross-section;  head  laterally  compressed;  eyes  small  and  el- 
liptical; gut  terminated  in  an  elongate  muscular  bulb  that  may 
extend  beyond  the  anal  fin  origin  but  not  beyond  the  margin  of 
the  finfold;  dorsal  and  anal  fins  form  in  adult  position  posteriorly 
on  the  body;  body  pigment  consists  of  one  or  more  melano- 
phores  dorsal  to  each  myomere,  one  or  more  melanophores  on 
the  hypaxial  myosepta  and,  in  some  genera,  on  the  epaxial  my- 
osepta.  Dorsal  and  lateral  pigmentation  tends  to  be  heavier  in 
forms  with  higher  meristic  counts.  The  genera  differ  principally 
in  body  size,  relative  body  depth,  relative  head  size,  jaw  size, 
gut  diameter,  size  and  shape  of  the  terminal  gut  section,  finfold 
height,  and  pigment  pattern. 

Present  knowledge  indicates  that  genera  apparently  have  dis- 
tinct facies,  tentative  descriptions  of  which  are  presented  below. 
Confirmation  awaits  identification  of  additional  species. 

Tactostoma  (Fig.  91  A). —  larvae  reach  44  mm  in  length;  body 
extremely  slender;  head  flat  and  elongate  initially,  becoming  less 
flat  and  relatively  smaller  with  development;  eye  size  moderate; 
gut  slender;  finfold  moderate;  pectoral  fin  lost  at  transformation; 
early  larvae  develop  one  melanophore  per  myomere  along  dor- 
sum and  1-3  melanophores  on  the  hypaxial  myosepta;  post- 
flexion  larvae  gradually  lose  the  dorsal  melanophores  and  then 
the  hypaxial  myosepta  pigment,  in  contrast  with  other  genera 
in  which  body  pigment  increases  with  development;  pigment 
on  lower  jaw  symphysis,  isthmus,  pectoral  fin  base,  cleithrum, 
and  above  gut  terminus;  dorsal  and  ventral  pigment  accentuated 
at  caudal  peduncle. 

Melanostomias  (Fig.  91B).— transforming  specimens  as  small 
as  16.4  mm;  body  slender;  head  small;  snout  short;  eye  size 
moderate;  gut  slender;  finfold  relatively  small;  one  melanophore 
per  myomere  along  dorsum  in  one  form  and  in  another  form 
the  zone  between  the  7th- 10th  myomere  and  the  dorsal  fin  lacks 
dorsal  pigment;  2-3  melanophores  in  hypaxial  myosepta;  pig- 
ment above  and  below  head,  below  liver,  on  terminal  gut  sec- 
tion, and  along  finfold  margins.  Larvae  tentatively  identified  as 
Echiostoma  have  similar  characters  (Table  46). 

Photonectes  (Fig.  9 IC).  — larvae  of  different  forms  transform  at 
sizes  between  16  and  28  mm;  body  somewhat  deep;  head  size 
and  snout  length  moderate;  eyes  small,  highly  elliptical;  several 
forms  of  dorsal  myomere  pigment  ( 1  melanophore  per  myomere 
in  Subgenus  Photonectes  and  3-7  per  myomere  in  Subgenus 
Trachinostomias);  hypaxial  myosepta  with  2-7  melanophores 
depending  on  form  (Table  46);  extensive  pattern  of  minute  me- 
lanophores on  head,  finfold,  and  median  fins. 

Flagellostomias (¥\g.  9  ID).  — larvae  may  reach  30-40  mm;  body 
somewhat  deep;  head  large,  deep,  with  steeply  sloping  snout  and 


large  jaws;  eyes  small;  gut  diameter  relatively  large;  finfolds 
large,  accentuating  body  depth;  one  large  melanophore  per  myo- 
mere along  dorsum;  1-3  melanophores  in  hypaxial  myosepta; 
some  scattered  lateral  melanophores  in  median  fin  region;  other 
pigment  scant;  a  few  melanophores  in  head  region,  some  on 
finfold  in  posterior  gut  region,  and  on  dorsal  and  anal  fins. 

Opostomias  (Fig.  9 IE).  — body  moderately  deep;  head  large,  deep 
posteriorly  with  elongate  sloping  snout;  eyes  small;  gut  slender; 
finfold  large;  one  melanophore  per  myomere  along  dorsum;  1- 
2  melanophores  in  hypaxial  myosepta;  epaxial  and  hypaxial 
myosepta  below  dorsal  fin  base  have  several  melanophores, 
giving  this  region  a  banded  appearance;  melanophores  on  dorsal 
head  region,  gill  arch  and  gut  terminus. 

Leplostomias  (Fig.  91F).  — larvae  may  reach  about  40  mm;  body 
somewhat  deep;  head  moderately  large,  deep;  eyes  small;  gut 
slender;  finfold  moderate;  pigmentation  heavy;  one  large  me- 
lanophore and  1-5  smaller  ones  per  myomere  along  dorsum; 
numerous  melanophores  on  epaxial  and  hypaxial  myosepta, 
increasing  with  development  to  completely  outline  myosepta; 
pigment  extensive  on  dorsal  and  ventral  head  regions,  on  gill 
arches;  pigment  below  liver,  on  finfold  margins,  above  gut  ter- 
minus and  on  dorsal  and  anal  fins. 

Bathophilus  (Figs.  92A-C).  — larvae  transform  at  25  mm  or  less; 
deep-bodied  compared  with  other  genera;  head  and  jaws  large; 
barbel  forms  in  late  postflexion  larvae,  particularly  in  B.  hrevis; 
eye  size  moderate;  gut  large  to  voluminous,  with  highly  devel- 
oped s-shaped  terminal  section;  finfolds,  particularly  dorsal,  large; 
one  or  several  melanophores  per  myomere  along  dorsum  and 
an  opposing  series  of  melanophores  along  ventral  surface  of 
myomeres;  no  lateral  pigment;  head,  finfolds  and  median  fins 
pigmented. 

Eustomias  (Fig.  92D).  — larvae  of  some  species  reach  45  mm; 
body  slender,  and  round  in  cross-section;  head  elongate  and  flat 
with  large  spatulate  snout;  large  jaws;  eyes  moderate  in  size, 
slightly  elliptical  to  round;  gut  slender,  deflected  ventrad  at  anal 
fin  origin  and  trailing  from  body;  body  pigment  consists  of  5- 
1 1  large  melanophores  along  the  dorsal  midline;  usually  pigment 
at  lower  jaw  symphysis. 

Malacosteidae  (Fig.  9iA  — Larvae  of  this  group  have  not  been 
described,  although  the  12-mm  larva  illustrated  by  Beebe  and 
Crane  (1939)  and  referred  to  "lEustomias"  is  apparently  Ar- 
istostomias.  We  have  examined  larval  series  and  transforming 
specimens  of  A.  scintillans  and  Photostomias  guernei  (Table  46). 

Aristostomias  scintillans  (Fig.  93A). —larvae  reach  47  mm  length; 
body  slender;  head  large,  flat;  snout  elongate;  jaws  large;  eyes 
slightly  elliptical;  opercle  markedly  reduced;  gut  slender,  de- 
flected ventrad  at  anal  fin  origin  and  trailing  from  body;  finfold 
moderate;  dorsal  and  anal  fins  form  in  adult  position  at  about 
flexion  stage;  pelvics  form  late;  initial  pigment  pattern  is  a  series 
of  paired  melanophores  along  the  dorsum,  beginning  with  14 


Fig.  91.  Larvae  of  Melanostomiatidae.  (A)  Tactostoma  macropiis,  CalCOFI  Norpac  Sta.  14;  (B)  Melanostomias  sp.,  16.0  mm,  ORI  KH73- 
2,  Sta.  49-7;  (C)  Photonectes  sp.,  22.2  mm,  SWFC,  Albacore  Oceanography  Cruise  71,  Sta.  99;  (D)  Ftagetloslomias  boureii.  36.4  mm.  SIO  Cat. 
No.  73-329,  Tasaday  I,  Tow  42;  (E)  Opostomias  mitsiiii.  1  5.0  mm,  ORI  KH  73-2  Sta.  2-3;  (F)  Leptoslomias  sp.,  24.5  mm,  MCZ  Cat  No.  59857. 


KAWAGUCHI  AND  MOSER:  STOMIATOIDEA 


177 


178 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


KAWAGUCHI  AND  MOSER:  STOMIATOIDEA 


179 


Fig.  93.     Larvae  of  Malacosteidae.  (A)  Aristostomias  scinlillans.  34.7  mm.  CalCOFI  5008  Sta.  70.30;  (B)  Photostomias  sp.,  26.7  mm.  ORI  KH 
73-5  Sla.  55-13.  Bn  24-12;  (C)  Malacosteidae,  34.5  mm.  from  Moser  (1981). 


Fig.  92.     Larvae  of  Melanostomiatidae.  (A)  Bathophilus  flemingi.  25.5  mm.  CalCOFI  4910.  Sta.  80.137;  (B)  B   hrevis.  15.7  mm.  ORI  KH 
81-1,  Sta.  17;  (C)  B.  nigernmus,  21.7  mm,  redrawn  from  Sanzo  (1931d);  (D)  Eustomias  sp.  33  mm,  redrawn  from  Regan  (1916). 


180 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  94.     Larva  of  Idiacanlhus  anirostoinus.  55  mm.  CalCOFI  6207  Sta.  90.120. 


KAWAGUCHI  AND  MOSER:  STOMIATOIDEA 


181 


pairs  and  increasing  in  numbers  with  development  to  cover  the 
entire  dorsum;  paired  ventral  series  develop,  initially  poste- 
riorly, and  increase  in  numbers  so  that  all  myomeres  have  me- 
lanophores  on  the  ventral  surface;  pigment  on  brain,  snout, 
lower  jaw,  gular-isthmus  region,  otic  region,  caudal  fin,  and  in 
vague  rings  along  trailing  gut.  Ahstostomias  larvae  were  iden- 
tified initially  by  E.  H.  Ahlstrom. 

Photostomias  giternei  (Fig.  93B).  — larvae  reach  about  30  mm; 
morphology  similar  to  A.  scintillans  except  eyes  smaller  and 
narrower  and  pelvic  fins  somewhat  elongate;  body  pigment  con- 
sists of  a  series  of  8  minute  dorsal  melanophore  pairs  and  8 
slightly  larger  opposing  pairs  along  the  ventral  surfaces  of  the 
myomeres;  melanophores  at  lower  jaw  symphysis,  large  mela- 
nophore on  each  pectoral  fin  base,  sparse  melanistic  rings  along 
trailing  gut. 

Malacosteid  C  (Fig.  93C).  — intact  specimen  (captured  by  Dr. 
Richard  Harbison,  WHOl)  has  morphological  and  meristic 
characters  of  malacosteid  larvae  but  lacks  pigment  except  on 
the  extensive  gut.  Shallow  capture  locality  of  this  specimen  and 
our  capture  of  large  A.  scintillans  larvae  in  MANTA  nets  in- 
dicates late-stage  malacosteid  larvae  have  a  shallow  distribution 
in  the  water  column. 

Idiacanihidae  (Fig.  94j.  —  Brauer  ( 1 906,  1 908)  described  the  re- 
markable larvae  of  Idiacanthus  and  named  them  Slylophthal- 
mus paradoxus.  Beebe  ( 1 934)  correctly  identified  the  larvae  and 
described  them  in  detail.  Idiacanthus  larvae  are  extremely  slen- 
der, reaching  a  length  of  35-70  mm  depending  on  the  species. 
Other  characteristics  are:  elongate  and  extremely  flat  head;  el- 
liptical eyes  on  long  stalks  with  cartilaginous  supporting  rods; 
stalk  length  up  to  27%  of  body  length  in  /.  antrostonms  (Weihs 
and  Moser,  1981);  gut  slender,  deflected  at  anal  fin  origin  and 
trailing;  finfold  small;  dorsal  fin  begins  forming  in  preflexion 
larvae;  dorsal  fin  larger  than  anal  fin  and  slightly  in  advance  of 
it  in  postfiexion  larvae;  during  transformation,  rays  added  se- 
quentially anteriad  so  that  in  adults  the  dorsal  extends  about  -A 
of  the  body  length  and  the  anal  about  'A;  pectoral  fins  well 
developed  but  lost  at  transformation  and  pelvic  fins  develop  in 
transforming  females,  but  not  at  all  in  males;  pigment  pattern 
consists  of  a  melanophore  on  the  posterior  margin  of  each  hy- 
paxial  myomere,  spreading  into  the  myosepta  when  expanded, 
several  elongate  internal  blotches  in  the  isthmus  region,  and  a 
series  of  melanophores  along  the  trailing  gut;  adult  males  of  /. 
fasciola  reach  32-42  mm  SL,  lack  teeth  and  paired  fins  and  have 
relatively  larger  eyes  and  an  enormous  luminous  gland. 


Relationships 

Information  on  larval  characters  of  18  of  the  26  stomiatoid 
genera  recognized  by  Fink  (this  volume),  representing  all  6  of 
the  families  recognized  by  Weitzman  (1974),  permits  some  pre- 
liminary generalizations  and  conclusions:  (1)  Larvae  of  Sto- 
miatidae  and  Chauliodontidae  are  similar  in  morphology  and 
are  distinct  from  other  stomiatoids.  Pigmentation  provides  fur- 
ther evidence  of  this;  Chaidiodus  larvae  are  unique  among  known 
stomiatoids  in  lacking  pigment  after  the  yolk-sac  stage  and  the 
median  series  of  gut  melanophores  of  Stomias  also  appear  to 
be  unique.  (2)  Larvae  of  Astronesthidae  are  diverse  in  mor- 
phology and  pigmentation  and  most  of  the  larval  specializations 
that  appear  in  other  stomiatoid  families  are  found  among  as- 
tronesthid  genera.  Larval  specializations  of  some  genera  (e.g., 
ornamented  trailing  gut,  trailing  gut  deflected  at  mid-body,  rud- 
der-like finfolds)  are  not  found  elsewhere  in  Stomiatoidea.  Het- 
erogeneity of  larval  characters  in  Astronesthidae  supports  Fink's 
view  that  the  group  is  paraphyletic.  (3)  In  the  Melanostomia- 
tidae,  larvae  of  Melanostomias.  Photonectes.  Echiostoma. 
Oposlomias.  Flagellostomias.  Odontostomias  and  Leptostomias 
are  similar  in  morphology,  have  paired  melanophore  series  on 
the  dorsum,  and  differ  chiefly  in  head  size,  body  depth,  and  in 
the  extent  of  myosepta  pigment.  Tactostoma  larvae  have  the 
characters  of  this  group  of  genera  except  that  the  body  is  ex- 
tremely slender  and  the  pigmentation  is  lost  in  the  postfiexion 
stage.  Larvae  of  Battiophilus  difler  from  those  of  the  above  group 
in  a  number  of  characters  (voluminous  gut  with  specialized 
terminal  section,  melanophore  series  on  the  ventral  surface  of 
the  myomeres,  lack  of  myosepta  pigment).  Larvae  of  Eustomias 
are  different  from  all  known  larvae  of  Melanostomiatidae  in 
having  a  trailing  gut,  flat  head  and  snout,  and  a  pigment  pattern 
consisting  of  a  median  series  of  up  to  11  large  melanophores 
on  the  dorsum.  Except  for  this  latter  feature,  Eustomias  larvae 
are  similar  to  those  of  Malacosteidae.  (4)  Idiacanthus  larvae 
have  a  combination  of  characters  unique  among  stomiatoids. 
The  stalked  eyes  are  autapomorphic.  Larval  characters  provide 
no  support  for  Fink's  hypothesis  that  this  genus  is  closely  related 
to  Tactostoma. 

Ocean  Research  Institute,  University  of  Tokyo,  1-15-1, 
MiNAMiDAi,  Nakano-ku,  Tokyo  164,  Japan,  and  Na- 
tional Marine  Fisheries  Service,  Southwest  Fisheries 
Center,  8604  La  Jolla  Shores  Drive,  La  Jolla,  Calif- 
ornia 92038. 


Stomii  forms:  Relationships 

W.  L.  Fink 


STOMIIFORMS  are  well  known  as  a  major  component  of 
the  midwater  oceanic  fauna.  Past  concepts  of  their  rela- 
tionships to  other  primitive  euleleosts  were  reviewed  by  Fink 
and  Weitzman  (1982),  but  in  brief  in  this  century,  they  have 


been  considered  isospondyls  (Parr,  1927;  Regan,  1923;  Morrow, 
1964)  or,  more  recently,  salmoniform  protacanthopterygians 
(Greenwood  et  al.,  1966).  In  1973,  Rosen  placed  these  fishes  as 
a  separate  order  (Stomiatiformes)  within  the  Neoteleostei,  as 


182 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


STERNOPTYCHIDAE 

GONOSTOMATIDAE 

PHOTICHTHYIDAE 

ASTRONESTHIDAE 

IDIACANTHIDAE 


MALACOSTEIDAE 


MELANOSTOMIIDAE 


STOMIIDAE 


CHAULIODONTIDAE 


Fig.  95.  Weitzman's  (1974)  hypothesis  of  relationships  of  the  sto- 
miiform  fishes.  The  Gonostomatidae  and  Stemoptychidae  comprise  the 
Gonostomata  and  the  remaining  families  comprise  the  Photichthya. 


sister  group  to  the  Eurypterygii.  Fink  and  Weitzman  (1982) 
agreed  with  this  placement,  provided  more  characters  to  sub- 
stantiate it,  and  demonstrated  monophyly  of  the  stomiiforms. 
Steyskal  (1980)  has  presented  arguments  that  the  root  of  the 
family-group  names  demands  that  these  be  altered  from  Sto- 
miatidae  and  Stomiatiformes  to  Stomiidae  and  Stomiiformes, 
respectively,  and  I  use  these  forms  throughout  this  paper. 

As  recognized  by  Weitzman  (1974),  there  are  two  major  sto- 
miiform  lineages,  Gonostomata  and  Photichthya,  both  classified 
at  infraordinal  rank,  with  families  Gonostomatidae  and  Ster- 
noptychidae  in  the  former  and  families  Photichthyidae,  Sto- 
miidae, Chauliodontidae,  Astronesthidae,  Melanostomiidae, 
Malacosteidae,  and  Idiacanthidae  in  the  latter  (Fig.  95).  I  have 
no  disagreement  with  Weitzman's  hypotheses  of  monophyly  of 
the  Stemoptychidae,  but  our  recent  work  on  Diplophos  (Fink 
and  Weitzman,  1982)  caused  us  to  question  the  monophyly  of 
the  Gonostomatidae  and  Photichthyidae,  and  my  work  on  the 
barbelled  stomiiforms,  comprising  the  remaining  families,  has 
cast  doubt  on  the  entire  traditional  arrangement  of  the  included 
26  genera  as  well  as  on  the  monophyly  of  the  Photichthya.  I 
have  found  features  which  support  new  hypotheses  of  relation- 
ship within  the  stomiiforms  and  will  present  some  of  these  ideas 
below.  Some  are  more  tentative  than  others.  Weitzman  is  cur- 
rently working  on  the  genera  he  placed  in  the  Gonostomatidae 
and  Photichthyidae. 

First,  I  have  found  no  evidence  that  Diplophos  is  the  sister 
group  of  any  other  genus  of  stomiiform  and  it  may  be,  as  Fink 
and  Weitzman  (1982)  suggested,  the  sister  group  of  the  rest  of 
the  order.  Specializations  in  the  adductor  muscles  indicate  that 


Diplophos 


GONOSTOMA 


Cyclothone 


Margrethia 


BONAPARTIA 


Triplophos 


STERNOPTYCHIDAE 


PHOTICHTHYA 


Fig.  96.     Hypotheses  of  stomiiforms  as  discussed  herein.  See  text  for 
explanation. 


Gonostoma.  Cyclothone,  Margrethia.  and  Bonapartia  form  a 
monophyletic  group,  but  what  relationships  within  that  group 
are  I  cannot  say,  and  presumably  this  will  be  treated  by  Weitz- 
man. These  hypotheses  would  cause  a  redefinition  of  the  Gon- 
ostomatidae, restricting  it  to  the  four  genera  mentioned  just 
above.  Relationships  of  Triplophos  are  also  unclear,  and  there 
is  evidence  in  the  hyoid  apparatus  that  it  may  be  related  to  some 
of  the  "photichthyans,"  rather  than  the  gonostomatids,  as 
Weitzman  ( 1 974)  supposed.  Weitzman  ( 1974)  established  mon- 
ophyly of  the  Stemoptychidae,  and  1  have  nothing  to  add  to  his 
conclusions.  Nevertheless,  since  he  did  not  deal  with  monophyly 
of  the  Gonostomatidae  or  with  the  sister  group  relationship  of 
the  Stemoptychidae,  there  is  no  current  evidence  that  the  latter 
is  more  closely  related  to  some  subset  of  the  former,  and  I  leave 
that  part  of  the  phylogeny  unresolved.  These  hypotheses  are 
summarized  in  Fig.  96.  See  also  the  paper  by  Ahlstrom,  Rich- 
ards, and  Weitzman  (this  volume)  on  the  Gonostomatidae,  Ster- 
noptychidae  and  other  stomiiforms. 

Within  the  "Photichthya,"  we  have  the  same  problem  as  with 
the  Gonostomatidae;  that  is,  there  is  a  diagnosable  monophy- 
letic unit  (the  barbelled  forms)  and  an  undiagnosed  grade  group, 
the  Photichthyidae. 

My  own  efforts  have  been  on  the  barbelled  forms,  currently 
distributed  in  six  families,  as  listed  above.  There  have  been  no 
strictly  phylogenetic  studies  of  relationships  within  the  group, 
but  they  were  examined  in  a  traditional  sense  by  Parr  (1927), 
Regan  and  Trewavas  ( 1 929,  1 930),  and  Beebe  and  Crane  (1939). 


FINK:  STOMIIFORMS 


183 


My  hypotheses  are  based  on  a  study  of  330  characters,  mostly 
taken  from  the  skeleton,  but  with  some  from  the  head  muscles, 
photophores,  and  other  parts  of  the  soft  anatomy.  The  conclu- 
sions are  presented  in  Fig.  97.  Traditional  families  are  not  rec- 
ognizable in  this  scheme  of  relationships. 

Evidence  for  the  arrangement  of  the  genera  is  presented  else- 
where (Fink,  in  prep.),  but  some  characters  will  be  discussed 
below,  particularly  those  relevant  to  some  of  the  larger  portions 
of  the  tree  or  in  areas  that  might  seem  controversial  to  some 
readers.  For  ease  of  communication,  I  will  state  here  that  my 
choice  of  classification  for  this  group  is  an  expansion  of  the 
traditional  Stomiidae  of  Regan  and  Trewavas  (see  Fig.  97). 

Monophyly  of  the  Stomiidae  is  established  on  the  basis  of  up 
to  1 7  characters,  including  1 )  presence  of  a  mental  barbel,  2)  5 
hypurals  in  the  caudal  skeleton  rather  than  6,  3)  lack  of  gill 
rakers  in  adults,  4)  a  divided  geniohyoideus  muscle,  and  5)  a 
portion  of  the  adductor  mandibulae  inserting  on  the  postorbital 
photophore. 

The  Astronesthidae,  as  most  recently  discussed  by  Weitzman 
(1967),  consisted  oi  Astronesthes.  Boroslomias.  Heterophotus. 
Neoncsthcs,  and  Rhadinesthes.  As  can  be  seen  in  Fig.  97,  the 
group  is  clearly  not  monophyletic.  Neonesthes  is  the  sister  group 
of  all  other  stomiids,  a  hypothesis  borne  out  by  many  characters 
shared  by  the  remaining  stomiid  genera,  including  lack  of  tooth- 
plates  on  basibranchial  1,  epibranchial  4,  and  on  the  posterior 
edges  of  gill  arches  1-4,  and  presence  of  rector  muscles  attaching 
to  the  fifth  ceratobranchial.  The  several  equally  parsimonious 
constructions  of  stomiid  relationships  leave  an  unresolved  tri- 
chotomy at  the  next  level,  there  being  insufficient  evidence  re- 
garding the  positions  of  Aslronesthes.  Boroslomias.  and  the  re- 
maining stomiids.  This  problem  will  be  further  discussed  by 
Fink  (in  prep.). 

The  remaining  stomiids  are  united  by  such  traits  as  lack  of 
toothplates  on  basibranchial  3  and  position  of  the  basihyal- 
hypohyal  ligament,  as  well  as  specializations  of  the  dorsal  and 
anal  fin  skeletons.  At  this  point  there  lies  another  unresolved 
trichotomy,  involving  the  groups  Heterophotus  plus  Rhadi- 
nesthes, Slomias  plus  Chauliodus,  and  the  remaining  stomiids. 
Heterophotus  and  Rhadinesthes  are  documented  as  sister  taxa 
by  several  characters,  including  an  elongate  dorsal  spine  on  the 
cleithrum  and  a  preopercle  that  is  narrow  at  the  area  of  the 
symplectic-hyomandibular  joint.  That  Chauliodus  and  Stomias 
are  sister  taxa  is  supported  by  numerous  characters,  including 
a  nasal  bone  which  forms  a  cup-like  wall  to  the  nasal  capsule; 
distribution  of  the  palatine  teeth  into  two  areas,  one  anterior 
and  one  well  posterior;  branchiostegals  deeply  bifurcated  dor- 
sally;  and  a  distinct  hexagonal  pigment  pattern  in  the  skin.  I  do 
not  recognize  the  genus  Macrostomias  since  work  in  progress 
shows  that  those  species  are  the  sister  group  to  a  derived  group 
within  Stomias. 

The  remaining  genera,  comprising  the  traditional  families 
Melanostomiidae.  Malacosteidae,  and  Idiacanthidae,  are  united 
by  presence  of  many  features,  including  no  more  than  one  pair 
of  toothplates  associated  with  any  basibranchial  ossification, 
and  reduction  of  the  distal  radials  of  the  pectoral  fins. 

As  postulated  by  Regan  and  Trewavas  (1930),  I  have  also 
found  that  Chirostomias  and  Tngonolampa  are  sister  taxa  based 
on  features  such  as  fusion  of  the  bilateral  toothplates  of  basi- 
branchials  2  and  3  and  reduction  of  the  supramaxiUa  to  a  sliver 
of  bone.  These  genera  are  the  sister  group  to  the  remaining 
genera,  a  hypothesis  supported  by  several  characters,  including 
fewer  than  6  branchiostegals  articulating  with  the  posterior  cer- 


Neonesthes 

astronesthes 

borostomias 

Heterophotus 

Rhadinesthes 

Chauliodus 

Stomias 

Chirostomias 

Trigonolampa 

Thysanactis 

leptostomias 

Opostomias 

Odontostomias 

Flagellostomias 

Photonectes 

Echiostoma 

Melanostomias 

Idiacanthus 

Tactostoma 

Grammatostomias 

Bathophilus 

eustomias 

Aristostomias 

Malacosteus 

Pachystomias 

Photostohias 

Fig.  97.     Hypothesis  of  relationships  within  the  Stomiidae,  as  dis- 
cussed herein. 


atohyal  ossification,  3  or  fewer  distal  pectoral  fin  radials,  and 
presence  of  a  modification  of  the  anterior  pectoral  fin  rays  into 
a  structure  I  call  the  "rod-ray  complex." 

For  the  remaining  genera,  I  will  concentrate  on  establishing 
the  major  lineages  as  monophyletic  and  on  areas  that  affect 
traditional  familial  classifications  of  the  group,  particularly  the 
relationships  of  the  "malacosteids"  and  Idiacanthus. 

One  monophyletic  group  is  comprised  of  Flagellostomias. 
Leptostomias.  Odontostomias.  Opostomias.  and  Thysanactis. 
Among  the  diagnostic  features  are  fusion  of  the  distal  cartila- 
ginous tips  of  the  lateral  ethmoid  and  supraethmoid,  and  an 
elongate  opercular  process  of  the  hyomandibula. 

The  remaining  genera  are  supported  as  monophyletic  by  nu- 
merous characters,  among  them  being  lack  of  a  retroarticular 
(also  lacking  in  Trigonolampa),  and  the  form  of  the  articulation 
of  the  interhyal.  The  latter  element  articulates  anterior  to  the 
front  margin  of  the  cartilage  between  the  hyomandibula  and 
symplectic  and  is  bound  to  the  metapterygoid  by  a  ligament 
from  the  anterior  margin  of  the  interhyal. 

The  Malacosteidae  has  traditionally  been  comprised  of  three 
genera,  Aristostomias.  Malacosteus.  and  Photostomias,  all  of 
which  lack  a  floor  to  the  mouth.  The  evidence  shows  that  Pachy- 


184 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES -AHLSTROM  SYMPOSIUM 


stomias  also  belongs  to  this  group,  and  not  with  the  other  "me- 
lanostomiids."  This  finding  is  not  particularly  radical,  since 
other  authors  have  noted  the  close  morphological  resemblance 
of  that  genus  to  the  other  three  and  indeed,  it  has  been  kept  out 
of  the  Malacosteidae  mostly  because  the  mouth  floor  is  still 
present,  though  thin,  in  members  of  the  genus.  The  data  are 
insufficient  to  allow  an  unambiguous  resolution  of  the  interre- 
lationships of  these  genera,  but  numerous  characters  support 
the  monophyly  of  the  assemblage,  including  the  suborbital  pho- 
tophore  being  ventral  or  posteroventral  to  the  eye  and  the  car- 
tilage of  the  palatine  arch  being  interrupted  between  the  pos- 
terior margin  of  the  palatine  and  the  rest  of  the  arch. 

Idiacanthus  has  usually  been  placed  in  a  family  by  itself  as 
was  done,  for  example,  by  Beebe  (1934),  primarily  on  the  basis 
of  the  specialized  stalked-eyed  larval  stages  and  the  degree  of 
sexual  dimorphism.  Beebe  recognized  that  the  genus  was  "closely 
related  to  the  Melanostomiatidae,"  as  did  Gibbs  (1964b).  Nei- 
ther author  suggested  more  precise  relationships,  and  Beebe  and 
Crane  (1939)  showed  Idiacanthus  in  a  large  multichotomy  in 
their  figure  of  "relationships."  Regan  and  Trewavas  ( 1 930)  con- 
sidered Idiacanthus  to  belong  with  Melanostomias,  Echiosto- 
ma.  and  Photonectes.  but  did  not  say  precisely  where.  My  data 
support  placement  of  the  genus  as  sister  group  to  Tactostoma, 
a  genus  described  in  1 939.  These  two  are  then  related  to  a  group 
of  genera  as  shown  in  Fig.  97.  Note  that  Melanostomias  and 
Echiostoma  are  excluded,  being  the  sister  group  of  the  entire 
assemblage.  I  am  confident  of  the  placement  of  Idiacanthus  and 
Tactostoma  together,  based  on  an  array  of  characters,  including 
reduction  of  the  basihyal  to  a  thin,  cylindrical  element,  origin 
of  the  dorsal  section  of  the  medial  division  of  the  adductor 
mandibulae  muscle  anterior  to  the  insertion  of  the  levator  arcus 
palatini  muscle,  and  an  extremely  elongate  body.  But  I  am  not 
particularly  confident  in  the  placement  of  these  two  genera  with 
the  others,  even  though  the  data  appear  impressive  at  first  glance. 
This  lack  of  confidence  is  attributable  to  the  fact  that  most  of 
those  characters  change  at  least  three  times  in  the  entire  tree, 
leaving  but  one,  lack  of  a  posttemporal  bone,  as  the  only  un- 
reversed character  supporting  the  hypothesis. 

Another  possibility  is  that  Idiacanthus  and  Tactostoma  are 
the  sister  group  of  Melanostomias  and  Echiostoma.  as  suggested 
in  part  by  Regan  and  Trewavas  (1930),  apparently  based  on  the 
close  morphological  resemblance  of  Idiacanthus  with  the  latter 
two  genera.  Such  a  hypothesis  would  require  some  additional 
reversals  or  independent  losses,  but  as  just  noted,  most  of  these 


characters  change  several  times  even  in  the  most  parsimonious 
tree.  This  part  of  the  total  phylogeny  deserves  more  critical 
examination,  and  it  is  hoped  that  larval  specializations  will  be 
found  which  will  be  found  which  will  cause  one  hypothesis  to 
be  clearly  preferred  over  the  other. 

Regarding  classification  of  the  stomiiform  fishes,  it  appears 
that  most  of  the  traditional  groups  will  cease  to  be  recognized, 
a  move  that  was  initiated  by  Weitzman  (1974).  A  period  of  flux 
should  be  expected  until  his  curtent  work  is  completed,  but  such 
temporary  instability  is  the  current  state  of  teleostean  classifi- 
cation at  all  levels,  as  phylogenetic  methodology  is  applied  with 
increasing  frequency.  One  might  expect,  however,  that  classi- 
fication within  the  Stomiiformes  will  be  stable  sooner  than  that 
in  many  other  groups,  because  phylogenetic  methods  already 
have  been  applied  to  it  for  several  years.  I  will  not  present  a 
classification  here,  but  I  do  provide  such  for  the  Stomiidae  in 
my  revision  of  the  group  (Fink,  in  prep). 

In  summary,  there  is  still  much  to  be  done  in  unravelling  the 
phylogenetic  history  of  the  main  lineages  of  stomiiform  fishes. 
1  have  outlined  above  areas  where  our  knowledge  is  either  in- 
complete or  poorly  developed,  and  these  should  be  the  areas 
where  workers  now  concentrate  their  attention— to  establish 
monophyletic  groups  among  the  "primitive"  stomiiforms  and 
to  critically  reexamine  some  of  the  hypotheses  I  have  produced 
within  the  barbelled  stomiiforms.  Some  of  this  work  is  under- 
way, using  adult  and  sub-adult  specimens,  but  the  usefulness  of 
larvae  is  as  yet  unknown.  The  data  presented  in  Ahlstrom's 
(1974)  work  on  patterns  of  metamorphosis  in  "gonostomatid" 
fishes  corroborate,  when  analyzed  by  phylogenetic  methods,  the 
placement  by  Weitzman  (1974)  of  many  of  those  genera  in  an 
expanded  Stemoptychidae.  An  example  of  this  is  the  presence 
of  photophores  in  clusters  with  common  bases  in  those  fishes 
recognized  by  Weitzman  as  stemoptychids.  Kawaguchi  and 
Moser  (this  volume)  present  the  most  comprehensive  infor- 
mation to  date  of  stomiid  larvae.  Their  data  indicate  that  there 
should  be  a  plethora  of  characters  for  phylogenetic  analysis  and 
that  study  of  larvae  should  indeed  prove  useful  in  testing  hy- 
potheses of  stomiid  relationships.  However,  even  a  cursory  ex- 
amination of  their  data  indicates  that,  as  with  characters  in 
adults,  there  appears  to  be  a  high  degree  of  homoplasy.  This  is 
an  interesting  phenomenon  deserving  further  study. 

Museum  of  Zoology,  University  of  Michigan,  Ann  Arbor, 
Michigan  48109. 


Families  Gonostomatidae,  Stemoptychidae,  and  Associated  Stomiiform  Groups: 

Development  and  Relationships 

E.  H.  Ahlstrom,  W.  J.  Richards  and  S.  H.  Weitzman 


A  summary  of  known  information  about  the  larvae  and  re-  formation,  both  published  and  unpublished,  gleaned  from  early 
lationships  of  the  stomiiforms  with  elongate  gill  rakers  in  life  history  stages  and  from  adults.  We  also  append  some  ten- 
adults  was  published  by  Ahlstrom  (1974).  The  present  paper  is  tative  new  hypotheses  of  relationships  within  this  "group"  of 
an  addendum  to  that  contribution  and  includes  additional  in-  stomiiforms. 


AHLSTROM  ET  AL.:  GONOSTOMATIDAE,  STERNOPTYCHIDAE 


185 


Table  47.    Summary  of  Diagnostic  Characters  for  Eggs  of  Certain  Stomiiform  Fishes. 


Illus- 

Species 

Egg  diameter 

Oil  globule 

Diameter 

Yolk 

Special  features 

trated 

Source 

Argyropelecus 

0.92-1.04 

1 

0.26-0.28 

segmented 

large  oil  globule 

Yes 

Sanzo,  1928 

hemigymniis 

Ichthyococcus 

0.80 

1 

0.24 

segmented 

large  oil  globule 

Yes 

Sanzo,  1930b 

ovalus 

Maurolicus 

1.63 

1 

0.25 

segmented 

hexagonal  pattern 

Yes 

Mito,  1961a 

muellen 

1.32-1.58 

1 

0.26-0.28 

segmented 

on  shell 

Yes 

Sanzo,  193  Id 

1  inciguerna 

0.58-0.74 

none 

irregularly 

thin  inner  shell 

Yes 

Ahlstrom  and 

lucc'lia 

segmented 

membrane 

Counts,  1958 

powenae 

0.75-0.85 

1 

0.17-0.19 

segmented 

no  thin  inner  shell 
membrane 

Yes 

Ahlstrom  and 
Counts,  1958 

nimhana 

0.64-0.72 

none 

irregularly 
segmented 

thin  inner  shell 
membrane 

No 

Ahlstrom  and 
Counts,  1958 

atlenuata 

0.84-0.92 

1 

0.18-0.195 

segmented 

no  thin  inner  shell 
membrane 

No 

Sanzo,  193  Id 

Gonosloma 

0.80-0.81 

1 

0.20-0.21 

— 

— 

No 

Sanzo,  193  Id 

denudatum 

Ahlstrom  (1974:672)  favored  recognition  of  one  family  for 
those  stomiiforms  with  elongate  gill  rakers  in  adults.  According 
to  the  rules  of  priority  this  would  be  the  Stemoptychidae.  Weitz- 
man  (1974:338)  recognized  three  families,  Gonostomalidae, 
Photichthyidae,  and  Stemoptychidae,  for  the  same  stomiiforms, 
the  last  family  including  the  "maurolicin"  genera  formerly  as- 
signed to  the  Gonostomalidae  and  the  deep-bodied  stemop- 
tychids  traditionally  assigned  to  the  family.  In  a  phylogenetic 
or  cladislic  analysis  this  elongate  gill  raker  bearing  "group,"  if 
recognized  as  a  single  family,  is  paraphyletic  if  one  considers 
certain  of  its  subgroups  as  equivalent  or  higher  taxonomic  cat- 
egories. For  example,  recognition  of  Ahlstrom's  Stemoptychi- 
dae, which  would  include  the  Stomiidae,  a  monophyletic  group 
with  its  members  having  a  median  barbel  attached  to  the  ventral 
surface  of  the  head  in  association  with  the  hyoid  bone  and 
lacking  elongate  gill  rakers  in  adults,  is  incompatible  with  a 
phylogenetic  classification  based  on  nested  monophyletic  groups, 
since  the  Stomiidae  is  the  sister  group  of  another  group  within 
Ahlstrom's  Stemoptychidae.  Furthermore,  the  character  used 
here  to  "define"  the  paraphyletic  Stemoptychidae,  the  presence 
of  elongate  gill  rakers  in  adults,  is  excellent  for  use  in  a  key  for 
identification  purposes,  but  cannot  be  used  as  a  synapomorphy 
relating  these  fishes  because  it  is  primitive  for  stomiiforms. 
Ahlstrom's  Stemoptychidae  is  undefinable  in  a  phylogenetic 
analysis  based  on  the  information  at  hand.  A  resolution  of  the 
use  of  familial  and  subordinal  names  in  stomiiform  fishes  must 
await  completion  of  ongoing  phylogenetic  studies  of  these  fishes. 
Because  these  studies  are  incomplete,  it  is  difficult  to  make 
recommendations  for  names  of  certain  stomiiform  subgroups. 
Among  the  stomiiforms  with  elongate  gill  rakers  in  adults,  the 
"family"  problem  is  more  complex  than  that  recognized  by 
Ahlstrom  (1974)  or  Weitzman  (1974).  We  here  recognize  two 
family  names  but  these  apply  to  only  some  of  the  24  genera 
listed  below.  We  recognize  the  Stemoptychidae  of  Weitzman 
(1974)  and  the  Gonostomatidae  in  a  new  and  restricted  sense. 
See  discussion  below. 

The  stomiiforms  discussed  here  include  the  following  24  gen- 
era, listed  alphabetically,  which  have  been  variously  recognized 
as  belonging  to  the  families  Gonostomatidae,  Stemoptychidae, 
Maurolicidae,  and  Photichthyidae: 


Araiophos  Grey  (two  species),  Argyripnus  Gilbert  and  Cramer 
(four,  possibly  a  few  more),  Argyropelecus  Cocco  (about  sev- 
en), Bonapartia  Goode  and  Bean  (one).  Cyclolhone  Goode 
and  Bean  (twelve).  Danaphos  Bruun  (one,  possibly  two),  Dip- 
lophos  GUnther  (two),  Gonostoma  Rafinesque  (six),  Ichthyo- 
coccus Bonaparte  (three),  Manducus  Goode  and  Bean  (two),' 
Margrethia  Jespersen  and  Tuning  (one,  possibly  two),  Mau- 
rolicus Cocco  (one,  possibly  two).  Photichthys  Hutton  (one), 
Pollichthys  Grey  (one),  Polyipnus  Giinther  (about  sixteen), 
Polymetme  McCulloch  (one,  possibly  four),  Sonoda  Grey  (two), 
Sternoptyx  Hermann  (two  or  three),  Thorophos  Bruun  (two, 
including  Neophos  Myers),  Triplophos  Brauer  (one),  Valen- 
ciennellus  Jordan  and  Evermann  (one,  possibly  two),  Vinci- 
guerria  Jordan  and  Evermann  (five),  Woodsia  Grey  (one), 
and  Yarella  Goode  and  Bean  (one). 


'  Grey  (1964:88)  recognized  Manducus  Goode  and  Bean,  1896  as  a 
junior  synonym  of  Diptophos  GxmXheT,  1873  because,  as  she  stated  ".  .  . 
the  differences  appear  to  be  of  a  specific  rather  than  a  generic  nature 
.  .  ."  This  was  in  the  context  of  the  kinds  of  differences  Grey  noted 
separating  other  species  of  "gonostomatids."  She  did  recognize  both  as 
subgenera  ot  Diphphos.  We  recognize  both  as  genera.  The  species  were 
most  recently  reviewed  by  Mukhacheva  (1978)  who  recognized  four 
species,  D.  maderensis  (Johnson),  D.  rebamsi  Krefft  and  Farm,  D.  greyae 
R.  K.  Johnson,  and  D.  taenia  Giinther.  We  have  examined  all  four 
species  and  find  that  D.  taenia  and  D.  rebamsi  have  the  cartilages  of 
the  two  medial  proximal  pectoral  radials,  radials  III  and  IV  in  the 
terminology  of  Fink  and  Weitzman  (1982:66),  fused  while  retaining  two 
bony  elements  separate  as  reported  for  D.  taenia  by  Fink  and  Weitzman 
(1982:65-67).  Furthermore,  one  of  the  distal  radials  is  out  of  line,  not 
in  a  single  series  in  these  two  species.  These  characters  are  specialized 
for  these  species.  In  Manducus  maderensis  and  A/,  greyae  there  are  four 
completely  distinct  proximal  radials  and  the  distal  radials  are  all  in  a 
simple  straight  series.  Because  the  pectoral  radial  morphology  in  Diplo- 
phos  taenia  and  D.  rebamsi  may  be  an  intermediate  stage  of  a  transition 
series  between  radials  such  as  are  found  in  Manducus  maderensis  and 
M.  greyae  and  those  in  the  "photichthyid"  genera,  we  recognize  Man- 
ducus &%  a  genus  and  apparent  sister  group  of  the  "photichthyid"  genera 
as  well  as  the  Stomiidae,  nearly  all  of  which  have  the  radials  111  and  IV 
completely  fused  to  one  bone.  A  few  stomiids  have  an  apparent  neo- 
morph  condition  in  which  the  third  proximal  radial  is  divided  into  two 
radials,  giving  a  total  of  four  proximal  radials.  See  also  text  discussion. 


186 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES- AHLSTROM  SYMPOSIUM 


Table  48.    Summary  of  Meristic  Characters  for  Adults  of  Certain  Stomiiform  Fishes. 


No. 
species 

Fin  rays 

Branchi- 

oslegal 

rays 

No.  of 
vertebrae 

Genera 

Dorsal 

Anal 

Pectoral 

Pelvic 

No.  of  gill  rakers 

Araiophos 

2 

13-20 

20-29 

16-18 

5 

9-11 

43-45 

2-3  +  12-19  =  14-22 

Argyripnus 

4+ 

11-12 

11-15  +  8-12  =  22-29 

15-19 

6-7 

8-10 

41-46 

4-7  +  12-19=  16-26 

Argyropelecus 

7 

(8)9(10) 

6-8  +  5-6=11 

-13 

10-11 

7 

7 

34-40 

15-24 

Bonapartia 

1 

17-20 

29-31 

14-16 

7-8 

13-16 

37 

5-6  +  11-12=  16-18 

Cydothone 

12  + 

12-15 

16-21 

9-13 

6-7 

10-14 

29-33 

4-10  +  9-18=  14-27 

Danaphos 

1 

6 

24-25 

13-14 

6 

9-10 

38 

2  +  11-13  =  13-15 

Diplophos 

2 

10-13 

47-69 

8-9 

7 

10-14 

44-94 

3  +  7-9=  10-12 

Gonostoina 

6 

10-18 

21-31 

9-13 

6-8 

10-13 

37-40 

5-11  +  10-17=  15-27 

Ichlhyococcus 

3 

10-15 

13-17 

7-8 

6-7 

11-12 

38^7 

7-11  +  15-26  =  22-37 

Manducus 

2 

11-13 

36-59 

9-11 

8 

11-14 

63-76 

3-5  +  8-10  =  12-14 

Margrethia 

1 

15-16 

21-26 

13-15 

8 

13 

34 

5  +  10-11  =  15-16 

Mauro/icus 

1 

9-12 

8-10  + 

11-15  = 

19-27 

17-20 

6-7 

9-10 

33-35 

4-8  +  17-22  =  22-30 

Photkhthys 

1 

12-13 

23-26 

9 

6-7 

20-21 

51 

4-5  +  11  =  15-16 

Pollichlhys 

1 

10-12 

22-30 

8 

6-7 

11-12 

40 

4-5  +  11-12=  15-17 

Polyipnus 

17 

10-17 

13-19 

12-16 

7 

9 

31-36 

10-28 

Polymetme 

3 

11-13 

24-33 

9-11 

7  (8?) 

12-14 

44-45 

5-8  +  9-12=  15-19 

Sonoda 

2 

8-9 

8-10  + 

14-16  = 

22-25 

13-15 

6 

8-10 

40? 

3-5  +  15-18=  18-21 

Sternoptyx 

3 

8-11 

14-16 

10-11 

7 

7 

28-31 

7-9 

Triplophos 

1 

10-12 

53-63 

9-11 

6-7 

11-14 

ca  60 

9  +  14-16  =  23-25 

Thorophos 

2 

8 

38 

13 

7 

7-8 

40-45 

5  +  13-14=  18-19 

Vatenciennellus 

2  or  3 

7-12 

22-25 

12-13 

6-9 

9-10 

32-33? 

2-3  +12=  14-15 

Vinciguerha 

4 

13-16 

12-17 

9-10 

7 

10-12 

38-42 

3  +  11-23-11  =  15-33 

Woodsia 

1 

11-12 

14 

9-10 

7-8 

17 

42-45 

3-5  +  13  =  16-18 

Yarella 

■> 

14-16(17) 

(28)29- 

U 

8-10 

6-7 

13-16 

45-54 

6-7  +  12-16=  18-22 

Table  49.    Position  of  the  Dorsal  and  Anal  Fin  and  Condition  of  the  Adipose  Fin  in  Certain  Stomiiform  Fishes. 


Dorsal  &n  position 


Genus 


Adipose  fin 


Anal  origin  in  advance  of  dorsal  fin.  Dorsal  origin 

opposite  5th  or  6th  anal  ray 
Anal  origin  opposite  dorsal  origin 
Anal  origin  opposite  last  dorsal  fin  ray 

Anal  origin  well  in  advance  of  dorsal  by  9  rays 

Anal  origin  opposite  dorsal  fin  or  slightly  behind 

Anal  origin  behind  dorsal  fin 

Anal  origin  beneath  5th  ray  or  behind  dorsal 
fin 

Anal  origin  opposite  or  3-4  rays  in  advance  of 
dorsal  origin 

Anal  origin  behind  dorsal  fin  by  a  space  =  '/2  dor- 
sal base 

Anal  origin  beneath  3rd  from  last  or  last  dorsal 
fin  ray 

Anal  origin  beneath  5th  dorsal  fin  ray 

Anal  origin  beneath  last  dorsal  fin  ray 


Anal  origin  behind  dorsal  fin 

Anal  origin  beneath  3rd  dorsal  fin  ray 

Anal  origin  usually  beneath  middle  of  dorsal  fin 

Anal  origin  beneath  end  of  dorsal  fin 

Anal  origin  in  advance  of  dorsal.  Dorsal  origin 

above  5th  anal  ray 
Anal  ongin  opposite  dorsal  origin 
Anal  origin  beneath  end  of  dorsal  fin 
Anal  origin  in  advance  of  dorsal  origin  by  3  or  4 

rays 
Anal  origin  1  or  2  rays  in  advance  of  dorsal  origin 
Anal  ongin  beneath  middle  of  dorsal  fin 
Anal  origin  behind  middle  of  dorsal  fin  by  dis- 
tance about  =  dorsal  base 
Anal  origin  beneath  middle  of  dorsal  fin 


Anal  opposite  dorsal  at  8  mm,  adult 

position  at  1 1  m 
Anal  origin  opposite  dorsal  origin 
Anal  origin  behind  dorsal  fin 

Same  as  adult 
Same  as  adult 
Same  as  adult 
Anal  origin  beneath  end  of  or  behind 

dorsal  fin 
Same  as  adult 


Anal  origin  behind  dorsal  fin 
Unknown 


Araiophos 

Argyripnus 
Argyropelecus 

Bonapartia 
Cydothone 
Danaphos 
Diplophos 

Gonostoma 

Ichthyococcus 

Manducus 

Margrethia 
Maurolicus 


Photichthys 
Pollichlhys 

Polyipnus 

Polymetme 
Sonoda 

Sternoptyx 
Triplophos 
Thorophos 

Valenciennellm 

Vinciguerria 

Woodsia 

Yarella 


Present  or  ab- 
sent 

Present 

Present  or  ab- 
sent 

Absent 

Absent 

Absent 

Absent 

Present  or  ab- 
sent 
Present 

Absent 


Same  as  adult 

Present 

Anal  origin  beneath  middle  of  dorsal 

Present 

fin,  advances  to  adult  condition  as 

juveniles 

Unknown 

Present 

Anal  origin  advances  forward  beneath 

Present 

dorsal  fin 

Same  as  adult 

Present  or  ab 

sent 

Unknown 

Present 

Unknown 

Absent 

Anal  origin  behind  dorsal  fin 

Present 

Unknown 

Absent 

Unknown 

Present  or  ab 

sent 

Same  as  adult 

Present 

Same  as  adult 

Present 

Same  as  adult 

Present 

Same  as  adult 


Absent 


AHLSTROM  ET  AL.:  GONOSTOMATIDAE,  STERNOPTYCHIDAE 


187 


Table  50.    Dernition  of  Alphabetical  Symbols  used  for  Designating  Photophores  in  Deep  Bodied  Sternoptychids  and  Other  Stomiiform 

Fishes. 


Other  slomiiforms 


Deep  bodied  slemoplychids 


SO  Symphyseal  photophores  (organs)  located  at  tip  of 

lower  jaw. 

Orb  Photophores  associated  with  the  eye  located  ante- 

rior and  posterior  of  orbit. 

Op  Photophores  on  opercle  series  generally  three,  cod- 

ed as  follows  1/(1  -I-  1). 

Br(BRP)  Photophores  located  on  the  branchiostegal  mem- 

branes. 

Is(I)  Photophores  located  on  the  isthmus. 

IP  Photophores  of  the  ventral  series  found  from  the 

isthmus  to  the  base  of  the  pectoral  fin. 

PV  Photophores  of  the  ventral  series  found  from  the 

pectoral  fin  base  to  the  pelvic  (ventral)  fin  base. 

VAV  Photophores  of  the  ventral  series  found  from  the 

pelvic  (ventral)  fin  base  to  the  anal  fin  base. 

AC  Photophores  of  the  ventral  senes  found  from  the 

anal  fin  base  to  caudal  fin  base  of  the  ventral  se- 
ries. 

IC  Summary  of  photophores  of  the  ventral  series  from 

the  isthmus  to  caudal  fin  base 
(IP  +  PV  +  VAV  +  AC). 

IV  Summary  of  photophores  of  the  ventral  series  from 

isthmus  to  pelvic  (ventral)  fin  base  (IP  +  PV). 

OV  Photophores  of  the  lateral  series  from  the  opercle 

to  pelvic  (ventral)  fin  base. 

VA(VALA)  Photophores  of  the  lateral  series  from  the  pelvic 

(ventral)  fin  base  to  the  anal  fin  base. 

OAA  Summary  of  photophores  of  OV  plus  VA  series. 

OA(OAB)  Summary  of  lateral  photophores  from  the  opercle 

to  anal  fin  base  (OV  +  VA). 

OAC(OC)  Entire  lateral  series  on  body  sides  just  dorsal  to 

ventral  series  and  extending  from  opercular 
border,  or  just  medial  to  it,  over  anal  fin  to  cau- 
dal fin  base. 

ODM  Photophores  (organs)  found  dorsal  to  the  lateral 

midline  (found  only  in  Gonosloma  gracile). 


SO 


PO 
PTO 


PRO 

Br 
Is 
AB 


PAN 

AN 
SC 


SAB 

SP 
L 

SAN 


Subopercle  photophore  which  is  equivalent  to  pos- 
teriomost  photophore  in  opercular  series  of  gon- 
ostomatids. 

Photophore  located  anterior  to  orbit. 

Photophore  located  posterior  to  orbit  and  may  be 
equivalent  to  upper  photophore  of  opercular  se- 
ries of  gonostomatids. 

Preopercular  photophore,  used  for  an  PO  photo- 
phore dorsal  to  ventral  limb  or  preopercle. 

Same  as  gonostomatid  definition. 

Same  as  gonostomatid  definition. 

Photophores  of  ventral  series  located  abdominally 
between  pectoral  fin  base  and  pelvic  fin  base  and 
equivalent  to  PV  in  gonostomatids,  plus  a  few 
posterior  photophores  of  the  IP  series. 

Photophores  found  anterior  to  anal  fin  and  may  be 
equivalent  to  VAV  or  VA  in  gonostomatids. 

Photophores  found  above  anal  fin. 

Photophores  found  on  lower  (sub)  caudal  peduncle. 
Together  with  AN  group  may  be  equivalent  to 
AC  in  gonostomatids. 

Photophores  located  above  (supra)  to  the  abdomi- 
nal series  and  may  be  equivalent  to  VA  in  gon- 
ostomatids. 

Photophores  located  above  (supra)  the  pectoral  fin 
and  may  be  equivalent  to  OV  in  gonostomatids. 

Photophore  located  laterally  above  PAN  (found 
only  in  Polyipnus). 

Photophores  located  above  (supra)  to  anal  photo- 
phores and  equivalent  to  part  of  AC  series. 


Some  genera  are  extremely  rare  (i.e.,  Thorophos  and  Sonoda) 
while  Others  represent  the  most  abundant  vertebrate  animals 
on  earth  (Cyclothone  and  I'incigiierria). 

Developmental  information  has  been  published  for  16  of  these 
genera  (12  prior  to  Ahlstrom,  1974;  3  by  Ahlstrom,  1974;  and 
one  by  Ozawa,  1976). 

Development 

Eggs.— Eggs  were  desciibed  for  Argyropelecus  hemigymnus  by 
Sanzo  (1928);  for  Ichthyococcus  ovatus  by  Sanzo  (1930b);  for 
Maurolicus  muelleri  by  Sanzo  (193 Id),  Mito  (1961a).  and  Oki- 
yama  (1971);  for  Vinciguerna  lucetia.  V.  poweriae.  and  I',  nim- 
baria  by  Ahlstrom  and  Counts  ( 1 958);  for  V.  attenuata  by  Sanzo 
(193 Id); and  for  Gonostomadenudatumby  Sanzo (\9'i\d).  Oth- 
er accounts  provide  minimal  details  of  ovarian  eggs  of  other 
species.  The  details  of  egg  characters  are  summarized  in  Table 
47. 

Larvae.  — Much  has  been  accomplished  for  the  identification  of 
the  larvae  of  these  stomiiform  genera  and  now  descriptions  are 


available  for  all  except  Manducus.  Triplophos,  Polymetme,  Pho- 
tichthys,  Thorophos,  and  Sonoda.  The  larvae  tentatively  iden- 
tified as  Polymetme  by  Ahlstrom  ( 1 974),  on  further  examination 
by  one  of  us  (Richards),  were  determined  to  be  Pollichthys.  One 
stomiiform  larval  form  has  been  described  but  not  assigned  to 
a  genus  [designated  "Maurolicine  Alpha"  by  Ahlstrom  (1974: 
670)].  It  presumably  is  the  larva  of  some  stemoptychid  (as  de- 
fined by  Weitzman,  1974).  Descriptive  details  and  illustrations 
of  several  species  were  given  by  Ahlstrom  (1974).  Here  we  pro- 
vide new  or  additional  data  including  characters  useful  in  iden- 
tifying these  larvae  and  illustrations  of  all  the  species  described 
to  date,  including  some  illustrated  for  the  first  time. 

The  identification  of  stomiiform  larvae  with  elongate  gill  rak- 
ers as  adults  requires  a  knowledge  of  developmental  data  from 
larvae,  juveniles,  and  data  from  adults  of  the  following  char- 
acters: counts  of  fin  rays,  teeth,  and  other  meristic  characters  as 
photophores;  patterns  of  photophore  development;  and  distri- 
butions (patterns)  of  dark  chromatophores  (dark  pigment  cells). 
With  those  sets  of  data,  nearly  all  species  should  be  identifiable 
at  least  to  genus,  and  in  cases  of  complete  data,  to  species.  A 


188 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  51.    Photophore  DrsTRiBUTiON  in  Certain  Stomtiform  Genera.  Refer  to  text  and  Table  50  for  definition  of  codes. 


Photo- 

phores in 

No.  of 

group   of 

rows 

so 

orb 

OP 

BR 

IS 

IV 

VAV 

AC 

glands 

Araiophos 

1 

No 

1 

1 

5-7 

Yes 

(2) 

+  (3)  +  3-4  +  (2) 
=  10-11 

3-5 

6-8 

Yes 

Argyripnus 

2 

No 

1 

3 

6 

Yes 

(6) +  (10) 

(18-28) 

(4-5)  +  (12-18) 
=  35-51 

Yes 

Argyropelecus 

2 

No 

2 

2 

6 

Yes 

18 

4 

10 

Yes 

Bonapartia 

1 

Yes 

1 

3 

11-13 

No 

14-16 

5-6 

18-20 

No 

Cydothone 

2 

No 

1 

2 

8-11 

No 

12-14 

4-5 

12-16 

No 

Danaphos 

2 

No 

1 

2-3 

6 

Yes 

18 

5 

22-26 

Yes 

Diplophus 

3  + 

Yes 

1 

3 

7-12  + 
0-3 
9 

Yes 

33-*9 

13-17 

33^9 

No 

Gonostoma 

2 

Yes 

1 

2-3 

No 

11-16 

3-10 

15-23 

No 

Ichthyococcus 

2 

No 

2 

3 

11-12 

Yes 

25-28 

9-14 

12-14 

No 

Manducus 

2  + 

Yes 

1 

3 

8-13 

Yes 

30-33 

12-14 

28-39 

No 

Margrethia 

1 

No 

1 

3 

9-12 

No 

13-15 

4 

17 

No 

Maurolicus 

2 

Yes 

1 

3 

(6) 

Yes 

(6) +  (12-13) 
=  18-19 

(6) 

1  +  (14-18)  +  (7- 

=  22-27 

9) 

Yes 

Pholichlhys 

2 

Yes 

2 

3 

17-18 

Yes 

10 

+  14-15  =  24-25 

15-17 

16-18 

Yes 

Poltichthys 

2 

Yes 

2 

3 

8 

Yes 

21-23 

7-9 

18-21 

No 

Polyipnus 

2 

No 

2 

2 

6 

Yes 

16 

5 

10-18 

Yes 

Polymetme 

2 

Yes 

1 

3 

9-10 

Yes 

19-21 

7-8 

21-25 

No 

Sonoda 

2 

No 

1 

3 

6-7 

Yes 

6  +  10=  16 

7-8 

(16-21)  +  (19-24) 
(5-6)  +  (5-6)  +  (5- 
=  36-43 

or 
-6) 

Yes 

Siernoptyx 

2 

No 

2 

2 

3 

Yes 

15 

3 

7 

Yes 

Tnplophos 

2  +  3  or  4 

Yes 

1 

3 

8-13 

Yes 

24-30 

5-7 

35-41 

No 

Thorophos 

2 

Yes  &  no 

1 

3 

6 

Yes 

17 

5 

13-15 

Yes 

Valenciennellus 

2 

No 

1 

3 

6 

Yes 

(3 

+  (4) +  (16-17) 
=  23-24 

(4H5) 

3-6  or  9-17 

Yes 

Vincignerna 

2 

Yes  or  no 

2 

3 

7-9 

Yes 

21-24 

7-11 

12-15 

No 

Woodsia 

2 

Yes 

2 

3 

14 

Yes 

25 

11-12 

12 

No 

Yarella 

2+  sev 

Yes 

1 

3 

11-13 

Yes 

23-25 

9-12 

20-28 

No 

summary  of  several  meristic  characters  for  genera  is  given  in 
Table  48.  The  position  of  the  dorsal  and  anal  fins  is  also  a  helpful 
aid,  but  caution  must  be  used  since  their  positions  relative  to 
other  structures  may  change  with  growth.  Also,  the  presence  or 
absence  of  the  adipose  fin  is  helpful,  but  again,  caution  is  in 
order  because  this  fin  is  fragile  and  often  damaged  or  lost  due 
to  contact  with  a  net.  These  fin  features  are  given  in  Table  49. 
Of  special  importance  in  identifying  lai^ae  and  adults  is  the 
distribution  and  patterns  of  the  photophores.  This  includes  the 
number  in  each  series,  the  patterns  of  their  distribution  in  re- 
lation to  each  other,  and  especially  the  sequence  of  development 
which  Ahlstrom  (1974)  stressed.  Some  confusion  appears  in  the 
literature  because  more  than  one  alphanumeric  code  has  been 
developed  to  indicate,  in  some  cases,  the  same  sets  of  photo- 
phores in  different  stomiiform  groups.  A  further  complication 
is  that  the  deep-bodied  stemoptychids  have  a  different  code 
because  of  their  altered  body  shape  as  adults  and  homologies 
were  considered  uncertain.  Weitzman  (1974:461),  because  he 
united  the  "maurolicin"  and  deep-bodied  stomiiforms  as  one 
family  considered  the  different  termmologies  "artificial"  and  as 
obscuring  homologies.  He  therefore  discussed  and  presented  a 
synonymy  of  stomiiform  photophores.  We  have  defined  the 
alphabetical  codes  in  Table  50  and  included  what  we  believe 
are  equivalent  photophores  in  stomiiforms.  In  this  code,  par- 
enthetical numbers  indicate  photophores  found  in  common 
glands  whereas  non-parenthetical  numbers  indicate  that  the 
photophores  are  single.  The  distribution  of  photophores  for  each 


genus  is  given  in  Table  51.  Table  52  provides  sequences  of 
photophore  formation  for  Bonapartia,  Margrethia,  and  Gon- 
ostoma.  Table  53  provides  similar  information  for  Araiophos. 
Maurolicus,  Danaphos,  Valenciennellus.  and  Argyripnus;  while 
Table  54  provides  similar  data  for  Polyipnus.  Argyropelecus, 
and  Siernoptyx.  Diagnostic  pigmentation  and  morphometric 
characters  are  summarized  in  Table  55.  Illustrations  (Figs.  98 
to  104)  are  provided  for  the  genera  for  which  larvae  are  known 
and  for  many  of  the  known  species.  In  addition,  the  following 
authors  provide  specific  information  which  will  aid  in  larval 
identifications:  Jespersen  and  TSnmg  (1919,  1926),  Sanzo 
(193 Id),  Ahlstrom  and  Counts  (1958),  Ahlstrom  and  Moser 
(1969),  Ozawa  (1976),  Grey  ( 1 964),  Badcock  and  Merrett  (1972), 
Kawaguchi  and  Marumo  (1967),  Okiyama  (1971),  Badcock 
( 1 982),  Rudometkina  (1981),  Gorbunova  (1981),  Mukhacheva 
(1964),  and  Ahlstrom  (1974). 

Relationships 

There  has  been  a  dichotomy  of  opinions  about  the  interre- 
lationships of  the  genera  and  the  use  of  family  names  based  on 
the  use  of  larval  versus  adult  morphological  characters.  Ahl- 
strom (1974:670-672)  presented  his  views  on  this  group  based 
on  larval  characters,  principally  the  mode  of  photophore  for- 
mation. The  suggested  relationships  resulting  from  his  analysis 
contrasted  in  part  with  those  of  Weitzman  (1974:472),  whose 
views  were  based  on  study  of  adult  osteology  and  soft  anatomy. 
Both  Ahlstrom  and  Weitzman  in  addition  to  their  own  data, 


AHLSTROM  ET  AL.:  GONOSTOMATIDAE,  STERNOPTYCHIDAE 


189 


Table  52.    Sequence  of  Photophore  Formation  in  Bonapania.  Margrethia.  and  Gonostoma. 


BR 


pv 


VAV 


AC 


Bonapartia 
pedaliota 


Margrclhia 
obtusirostra 


Gonostoma 
etongatum 


Gonostoma 
demidalum 


Gonostoma 
gracile 


Gonostoma 
ehelingi 

Gonostoma 
bathyphilum 

Gonostoma 

allanlicum 


adult 
9.5 
11.5 
12.0 
14.0 
15.0 
16.0 
23.0 

adult 

5.8 

6.4 

8.0 
11.3 
15.0 

adult 

6.0 

7.5 

7.9 
10.2 
13.0 
14.0 
16.7 
22.5 

adult 

18.25 

19.0 

20.75 

24.75 

29.65 

34.0 

39.0 

adult 
15.5-5-17.0 
20.0 
22.0 

adult 

13.8 

15.0 

adult 

11.0 

14.8 

adult 

12.0 
13.0 
14.5 
17.8 

18.8 
23.8 


I 

0 
0 
0 
0 
0 
1 
1 
1 

1 

0 
0 
0 
0 
0 
I 
1 

I 

0 

1 
1 

1 

0 
0 

1 

0 

1 
1 

0 
0 
0 

1 
1 
1 


I 

0 
0 
0 
0 
0 
0 
0 

0 
0 
0 
0 
0 
0 

I 

0 
0 
0 
0 
0 
0 

n 
I 

1 

0 
0 
0 
0 
0 
0 

I 
1 

0 

1 
1 

1 

0 
0 

0 
0 
0 

1 

0 
0 
0 
0 
0 
0 


11-13 

2 
3 
4 
5 
5 
6 

II 

9-12 
0 
0 
2 
6 
9 

9 
0 
0 
0 

2/1 
2 
2 
3 
9 

9 
0 
0 
I 

3 
5 
9 
9 

9 
0 
2 
9 

9 
0 
0 

9 
0 
4 

9 
0 
0 
0 
4 
9 
9 


14-15 
3 

5 

5 
10 

9 
II 
14 

13-15 

2 

6 

10 
14 
14 

15 

0 

5 

4 
10 
11 
II 
II 
15 

15-16 

1 

2 

3 

6 
14 
16 
16 

13-15 

0 

13 

14 

15 

7 
9 

11-12 

5 

10 

15-16 
0 
1 

2 

13 
16 
16 


5-<6) 
0 

2 
2 
4 
3 
5 
5 

4 
0 
2 
4 
4 
4 

(4)-5 
0 
0 
0 
2 
3 

2/3 
4 
5 

5 

0 
0 
1 
3 
5 
5 
5 

4-5 
0 
5 
4 

10 
0 
0 

4-5 
0 

2 

5 
0 
0 
0 
3 
5 
5 


16-18  +  2-3 

0 

0 

0 
3  +  1 
I  +  1 
5  +  2 
14  +  2 

13-14  +  3-4 

0 

1  +  2 

1  +  2 

5  +  3 

11+4 

21-23 
0 
0 
0 
0 
0 

1  + 
1  + 
22 

17-20 
0 

+  2 

+  3 

3  +  3 
11+3 
15  +  5 
15  +  5 


17-19 

0 

17 

18 

19 
0 
0 

20-21 
0 
0 

19 

0 

0 

0 

1 
19 
19 


II- 


0  Grey,  1964 

0  Grey,  1964 

0  Original 

0  Grey,  1964 

0  Grey,  1964 

0  Original 

0  Jespersen  and  TSning,  1919 

0  Grey,  1964 

0  Grey,  1964 

0  Ahlstrom,  1974 

0  Ahlstrom,  1974 

0  Ahlstrom,  1974 

0  Ahlstrom,  1974 

0  Ahlstrom,  1974 

13-15  Grey,  1964 

0  Ahlstrom,  1974 

0  Ahlstrom,  1974 

0  Original 

0  Ahlstrom,  1974 

0  Ahlstrom,  1974 

0  Grey,  1964 

0  Jespersen  and  Tuning,  1919 

13  Grey,  1964 

13-15  Grey,  1964 

0  Sanzo,  1912b 

0  Sanzo,  1912b 

0  Sanzo,  1912b 

0  Sanzo,  1912b 

0  Sanzo,  1912b 

13  Sanzo,  1912b 

13  Sanzo,  1912b 

1 2  +  6-7        Kawaguchi  and  Marumo,  1 967 

0  Kawaguchi  and  Marumo,  1967 

0  Kawaguchi  and  Marumo,  1967 

12  +  4  Kawaguchi  and  Marumo,  1967 

21  Grey,  1964 

0  Ahlstrom,  1974 

0  Ahlstrom,  1974 

14  Grey,  1964 

0  Ahlstrom,  1974 

0  Ahlstrom,  1974 

13  Grey,  1964 

0  Ahlstrom,  1974 

0  Ahlstrom,  1974 

0  Ahlstrom,  1974 

0  Original 

0  Ahlstrom,  1974 

13  Original 


used  the  results  of  photophore  anatomy  research  by  Bassot  ( 1 966, 
1971)  to  support  their  conclusions.  These  results  seemingly 
completely  supported  Weitzman's  referral  of  genera  to  family 
groups  and  agreed  with  Ahlstrom  except  for  placement  of  three 
genera  — C'lr/or/iowc,  Diplophos  (including  Mandncus),  and 
Tnplophos. 

One  of  us  (Weitzman),  continues  to  study  relationships  of  the 
stomiiforms  with  elongate  gill  rakers  in  adults  and  we  offer  the 
following  analysis  as  a  current  comment  on  the  status  of  our 
knowledge  of  these  fishes.  The  two  concepts  of  relationships  by 
Ahlstrom  and  Weitzman  may  be  compared  as  follows:  Ahlstrom 


(1974:670-672)  stressed  relationships  of  taxa  based  on  photo- 
phore patterns  and  development.  Ahlstrom  (1974:672)  consid- 
ered the  stomiiforms  with  elongate  gill  rakers  in  adults  as  a 
group  comprised  of  three  groups  of  genera,  with  any  subdivision 
being  into  two  subfamilies  based  on  photophores  occurring  in- 
dividually or  in  clustered  groups.  These  groups  of  genera  in- 
clude: (I)  Those  with  individual  separate  photophores,  most  of 
the  photophores  developing  simultaneously  and  initiated  as  a 
"white"  photophore  stage.  This  group  includes  Manducus,  Dip- 
lophos, Cyclothone.  Yarella,  Pollichthys,  V'inciguerha.  Wood- 
sia,  Ichlhyococcus.  and  presumably  Triplophos  and  Polymetme, 


190 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES -AHLSTROM  SYMPOSIUM 


Table  53. 

Sequence  of 

Photophore  Formation  in  Araiophos,  Maurolicus,  Danaphos,  Valenciennellus,  and  Argyripnus. 

ORB 

OP 

so 

BR 

IP 

PV 

VAV 

AC 

OA 

Source 

Araiophos 

adult 

1 

1 

0 

(6) 

(2) 

(3)  +  3- 

(3) 

(2)  +  2  +  (2) 

No 

Ahlstrotn  and  Moser, 

1969 

eastropas 

4 +  (2) 

11.2 

0 

0 

0 

(3) 

0 

(2) 

0 

0 

- 

Ahlstrom  and  Moser, 

1969 

Maurolicus 

adult 

1 

3 

1 

(6) 

(6) 

(12) 

(6) 

3/(4)  +  (8) 

(2)  +  7 

Ahlstrom, 

1974 

muelleri 

5.5 

0 

0 

0 

(1/2) 

0 

0 

0 

0 

0 

Ahlstrom, 

1974 

6.2 

0 

0 

0 

(2) 

0 

(2) 

0 

0 

0 

Ahlstrom, 

1974 

6.5 

0 

0 

0 

(2) 

0 

(4) 

0 

0 

0 

Ahlstrom, 

1974 

6.7 

0 

1 

0 

(3) 

0 

(5) 

0 

0 

0 

Ahlstrom, 

1974 

6.9 

1 

0 

(4) 

0 

(8) 

0 

0 

0 

Ahlstrom, 

1974 

7.5 

1 

0 

(4) 

I 

(9) 

0 

0  +  (2)  +  0 

0 

Ahlstrom, 

1974 

8.6 

2 

0 

(5) 

(3) 

(12) 

(2) 

0  +  (3)  +  (3) 

0 

Ahlstrom, 

1974 

9.0 

2 

0 

(5) 

(3) 

(11) 

(2) 

0  +  (3)  +  (3) 

1 

Ahlstrom, 

1974 

9.7 

3 

0 

(5) 

(5) 

(11) 

(3) 

0  +  (4)  +  (6) 

(2)+  1 

Ahlstrom, 

1974 

10.8 

3 

0 

(6) 

(5) 

(12) 

(4) 

0  +  (5)  +  (6) 

(2) +  2 

Ahlstrom, 

1974 

13.5 

3 

0 

(6) 

(6) 

(12) 

(6) 

0  +  (9)  +  (7) 

(2) +  6 

Ahlstrom, 

1974 

Danaphos 

adult 

1 

3 

0 

(6) 

(3)  +  (4) 

(11) 

(5) 

(3)  +  16  + 

6 

Ahlstrom, 

1974 

oculatus 

(4)+  1 

Ahlstrom, 

1974 

16.5 

0 

0 

0 

(2) 

0 

0 

0 

0 

0 

Ahlstrom, 

1974 

16.5 

0 

0 

0 

(3) 

0 

(3) 

0 

0 

0 

Ahlstrom, 

1974 

19.2 

0 

0 

0 

(4) 

0 

(10) 

0 

0 

0 

Ahlstrom, 

1974 

21.0 

1 

1 

0 

(5) 

(2)  +  (4) 

(10/11) 

0 

(2)  +  0  +  0  +  0 

0 

Ahlstrom, 

1974 

21.3 

1 

1 

0 

(4/5) 

(3)  +  (4) 

(10) 

0 

(3)  +  0  +  (2)  +  0 

0 

Ahlstrom, 

1974 

21.8 

I 

2 

0 

(5) 

(3)  +  (4) 

(11) 

(2) 

(3)  +  8  +  (4)  +  0 

2 

Ahlstrom, 

1974 

24.2 

I 

2 

0 

(6) 

(3)  +  (4) 

(11) 

(2) 

(3)  +  9  +  (4)  +  0 

2 

Ahlstrom, 

1974 

Valenannellus 

adult 

1 

3 

0 

(6) 

(3)  +  (4) 

(16-17) 

(4-5) 

(3)  +  (3)  +  (3)  + 

(2) +  3 

Ahlstrom, 

1974 

thpunculatus 

(2)  +  (4) 

7.8 

0 

0 

0 

0 

0 

0 

0 

0 

0 

Original 

8.6 

0 

0 

0 

(3) 

0 

(3) 

0 

0 

0 

Ahlstrom, 

1974 

9.5 

0 

0 

0 

(4) 

0 

(6) 

0 

0 

0 

Ahlstrom, 

1974 

11.0 

0 

0 

0 

(4) 

0 

(10) 

0 

0 

0 

Original 

12.0 

0 

0 

0 

(4) 

0 

(13) 

(2) 

0 

0 

Ahlstrom, 

1974 

12.4 

1 

0 

0 

(5) 

0 

(15) 

(2) 

0 

0 

Original 

13.0 

1 

0 

0 

(5) 

(2) 

(15) 

(2) 

0 

0 

Original 

13.2 

0 

0 

0 

(4) 

0 

(14) 

(3) 

0 

0 

Ahlstrom, 

1974 

14.0 

1 

0 

0 

(5) 

(4) 

(15) 

(5) 

0 

0 

Original 

17.0 

1 

2 

0 

(4-5) 

(3)  +  (4) 

(15) 

(5) 

(3)  +  (3)  +  0  + 
(3)  +  (4) 

(2) 

Grey,  ige-) 

Argyripmis 

adult 

1 

3 

0 

(6) 

(6) 

(10) 

(26) 

(5) +  (17) 

(3) +  4 

Badcock  and  Merrett, 

1972 

atlamicus 

18.7 

1 

2 

0 

(6) 

(3) 

(10) 

(3) 

(4)  +  (4) 

0 

Badcock  and  Merrett, 

1972 

16.8 

1 

2 

0 

(6) 

(3) 

(10) 

(2) 

(4)  +  (3) 

0 

Badcock  and  Merrett, 

1972 

although  their  development  is  not  known.  (2)  Those  with  in- 
dividual, separate  photophores  that  have  a  gradual,  protracted 
metamorphosis.  This  group  includes  Bonapartia,  Margrelhia. 
and  Gonostoma.  (3)  Those  with  some  individual  photophores 
but  some  or  most  of  the  photophores  with  common  bases  [ac- 
tually a  common  lumen,  during  development  at  least]  and  hav- 
ing a  gradual,  protracted  metamorphosis.  This  group  includes 
Araiophos,  Maurolicus,  Danaphos,  Valenciennellus,  .Argyrip- 
nus, Polyipnus,  Argyropelecus,  Sternoptyx,  and  presumably 
Thorophos  and  Sonoda  ahhough  their  development  is  unknown. 
Groups  (1)  and  (2)  comprised  the  subfamily  Gonostomatinae 
and  Group  (3)  comprised  the  Stemoptychinae  in  Ahlstrom's 
concept.  Group  (3)  is  equivalent  to  Weitzman's  Stemoptychi- 
dae.  The  genus  Gonostoma  was  considered  "pivotal"  by  Ahl- 
strom; that  is,  its  relationships  could  be  with  either  the  gonos- 
tomatines  or  the  stemoptychines  of  his  concept.  In  Ahlstrom's 
conclusions,  the  photophore  pattern  of  Group  (1)  is  most  like 
that  of  the  stomiid  groups  discussed  by  Fink  in  this  volume. 

Weitzman's  classification  (1974)  concentrated  in  most  detail 
on  a  hypothesis  of  phylogenetic  relationships  within  the  family 
Stemoptychidae  as  he  defined  it.  Weitzman  (1974)  pointed  out 


that  more  detailed  studies  should  be  conducted  on  other  sto- 
miiform  genera  in  the  future,  but  he  did  discuss  their  possible 
relationships.  Based  on  the  number  of  proximal  pectoral-fin 
radials,  he  established  two  infraorders  for  stomiiform  fishes. 
Members  of  the  Infraorder  Gonostomata  were  considered  to 
have  four  proximal  pectoral-fin  radials  (except  Cyclothone  with 
one).  This  infraorder  was  divided  into  two  families  based  prin- 
cipally on  Bassot's  photophore  findings;  Family  Gonostomati- 
dae  with  Beta  type  photophores  comprised  of  Diplophos  in- 
cluding Manducus),  Triplophos,  Bonapartia,  Margrethia. 
Gonostoma,  and  Cyclothone  and  the  Family  Stemoptychidae 
with  Alpha  type  photophores  comprised  of  Thorophos,  Araio- 
phos, Maurolicus,  Danaphos,  Valenciennellus,  Argyripnus,  Son- 
oda,  Polyipnus,  Argyropelecus,  and  Sternoptyx.  The  problem 
with  Weitzman's  Gonostomata  is  that  it  was  based  on  a  prim- 
itive character  for  the  stomiiforms,  four  pectoral-fin  radials,  and 
this  character  cannot  be  used  as  a  synapomorphy  to  define  a 
subgroup  of  stomiiforms.  The  non-stemoptychid  and  non-gon- 
ostomatid  genera,  along  with  the  stomiiform  families  possessing 
barbels  originating  from  the  hyoid  bone  and  lacking  elongate 
gill  rakers  in  the  adults  (the  Stomiidae  of  Fink,  this  volume). 


AHLSTROM  ET  AL.:  GONOSTOMATIDAE,  STERNOPTYCHIDAE 


191 


Table  54.    Sequence  of  Photophore  Formation  in  Polyipnvs.  Arcyropelecvs  and  Sternoptyx. 


OP 

PRO  + 

Size 

PC 

PTO 

BR 

IS 

SO 

SP 

AB 

SAB 

PAN 

AN 

SAN 

LSC 

Source 

Polyipnus  polli 

adult 

1 

1 

6 

6 

1    +    1 

3 

10 

3 

5 

8 

3 

14 

Baird,  1971 

4.3 

0 

0 

2 

0 

0+  1 

0 

1 

0 

0 

0 

0 

0 

Original 

4.8 

1 

0 

4 

2 

1  +  1 

0 

3 

0 

0 

0 

0 

0 

Original 

5.5 

1 

1 

6 

4 

1  +  1 

2 

8 

0 

0 

0 

0 

0 

Original 

6.0 

1 

1 

6 

6 

1  +  1 

2 

10 

0 

1 

0 

0 

0 

Original 

7.5 

1 

1 

6 

6 

1  +  1 

2 

10 

0 

3 

0 

0 

0 

Original 

9.0 

1 

1 

6 

6 

1  +  1 

2 

10 

0 

3 

2 

0 

2 

Original 

9.6 

1 

1 

6 

6 

1  +  1 

2 

10 

1 

3 

2 

0 

2 

Original 

13.5 

1 

1 

6 

6 

1  +  1 

3 

10 

3 

5 

4 

0 

14 

Original 

15.3 

1 

1 

6 

6 

1  +  1 

3 

10 

3 

5 

4 

1 

14 

Original 

17.0 

1 

1 

6 

6 

1  +  1 

3 

10 

3 

5 

4 

2 

14 

Original 

18.4 

1 

1 

6 

6 

1  +  1 

3 

10 

3 

5 

6 

3 

14 

Onginal 

23.5 

1 

1 

6 

6 

1  +  1 

3 

10 

3 

5 

7 

3 

14 

Original 

Argyropelecus 

adult 

1 

1 

6 

6 

1  +  1 

2 

12 

6 

4 

6 

0 

4 

Baird,  1971 

hemigymnus 

10.92 

0 

0 

4 

6 

0  +  1 

0 

7 

0 

0 

1 

0 

2 

Sanzo,  193  Id 

9.92 

0 

0 

6 

6 

0+  1 

0 

9 

0 

0 

2 

0 

3 

Sanzo.  193  Id 

7.84 

1 

0 

6 

6 

1  +  1 

2 

12 

0 

0 

3 

0 

4 

Sanzo.  193  Id 

11.20 

1 

1 

6 

6 

1  +  1 

2 

12 

2 

3 

4 

0 

4 

Sanzo,  193  Id 

Arygropelecus  sp. 

adult 

1 

1 

6 

6 

1  +  1 

2 

12 

6 

4 

6 

0 

4 

Baird,  1971 

4.5 

0 

0 

0 

0 

0  +  0 

0 

0 

0 

0 

0 

0 

0 

Original 

9.5 

0 

0 

6 

6 

0  +  1 

0 

6 

0 

0 

1 

0 

0 

Original 

9.5 

0 

0 

6 

6 

0  +  1 

0 

8 

0 

0 

3 

0 

0 

Original 

7.0 

1 

0 

6 

6 

1  +  1 

2 

12 

0 

0 

3 

0 

3 

Original 

7.0 

1 

0 

6 

6 

1  +  1 

2 

10 

0 

0 

3 

0 

4 

Original 

7.4 

1 

1 

6 

6 

1  +  1 

2 

12 

4 

4 

4 

0 

3 

Original 

10.0 

1 

1 

6 

6 

1  +  1 

2 

12 

5 

4 

5 

0 

4 

Original 

Sternoplyx  sp. 

adult 

1 

1 

3 

5 

1  +  1 

3 

10 

0 

3 

3 

I 

4 

Baird,  1971 

4.8 

0 

0 

0 

0 

0  +  1 

0 

0 

0 

0 

0 

0 

0 

Original 

7.5 

0 

0 

0 

0 

0  +  1 

0 

0 

0 

0 

0 

0 

0 

Original 

7.8 

0 

1 

2 

3 

0  +  1 

0 

4 

0 

0 

0 

0 

0 

Onginal 

8.1 

0 

1 

2 

4 

0  +  1 

2 

7 

0 

0 

0 

0 

0 

Original 

7.6 

0 

1 

2 

5 

0+  1 

3 

10 

0 

1 

3 

0 

1 

Original 

Table  55.    Diagnostic  Pigment  Characters  and  Unusual  Morphometric  Features  of  Some  Stomiiform  Larvae. 


Genus/species 


Diagnostic  character 


Diplophos  taenia 
Bonapartia  pedaliota 

Margrelhia  obtusirostre 
Gonostoma 


Cyclolhone 

Yarella  blackfordi 
Pollichthys  mauli 
Vincigucrria 


Woodsia  nonsuchae 


Ichlhvococcus  ovatus 


Pigment  spots  on  dorsal  and  ventral  midline.  Extremely  elongated  larvae. 

Similar  to  Gonostoma  but  lacks  deep  pigment  spot  behind  eyes  and  has  pigment  on  medial  portion  of  caudal 
Ijeduncle. 

A  distinct  vertical  streak  of  pigment  on  caudal  peduncle  in  most  specimens. 

All  species  usually  have  deep  pigment  spot  behind  eyes.  Specific  differences  among  the  species  are  as  follows: 
G.  elongatum.  G.  gracile  and  G.  ehelingi  lack  pigment  on  caudal  peduncle;  G.  bathyphilum  has  pigment 
spots  on  dorsal  edge  of  caudal  peduncle;  G.  atlanticum  has  pigment  over  medial  portion  of  caudal  peduncle 
(closely  resembles  Cyclothone  in  ventral  pigmentation  and  swimbladder  position);  G.  denudatum  has  broad 
streak  of  pigment  diagonally  over  caudal  fin  base  from  dorsal  caudal  peduncle  to  base  of  lower  caudal  fin 
rays. 

A  distinct,  dark  streak  or  intense  melanophore  over  and  parallel  to  the  parhypural  on  the  caudal  fin  base, 
pigmentation  over  gut  and  along  ventral  margin  of  tail  and  a  conspicuous  swimbladder. 

Myosepta  pigmented  over  caudal  peduncle  giving  chevron  appearance. 

No  pigment  except  for  the  eyes.  Very  similar  to  Vinciguerria  in  other  aspects. 

All  species  have  medial  or  ventral  margin  caudal  pigment  spot.  I '.  nimbaria  and  V.  lucetia  have  the  caudal 
pigment  spot  restncted  to  the  ventral  margin  of  the  caudal  fin  base  and  pigment  above  the  anal  fin.  V. 
attenuata  and  I '.  poweriae  has  the  caudal  pigment  spot  in  a  medial  position  and  no  pigment  above  the  anal 
fin.  r.  attentuata  has  pigment  over  the  airbladder  which  is  lacking  in  C.  poweriae.  V.  poweriae  has  a  struc- 
ture above  the  anal  papilla  which  may  appear  as  pigment.  V.  mabahiss  is  similar  to  V.  nimbaria  and  is 
restricted  to  the  Red  Sea  (Johnson  and  Feltes,  1984). 

Melanophores  profusely  distributed  on  all  myomeres  below  the  lateral  midline.  Broad  pigment  band  along 
roof  of  mouth  continuous  with  trunk  pigment.  Also  has  a  trailing  gut  and  elongated  rays  on  pectoral  fin, 
both  of  which  may  be  missing. 

Pigment  profusely  distributed  on  all  myomeres  below  the  lateral  midline.  Elongate  pectoral  fin  rays  and  a 
trailing  gut,  both  of  which  may  be  missing. 


192 


ONTOGENY  AND  SYSTEM ATICS  OF  FISHES -AHLSTROM  SYMPOSIUM 


Fig.  98.  Lateral  views  from  top  to  bottom:  Diplophos  taenia  22.0  mm  SL,  R/V  OREGON  II  Cr.  126,  Sta.  36754,  27''30'N,  092°30'W,  May 
10,  1982,  drawn  by  J.  C.  Javech;  I'lncigiierria  lucelia  9.0  mm  SL  modified  after  Ahlstrom  and  Counts  (1958);  I'lnciguerna  powcnae  1 1.5  mm 
SL,  R/V  OREGON  II  Cr.  126,  Sta.  36746,  27°59.9'N,  088°00'W,  May  8,  1982,  drawn  by  J.  C.  Javech;  and  linciguerna  atlenuata  9.7  mm  SL 
modified  after  Jespersen  and  Tuning  (1926). 


Fig.  99.  Lateral  views  from  top  to  bottom:  Pollichlhvs  niauli  14.5  mm  SL,  R/V  OREGON  II  Cr.  126,  Sta.  36688.  26°00.5'N.  0.88°00.4W, 
April  20,  1982,  drawn  by  J.  C.  Javech;  Yarella  blackfordi  23.5  mm  SL,  R/V  OREGON  II  Cr.  126,  Sta  36752,  27°30'N,  094°30.3'W,  May  9, 
1982,  drawn  by  J.  C.  Javech;  Woodsia  nonsuchae  1 1.5  mm  SL,  Eastropac,  Sta.  75.225,  drawn  by  J.  C.  Javech;  and  Ichthyococcus  ovalus  18.1 
mm  SL,  R/V  OREGON  II  Cr.  126,  Sta.  36746,  27°59.9'N,  0.88°00'W,  May  8,  1982,  drawn  by  J.  C.  Javech. 


AHLSTROM  ET  AL.:  GONOSTOMATIDAE,  STERNOPTYCHIDAE 


193 


Fig.  100.  Lateral  views  from  top  to  bottom:  Bonapar/ia  pedaliota  1 1 .5  mm  SL.  R/V  OREGON  II  Cr.  1 26,  Sta.  36688.  26°00.5'N,  088''00.4'W, 
April  20,  1982,  drawn  by  J.  C.  Javech;  Margrethia  obtusirostra  6.7  mm  SL,  R/V  OREGON  II  Cr.  126,  Sla.  36773.  26°00.rN,  094°00.2'W,  May 
23,  1982,  drawn  by  J.  C.  Javech. 


were  placed  in  the  Infraorder  Photichthya.  Nearly  all  have  three, 
or  rarely  fewer,  proximal  pectoral-fin  radials,  a  specialized  char- 
acter which  can  be  used  as  a  synapomorphy  uniting  this  group. 
As  noted  above,  there  are  a  few  exceptions  which  bear  four 
proximal  radials  but  these  appear  to  be  either  reversals  or  are 
neomoiphic.  Within  the  Infraorder  Photichthya  the  stomiiform 
genera  with  elongate  gill  rakers  in  adults  were  placed  in  the 
Family  Photichthyidae  comprised  of  the  genera  Polymetme. 
Yarella.  Pollichthys.  Pholichlhys.  Vinciguerria.  Woodsia,  and 
Ichthyococcm.  This  placement  was  done  on  the  basis  of  the 
presence  of  Type  Gamma  photophores  in  at  least  most  of  the 
genera,  a  specialization  for  the  group  (as  well  as  for  at  least  some 
of  the  stomiid  genera)  and  therefore  a  synapomoiTahy.  The  pres- 
ence of  elongate  gill  rakers  in  this  group  is  not  a  synapomorphy 
because  it  is  primitive  for  the  group. 

Essentially,  Ahlstrom  and  Weitzman  disagreed  on  the  rela- 
tionships of  three  genera.  Alhstrom's  Group  ( 1 )  was  mostly 
equivalent  to  Weitzman's  Photichthyidae  but  included  three 
genera,  Cyclothone,  Dtplophos  (including  Manducus),  and  Trip- 
lophos.  placed  in  the  Gonostomatidae  by  Weitzman.  Otherwise, 
Weitzman's  Gonostomatidae  was  equivalent  to  Ahlstrom's 
Group  (2).  Based  on  evidence  available  to  Ahlstrom  and  Weitz- 
man, on  some  supplementary  evidence  provided  by  Fink  and 
Weitzman  (1982),  and  on  some  of  our  own  data,  we  here  present 
a  somewhat  different  arrangement  based  on  a  more  rigorous 
phylogenetic  analysis  than  done  by  Weitzman  (1974).  It  turns 
out  that  Weitzman's  analysis  of  the  Stemoptychidae  and  its 
genera  is  consistently  phylogenetic  but  that  of  outgroup  sto- 
miiforms  is  not.  Ahlstrom  (1974)  did  not  attempt  to  analyze 
his  groups  phylogenetically.  The  evidence  available  now  seems 


to  resolve  the  conflict  between  Ahlstrom  (1974)  and  Weitzman 
(1974).  However,  we  would  note  that  the  analysis  below  is  to 
be  regarded  as  a  guide  to  future  studies  rather  than  any  sort  of 
well-corroborated  phylogeny.  Parts,  at  least,  of  the  arrangement 
need  much  additional  study.  Furthermore,  the  relationships  of 
the  genera  in  contention  by  Ahlstrom  and  Weitzman  are  still 
not  fully  clear.  Some  of  these  genera,  Manducus,  Diplophos.  and 
perhaps  Triplophos,  are  relatively  primitive  within  stomiiforms 
with  few  characters  specialized  beyond  the  stomiiform  level. 
This  makes  placing  them  in  stomiiform  subgroups  difficult.  Cy- 
clothone is  more  derived  but  retains  certain  primitive  stomi- 
iform features  and  its  relationship,  although  in  our  view  is  un- 
doubtedly with  the  gonostomatids,  at  this  time  is  somewhat 
uncertain  because  our  data  are  not  fully  analyzed. 

The  conflict  between  Ahlstrom  (1974)  and  Weitzman  (1974) 
arose  in  part  because  they  both  utilized  one  or  the  other  of 
certain  characters.  Type  Beta  photophores  and  "white"  pho- 
tophore  development,  as  though  they  were  shared  specialized 
characters,  synapomorphies  indicating  relationships.  Instead, 
we  believe  these  features  are  plesiomorphous  for  stomiiform 
subgroups  and  cannot  be  used  to  support  a  hypothesis  of  rela- 
tionships among  stomiiform  genera.  Our  current  analysis  is  as 
follows. 

Fink  and  Weitzman  (1982:69-75)  list  and  discuss  eight  syn- 
apomorphies for  stomiiform  fishes.  One  of  these,  stomiiform- 
type  photophores,  was  described  in  some  detail  based  in  part 
on  Bassott  (1966,  1971).  Bassot  (1966:574-576),  Weitzman 
( 1974:338),  and  Fink  and  Weitzman  ( 1 982:70)  recognized  Type 
Beta  photophores  as  primitive  for  stomiiforms.  Bassot  (1966, 
1971)  recognized  two  other  types  of  photophores.  Type  Alpha 


194 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES- AHLSTROM  SYMPOSIUM 


Fig.  101.  Lateral  view  from  top  to  bottom:  Gonosloma  bathyphilum  1 1 .0  mm  SL  modified  after  Ahlstrom  ( 1 974);  Gonostoma  elongatum  9.8 
mm  SL  modified  after  Ahlstrom  (1974);  Gonostoma  ebeUngi  15.0  mm  SL  modified  after  Ahlstrom  (1974);  Gonostoma  atlanticum  12.0  mm  SL 
modified  after  Ahlstrom  (1974);  Gonostoma  denudalum  20.7  mm  SL  modified  after  Sanzo  (193 Id). 


AHLSTROM  ET  AL.:  GONOSTOMATIDAE,  STERNOPTYCHIDAE 


195 


Fig.  102.     Cydothone  signata  9.0  mm  SL,  drawn  by  H.  Orr. 


and  Type  Gamma,  as  being  more  specialized.  This  recognition, 
although  not  stated  by  these  authors,  is  based  on  a  concept  that 
Types  Alpha  and  Gamma  photophores  of  some  stomiiformes 
appear  to  be  elaborations  of  Type  Beta  photophores.  In  other 
words,  their  particular  features  appear  to  be  developmental  ter- 
minal additions  to  Type  Beta  photophores  and  are  therefore 
available  for  use  as  synapomorphies  for  stomiiform  subgroups. 
Although  more  detailed  analyses  of  these  features  are  needed, 
for  the  sake  of  discussion  we  here  accept  that  Type  Beta  pho- 
tophores are  primitive  for  stomiiforms. 

Weitzman  (1974:338),  on  the  basis  of  outgroup  comparison 
(not  described  or  discussed  in  his  text),  considered  four  proximal 
pectoral-fin  radials  to  be  primitive  for  stomiiforms,  their  re- 
duction to  three  or  fewer  to  be  specialized.  We  see  no  reason 
to  change  that  analysis.  Thus  three  or  fewer  proximal  pectoral- 
fin  radials  are  available  as  synapomorphous  characters  for  sto- 
miiform subgroups. 

Ahlstrom  ( 1 974:660)  described  what  can  be  labeled  as  "white" 
photophore  development  in  which  most,  or  at  least  the  ventral 
series  of  photophores,  are  "laid  down  initially  during  a  white 
photophore  stage  [before  black  pigment  develops]  and  only  a 
few  photophores  are  late  forming."  One  form  or  another  of 
"white"  photophore  development  is  common  to  all  stomiiforms 
except  those  including  the  gonostomatid  genera  Bonapariia. 
Margrethia.  and  Gonostoma,  and  the  stemoptychids  of  Weitz- 
man ( 1 974).  Members  of  these  gonostomatid  and  stemoptychid 
genera  have  a  protracted  metamorphosis  from  the  larval  stage 
as  well  as  a  gradual,  more  extended  photophore  formation.  This 
latter  type  of  photophore  development  appears  to  be  an  elab- 
oration of  "white"  photophore  development  and  thus  we  con- 
sider white  photophore  development  primitive  with  respect  to 
the  more  complicated  forms  having  prolonged  photophore  de- 
velopment. Again,  much  information  of  an  anatomical  and  de- 
velopmental nature  remains  to  be  gathered  from  the  process  of 
photophore  development. 

If  "white"  photophore  development  and  Type  Beta  photo- 
phores are  primitive  in  regard  to  stomiiform  subgroups  and 
therefore  unavailable  as  synapomorphies  for  stomiiform 
subgroups,  then  the  conflict  regarding  the  distribution  of  char- 
acters among  taxa  between  Ahlstrom  (1974)  and  Weitzman 
(1974)  disappears  in  a  phylogenetic  analysis  by  somewhat  al- 
tering certain  of  the  groups  of  both  authors  as  follows. 

In  our  tentative  scheme  of  relationships,  Weitzman's  Ster- 
noptychidae  and  Ahlstrom's  Group  (2)  genera  (Ahlstrom,  1 974: 
671),  Bonapariia,  Margrethia,  and  Gonostoma,  the  Gonosto- 
matidae  in  the  strictest  sense,  are  united  by  a  synapomorphy 
consisting  of  a  specialized  form  of  prolonged  metamorphosis 


and  photophore  development  described  by  Ahlstrom  (1974: 
660-661).  See  also  Tables  52-54  herein.  These  three  gonosto- 
matid genera  and  Cydothone  apparently  share  derived  char- 
acters of  the  jaws  and  associated  head  parts  which  will  be  ex- 
plained in  a  later  contribution.  These  four  genera  retain  the 
primitive  Type  Beta  photophores,  a  character  relating  stomi- 
iforms only  at  the  ordinal  level.  In  our  opinion  these  four  genera 
constitute  the  Gonostomatidae  and  Cydothone  may  have  lost 
prolonged  photophore  development  through  paedomorphic  re- 
versal associated  with  the  small  size  of  most  of  its  members,  a 
situation  needing  further  study. 

The  Stemoptychidae  have  specialized  Type  Alpha  photo- 
phores and  the  several  other  synapomorphies  listed  by  Weitz- 
man ( 1 974:446-448).  In  addition  they  apparently  share  a  unique 
photophore  growth  pattern  previously  unrecorded.  One  of  us 
(Weitzman)  has  been  studying  photophore  development  in  re- 
lation to  phylogenetic  studies  in  stomiiforms  and  has  found  that 
each  cluster  or  group  of  photophores  of  the  stemoptychids  ap- 
pears to  develop  by  budding  from  one  single  photophore  rather 
than  by  fusion  at  a  later  growth  stage  of  separately  developed 
photophores.  This  is  a  terminal  developmental  addition  in  pho- 
tophore ontogeny  and  both  outgroup  comparison  and  devel- 
opmental information  indicate  that  this  pattern  of  photophore 
formation  is  a  specialization  in  comparison  to  the  simpler  ap- 
pearance of  single,  separate  body  photophores  (usually  one  per 
scale  in  any  given  series  found  in  other  stomiiforms).  This  growth 
character  appears  to  be  present  in  all  stemoptychid  genera  for 
which  we  have  developmental  information.  It  is  therefore  a 
likely  synapomorphy  for  the  group. 

Manducus  (based  on  the  type  species,  Gonostoma  maderense 
Johnson)  is  a  primitive  stomiiform,  having  ordinal-level  char- 
acters with  no  known  specialized  characters  except  the  absence 
of  an  adipose  fin  and  a  short  neural  spine  on  the  preural  centmm. 
The  latter  may  be  a  primitive  rather  than  a  specialized  stomi- 
iform feature.  Diplophos  (based  on  the  type  species  Diplophos 
taenia  Gunther)  appears  to  have  a  transitional  stage  pectoral 
radial  morphology  between  Manducus  on  the  one  hand  and  the 
Photichthyidae  of  Weitzman  (1974)  (an  ill-defined  group)  and 
the  Stomiidae  on  the  other.  In  Manducus  the  cartilages  and 
bones  of  proximal  pectoral-fin  radials  III  and  IV  remain  separate 
whereas  Diplophos  has  the  cartilages,  but  not  the  bones,  of  the 
two  elements  fused.  Fink  and  Weitzman  (1982:65-67).  In  the 
"photichthyids"  and  stomiids  the  cartilages  and  bones  of  the 
two  medial  pectoral-fin  radials  are  fused.  This  represents  the 
terminal  condition  in  the  transition  series  except  that  in  some 
genera  there  is  a  reversal  of  radial  numbers  and  in  Eustomias 
there  occurs  a  further  specialized,  reduced  pectoral-fin  radial 


196 


ONTOGENY  AND  SYSTEM ATICS  OF  FISHES -AHLSTROM  SYMPOSIUM 


AHLSTROM  ET  AL.:  GONOSTOMATIDAE,  STERNOPTYCHIDAE 


197 


Fig.  104.  Lateral  views  from  top  to  bottom:  Polyipnus  polli  5.2  mm  SL  R/V  GERONIMO  Cr.  2,  Sta.  155,  05°28S,  01°120'E,  August  21, 
1963,  drawn  by  J.  C.  Javech;  Argyropelecus  hemigy'mnus  7.8  mm  SL  modified  after  Sanzo  (193 Id);  and  Sternoptyx  sp.  8.8  mm  SL.  drawn  by  H. 
C.  Orr. 


Fig.  103.  Lateral  views  from  top  to  bottom:  Araiophos  eastropas  8.8  mm  SL  modified  after  Ahlstrom  and  Moser  (1969);  Maurolicus  muelleri 
10.8  mm  SL  modified  after  Ahlstrom  (1974);  Danaphos  oculatus  middle  metamorphosis  modified  after  Ahlstrom  (1974);  Valenaennettus  tri- 
putulutatus  middle  metamorphosis  modified  after  Ahlstrom  (1974);  Argyripnus  atlanticus  18.7  mm  SL  modified  after  Badcock  and  Merrett 
(1972);  and  maurolicine  Alpha  7.5  mm  SL  modified  after  Ahlstrom  (1974). 


198 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


condition.  The  "photichthyids"  and  stomiids  have  specialized 
Type  Gamma  photophores,  although  it  is  not  known  that  all 
genera  in  these  groups  have  Type  Gamma  photophores;  this  is 
a  problem  for  further  investigation.  Manducus  and  Diplophos 
retain  Type  Beta  photophores  and  all  of  these  fishes  apparently 
retain  "white"  photophore  development  of  one  kind  or  another. 
These  two  characters  are  only  useful  at  the  ordinal  level  as 
synapomorphies.  Again,  further  research  on  "white"  photo- 
phore formation  is  needed  since  there  appears  to  be  more  than 
one  form  of  this  development. 

The  monotypic  Triplophos  may  or  may  not  be  related  to 
Manducus  and/or  Diplophos.  Triplophos  has  a  variety  of  derived 
features  not  shared  by  Manducus  or  Diplophos.  However,  this 
tells  us  nothing  about  its  possible  relationships  with  these  gen- 
era. Triplophos  has  four  proximal  pectoral-fin  radials  but  with 
some  reduction  in  radial  IV,  Type  Beta  photophores,  and  prob- 
ably "white"  photophore  development,  the  last  two  characters 
synapomorphous  only  at  the  ordinal  level.  Four  pectoral-fin 
radials  are  not  a  synapomorphy  for  stomiiforms  at  any  level 
since  the  feature  is  found  in  most  teleost  outgroups.  Triplophos 
appears  to  be  a  primitive  stomiiform  with  certain  autapomorph- 
ic  features  associated  with  an  elongate  body.  Its  relationships 
are  uncertain  and  there  may  be  indications  in  the  head  and 
pectoral  girdle  anatomy  of  a  relationship  with  certain  photich- 
thyid  genera.  The  problem  needs  much  study. 

Cyclothone  retains  Type  Beta  photophores  and  "white"  pho- 
tophore development  but  has  its  own  specialized  features  such 
as  only  one  pectoral-fin  radial.  It  has  a  modified  head  and  jaws, 
which  resemble  and  are,  in  our  opinion,  synapomorphous  with 
those  of  Gonostoma.  The  single  pectoral-fin  radial  might  be 
thought  of  as  a  terminal  stage  in  a  transition  series  from  Man- 
ducus (with  four  pectoral-fin  radials)  to  Diplophos  to  some  mem- 
bers of  the  "Photichthyidae"  and  then  to  Cyclothone.  However, 
Cyclothone  does  not  have  specialized  Type  Gamma  photo- 
phores of  the  "photichthyid"  genera.  The  phylogenetic  rela- 
tionships of  Cyclothone  may  not  be  certain  as  yet,  but  in  many 
respects  it  bears  a  resemblance  to  the  three  gonostomatid  genera 
and  we  favor  its  placement  with  these  genera.  See  also  discussion 
above. 

Although  we  have  perhaps  resolved  the  differences  between 
Ahlstrom  (1974)  and  Weitzman  (1974),  we  have  not  achieved 
a  useful  phylogeny  of  most  stomiiform  groups.  Rather,  we  have 
attempted  to  outline  certain  suggested  hypotheses  of  relation- 
ships to  be  investigated  in  the  future  with  additional  data.  Adult 
morphological  data  of  the  kind  used  by  Weitzman  to  define  and 
relate  the  stemoptychid  genera  are  available  in  abundance  and 
may  be  useful  for  other  stomiiform  groups.  A  closer  look  at 
growth  stages  with  the  specific  purpose  of  looking  for  possible 
developmental  specializations  and  terminal  additions  to  char- 
acters found  in  outgroups  should  greatly  aid  in  delineating  re- 
lationships among  the  stomiiform  genera.  However,  problems 
associated  with  a  high  percentage  of  homoplasy  can  be  expected 
for  some  groups.  The  answers  to  problems  of  stomiiform  in- 
terrelationships will  not  come  easily. 

Consideration  of  certain  features  is  in  order.  For  example, 
larvae  of  Diplophos  superficially  resemble  those  of  Chauliodus 
with  their  prolonged  development  to  a  large  larval  size  and  great 
elongation  with  bodies  that  are  circular  in  cross  section.  Are 
these  convergent  larval  specializations  or  primitive  stomiiform 
features  found  only  in  certain  stomiiform  genera?  The  ventral 
pigmentation  on  the  body  of  developing  Diplophos  resembles 


that  of  developing  paralepidids  and  myctophoids.  Is  this  a  prim- 
itive stomiiform  feature  of  Diplophos  shared  with  certain  sto- 
miiform outgroups  or  a  gross  convergence  of  pigment  patterns? 

Woodsia  and  Ichthyococcus  share  with  certain  stomiid  genera 
(for  example,  Eustomias)  such  developmental  features  as  elon- 
gate pectoral-fin  rays,  trailing  guts,  pigmentation  patterns,  and 
bodies  with  a  circular  cross  section.  Some,  if  not  all,  of  these 
may  be  shared  larval  specializations.  But  again,  independent 
appearance  of  these  characters  indicated  by  a  high  degree  of 
homoplasy  may  be  a  vexing  problem.  Larvae  of  other  genera 
such  as  Vinciguerria.  Pollichthys,  and  Cyclothone  have  body 
shapes  and  certain  other  features  that  closely,  but  presumably 
superficially,  resemble  those  of  clupeoid  larvae.  Detailed  com- 
parisons of  these  similarities  may  possibly  distinguish  between 
homology  and  convergence  among  these  taxa. 

In  summary,  a  future  phylogenetic  analysis  based  on  much 
additional  data  may  clear  up  many  of  the  problems  of  stomi- 
iform generic  relationships.  However,  at  present  we  are  left  with 
numerous  phylogenetic  problems  and  assignment  of  certain  gen- 
era to  family-level  groups  at  this  time  would  be  misleading.  The 
above  analysis  retains  Weitzman's  Stemoptychidae.  It  restricts 
the  Gonostomatidae  to  the  genera  Bonapartia.  Margrethia,  and 
Gonostoma.  and  we  recommend  the  inclusion  of  Cyclothone. 

The  other  groups  of  non-stomiid  stomiiforms  remain  unclear 
as  to  family  relationships.  We  agree  with  Fink  and  Weitzman 
(1982)  that  Manducus  and  Diplophos  are  primitive  stomiiforms, 
but  we  cannot  provide  a  stable  classification  for  Manducus. 
Diplophos.  and  Triplophos.  Manducus  and  Diplophos  might  seem 
to  be  sister  taxa  because  of  their  similarity  of  appearance.  How- 
ever, they  share  no  known  specialized  character  or  characters 
that  would  unite  them  as  a  stomiiform  subgroup  except  the 
absence  of  an  adipose  fin  and  possibly  a  short  neural  spine  on 
the  preural  centrum.  Currently  all  their  other  shared  characters 
seem  primitive  for  stomiiforms.  Further  analysis  of  this  situa- 
tion is  needed. 

Triplophos  is  again  very  much  like  a  primitive  stomiiform  in 
its  head  especially,  but  it  has  a  number  of  specialized  stomiiform 
features  as  listed  by  Grey  (1964:106)  and  may  show  some  re- 
lationship to  some  of  the  "photichthyid"  genera. 

That  the  genera  classified  in  the  "Photichthyidae"  by  Weitz- 
man (1974)  form  some  kind  of  related  group  seems  reasonable. 
However,  relationships  among  these  genera  are  not  known.  That 
these  "photichthyid"  genera  are  related  to  Diplophos  is  possible, 
and  that  the  stomiids  are  related  to  the  "photichthyids"  is,  in 
our  view,  very  probable.  The  larval  specializations  of  Woodsia 
and  Ichthyococcus  noted  above,  may  be  important  here  because 
they  may  be  synapomorphies  relating  these  genera  to  the  sto- 
miids. 

Until  the  developmental  and  adult  morphological  features  of 
many  stomiiform  genera  are  analyzed  in  detail,  certain  aspects 
of  their  developmental  stages  outlined,  and  detailed  outgroup 
analysis  performed  on  all  putatively  useful  characters,  we  can 
make  no  certain  predictions  about  relationships  and  classifi- 
cation. 

(W.J.R.)  National  Marine  Fisheries  Service,  Southeast 
Fisheries  Center,  75  Virginia  Beach  Drive,  Miami, 
Florida  33149;  (S.H.W.)  Division  of  Fishes,  National 
Museum  of  Natural  History,  Smithsonian  Institution, 
Washington,  D.C.  20560. 


Giganturidae:  Development  and  Relationships 
R.  K.  Johnson 


THE  Giganturidae  contains  two  highly-specialized  bathy- 
pelagic  species  placed  in  two  monotypic  genera;  Gigantura 
chum  Brauer,  1901  and  Rosaura  mdica  (Brauer,  1901).  Adults 
now  placed  in  Rosaura  were  formerly  recognized  as  Bathyleptus 
Walters,  1961.  Morphological  specializations  of  giganturids  are 
sufficiently  divergent  and  numerous  that  the  group  has  usually 
been  accorded  subordinal  or  ordinal  status  somewhere  within 
the  group  now  recognized  as  basal  neoteleosts  (Stomiiformes  + 
"Aulopiformes"  +  "Myctophiformes,"  see  Rosen,  1973;  John- 
son, 1982;  Fink  and  Weitzman,  1982). 

Giganturids  are  oceanic  and  deep  mesopelagic  or  bathypelagic 
as  juveniles  and  adults.  Most  hauls  successful  for  juveniles  and 
adults  have  been  at  depths  in  excess  of  500  m  (with  closing  net 
captures  as  deep  as  2,000-2,500  m).  There  is  no  evidence  for 
diel  vertical  migration.  G.  chum  is  tropical,  R.  indica  tropical- 
subtropical  (sensu  Johnson,  1982;  185).  Giganturids  are  un- 
known from  the  Southern  Ocean,  Pacific  Subarctic,  temperate 
North  Atlantic  (including  Mediterranean),  and  only  a  single 
specimen  (G.  chum')  is  known  from  the  eastern  tropical  Pacific. 

Giganturids  are  relatively  large-bodied  with  adults  of  Rosaura 
achieving  more  than  220  mm  SL,  adults  of  Gigantura  more 
than  1 70  mm  SL.  Giganturids  are  well-known  swallowers  with 
greatly  expandable  pouchlike  stomachs.  Most  identifiable  gut 
contents  have  been  fishes,  often  single  large  fish  ingested  whole 
(e.g.,  Regan,  1925).  Transformed  giganturids  are  distinguished 
from  most  or  all  other  teleosts  by  the  following  combination  of 
characters;  (A)  eyes  tubular,  directed  straight  forward,  in  parallel 
with  main  axis  of  body;  (B)  gape  of  mouth  extends  far  behind 
eye;  teeth  fang-like,  unbarbed,  recurved,  depressible;  teeth  bi- 
serial  on  each  jaw,  a  medial  row  of  enlarged  canines  and  a  lateral, 
more  irregular  row  of  smaller  canines;  anteriormost  canine  in 
each  jaw  recurving  anteriad;  (C)  bases  of  pectoral  fins  nearly 
horizontal,  above  the  gill  openings;  pectoral  fins  with  a  very 
high  fin-ray  count,  37  to  43  in  Rosaura.  30-33  in  Gigantura; 
(D)  caudal  forked,  middle  rays  of  lower  lobe  lengthened  enor- 
mously; in  one  120.3  mm  SL  specimen  of  G.  chuni  the  fila- 
mentous extension  of  the  lower  caudal  lobe  adds  243  mm  to 
the  length  of  the  fish;  (E)  skin  loose,  scaleless,  with  a  thick  layer 
of  mesenchymal  jelly  adding  substantially  to  an  overall  char- 
acteristic flabbiness;  (F)  stomach  a  thickwalled  blind  pouch, 
giving  rise  to  the  intestine  ventrally,  near  midline;  intestine 
passing  laterad  and  dorsad,  to  right,  continuing  along  dorsal 
contour  of  stomach  until  finally  turning  ventrad  behind  poste- 
rior terminus  of  stomach  and  ending  at  anal  papilla;  (G)  lack 
of  pelvic  fins,  dorsal  adipose  fin,  branchiostegal  rays,  gill  rakers; 
loss  of  most  of  gill  arch  elements  on  arches  I-III,  but  with  strong, 
recurved  teeth  on  3rd  pharyngobranchial  (pb)  and  4th  pb  tooth- 
plate;  loss  of  numerous  other  skeletal  elements  (cf  Regan,  1 925; 
Walters,  1961,  1964;  Rosen,  1973);  and  (H)  considerable  con- 
solidation of  caudal  fin  skeleton  with  two  presumably  com- 
pound hypurals  (Rosen,  1973). 

Development 

Eggs  of  giganturids  are  unknown.  Larvae  are  known  for  both 
species  but  only  the  larva  of  Rosaura  (a  single  8.4  mm  specimen. 


Fig.  105)  has  been  illustrated  (Tucker,  1954).  For  both  species 
larvae  have  commonly  been  taken  in  the  upper  100  m.  The 
distributional  ranges  of  larvae  and  adults  are  coextensive  and 
there  is  no  evidence  for  seasonality  in  reproductive  effort  (with 
only  ca  400  known  larval  specimens,  the  data  are  far  from 
complete).  The  sexes  are  separate  and  according  to  Clarke  and 
Wagner  (1976)  the  females  may  reach  twice  the  size  of  males, 
although  available  data  are  sparse.  Osteological  examination 
has  been  confined  to  adults  except  for  those  elements  visible 
and  described  in  Tucker's  (1954)  astonishingly  detailed  decrip- 
tion  of  the  holotype  of  Rosaura  rotunda.  Development  is  direct 
but  transformation  is  abrupt  with  the  change  from  larval  to 
adult  morphology  occurring  over  the  approximate  size  range  of 
30-40  mm  SL  in  Gigantura  and  40-60  mm  SL  in  Rosaura. 
Transformation  series  are  now  known  for  both  species  (only  8 
transforming  specimens  of  Gigantura  are  known,  for  Rosaura 
the  count  stands  at  34)  but  these  results  remain  unpublished. 
The  interim  account  below  is  thus  based  on  work  in  progress. 

Gross  aspect  (Fig.  105).  — "Rosaura"  larvae  are  short,  deep,  glo- 
bose, translucent  and  virtually  colorless.  The  forehead  is  steep, 
the  eyes  small,  round  and  directed  laterad.  The  snout  is  pointed. 
The  body  is  deepest  at  a  vertical  through  the  center  of  the 
opercle.  The  pectoral  insertion  is  nearly  vertical.  A  dorsal  adi- 
pose and  distinct  partly-stalked  5-rayed  pelvic  fins  are  present. 
Large,  readily  visible,  rather  platelike  branchiostegal  rays  are 
present.  Raptorial  jaw  teeth  are  present  in  the  smallest  known 
larvae  (4  mm  SL).  Teeth  on  the  jaws  are  biserial  with  an  inner 
series  of  prominent  canines  and  an  outer  series  of  shorter  more 
broadbased  teeth  on  the  premaxillaries  and  dentaries.  There  are 
2-4  recurved  smaller  fangs  on  the  basihyal.  The  maxillary  is 
included  in  the  gape  but  is  edentulous.  The  abdominal  body 
wall  is  nearly  transparent  and  balloonlike,  enclosing  an  expan- 
sive gut  cavity.  The  body  form  remains  essentially  unchanged 
over  a  period  of  larval  growth  extending  to  ca  30  mm  SL  (Gi- 
gantura) and  to  ca  35  mm  SL  {Rosaura).  when  transformation 
begins.  Changes  during  transformation  are  striking,  as  described 
below.  At  all  stages— larvae,  transforming  specimens,  and  ju- 
veniles and  adults— the  species  can  be  distinguished  on  the  basis 
of  relative  depth  of  the  caudal  peduncle.  The  value  of  this  char- 
acter varies  ontogenetically  but  the  relative  peduncle  depth  is 
always  greater  in  Gigantura. 

Meristic  characters.— Courtis  of  fin  rays  do  not  differ  between 
larvae  and  adults  except  that  semi-stalked  pelvic  fins  (5  rayed) 
are  universally  present  in  larvae  and  early  transforming  speci- 
mens but  are  completely  lost  during  transformation.  Values  for 
anal-fin  ray  counts  (8  to  10  in  G.  chuni.  II  to  14  in  i?.  indica) 
and  pectoral-fin  ray  counts  (30  to  33  in  G.  chuni.  36  to  42  in 
R.  indica)  separate  the  two  species  without  overlap.  Dorsal-fin 
ray  counts  ( 1 6  to  19)  have  the  same  range  in  both  species.  The 
caudal  is  the  first  fin  to  form;  it  is  asymmetric  with  10  -t-  6(7) 
principle  caudal  rays  and  (3)4(5)  procurrent  caudal  rays  above 
and  below.  Next  to  form,  in  order,  are  the  dorsal  +  anal  fins, 
pelvic  fins,  and  pectoral  fins  (the  dorsalmost  pectoral  rays  begin 


199 


200 


ONTOGENY  AND  SYSTEMATICS  OF  RSHES-AHLSTROM  SYMPOSIUM 


."//,/., 


Fig.  105.     Giganturidae.  (Upper)  Larva  oi  Rosaura  indica.  8.4  mm  SL  (=holotype  oi  Rosaura  rotunda  from  Tucker,  1954).  (Lower)  Adult 
Rosaura  indica.  182  mm  SL  (from  Berry  and  Perkms,  1966). 


to  differentiate  in  larvae  as  small  as  5.5  mm  SL.  but  the  ventral- 
most  pectoral  rays  are  the  last  fin  rays  to  be  formed).  The  pelvic 
fins  appear  just  below  the  dorsal-fin  origin  and  do  not  greatly 
shift  in  relative  position  until  transformation.  A  dorsal  finfold 
connects  the  incipient  dorsal  fin  with  the  caudal  fin  in  small 
larvae,  but  loses  this  connection  in  larvae  larger  than  6  mm  SL. 
and  shrinks  in  extent  but  remains  as  a  highly  visible  adipose  fin 
until  transformation,  when  it  is  resorbed. 

Peritoneal  pigment  sections.  — A  single  peritoneal  pigment  sec- 
tion characterizes  the  larvae  of  both  species.  This  section  lies 
just  above  and  posterior  to  the  dorsal  transverse  limb  of  the 
intestine.  The  section  is  never  paired  as  in  synodontoids  and 
remains  proportionately  constant  in  size  throughout  larval  life 
and  is  represented  in  adults  as  a  small,  intensely-black  oval 
pigment  patch  above  the  stomach  (growth  of  the  section  ap- 
parently ceases  at  about  the  onset  of  transformation,  but  the 
section  apparently  remains  in  both  juveniles  and  adults  of  both 
species).  The  dense  brown  or  black  pigment  enclosing  the  gut 
is  not  derived  from  this  peritoneal  pigment  section,  as  is  true 
for  many  "inioms"  (see  Johnson,  1 982)  but  develops  separately 


dunng  transformation  (as  in  Aleptsaurus  and  Omosudis,  Was- 
sersug  and  Johnson.  1976). 

Other  pigmentation.  — \n  both  species  pigmentation  in  larvae 
occurs  in  three  areas  (other  than  the  peritoneal  section),  the 
eyes,  over  the  optic  lobes,  and  on  the  sides  of  the  body  posterior 
to  the  dorsal-fin  base.  In  some  but  not  all  pre-transformation 
specimens  of  Gigantura,  very  small  punctate  melanophores  ap- 
pear over  the  still  otherwise  essentially  transparent  lateral  ab- 
dominal body  wall. 

Gut  morphology.— The  stomach  is  enlarged  and  sac-like.  The 
mtestine  leaves  the  pyloric  region  of  the  stomach,  descends 
round  the  left  margin  of  the  abdominal  cavity,  crosses  trans- 
versely upon  the  ventral  body  wall,  reascends  the  right  side  and 
then  turns  again,  descending  abruptly  and  obliquely  down  and 
posteriad  to  the  vent. 

Transformation —Changes  during  transformation  are  numer- 
ous and  striking:  (A)  Body  shape.  The  body  changes  in  shape 
from  short,  rotund  and  deep,  rather  as  in  some  ceratioid  larvae 


JOHNSON:  GIGANTURIDAE 


201 


(Bertelsen,  1 95 1 )  or  the  larvae  of  certain  scopelarchids  (Johnson, 
1974b,  1982)  to  the  elongate,  shallow,  slender  shape  of  the  gi- 
ganturids.  The  head  while  still  massive  is  proportionately  much 
less  so  ('/«  vs  'A  SL  in  Rosarua)  and  the  dorsal  head  profile  is 
essentially  horizontal  rather  than  steeply  oblique  (Fig.  105).  (B) 
Eyes.  Eyes  in  larvae  are  round,  small  and  directed  laterad;  eyes 
in  adults  are  fully  tubular  and  directed  rostrad.  (C)  Fins.  Dis- 
tinct, partly-stalked,  5-rayed  pelvic  fins  are  present  in  larvae, 
resorbed  or  shed  during  transformation,  and  lacking  in  adults. 
The  line  of  insertion  of  the  pectoral-fin  rays  is  obliquely  vertical 
in  larvae,  essentially  horizontal  in  adults.  In  larvae  the  pectoral 
insertion  is  behind  the  gill  slit,  in  adults  (especially  prominent 
in  Gigantura)  the  pectoral  insertion  is  substantially  above  the 
gill  slit.  A  distinct  dorsal  adipose  fin  is  present  in  larvae,  absent 
in  adults.  Procurrent  caudal  fin  rays  number  (3)4(5)  in  larvae 
and  are  prominent,  in  adults  procurrent  caudal  rays  are  fre- 
quently embedded  in  the  skin,  difficult  to  see,  and  number 
(0)1(2,3).  (D)  Teeth.  Among  the  most  striking  changes  occurring 
dunng  transformation  is  the  total  loss  of  all  larval  teeth  (in- 
cluding basihyal  teeth).  Transforming  specimens  are  character- 
ized by  a  scalloped,  irregularly-emarginate  jaw  edge  (upper  and 
lower)  which  is  edentulous.  None  of  the  40  known  transforming 
specimens  shows  development  of  adult  teeth  and  the  smallest 
known  post-transformation  specimen  (36.4  mm  SL,  G.  chuni; 
47.9  mm  SL,  Rosaura  indica)  possess  a  full  complement  of 
adult  teeth.  (E)  Color.  Larvae  are  essentially  translucent  with 
very  little  development  of  pigment,  adults  are  entirely  blackish/ 
brown  (often  with  the  development  of  an  iridescent  finish  in 
Gigantura).  Onset  of  transformation  is  indicated  by  the  "sud- 
den" widespread  development  of  pigmentation.  (F)  Loss  of  skel- 
etal elements.  Larvae  possess  at  least  the  following  skeletal  ele- 
ments not  seen  in  adults:  symplectic,  coracoid,  cleithrum, 
posttemporal,  supracleithrum,  branchiostegal  rays. 

Relationships 

The  first  association  of  "Rosaura"  with  the  giganturids  was 
by  Ahlstrom  and  Berry  about  1960  (letters  and  mss  material 
made  available  by  H.  G.  Moser)  with  the  first  published  sug- 
gestion made  in  Berry  and  Perkins  (1966).  Key  characters  sug- 
gesting relationship  included  the  very  high  pectoral-fin  ray  count 
and  the  highly  unusual  10-1-6(7)  distribution  of  principle  caudal 
rays,  apparently  unique  to  "Rosaura"  and  the  giganturids.  The 
disparities  between  "Rosaura"  larvae  and  adult  giganturids— 
briefly  outlined  above— left  doubt  in  many  minds,  but  the  cap- 
ture of  essentially  complete  transformation  series  (to  be  de- 
scribed and  illustrated  in  detail  elsewhere)  make  it  unquestion- 
able that  "Rosaura"  is  the  larval  form  of  the  giganturids.  With 
a  caudal  peduncle  depth  of  ca  9.9%  of  SL  (Tucker,  1954:168) 
there  is  likewise  no  doubt  that  the  type  of  Rosaura  rotunda 


represents  a  larva  of  "Balhyleptus,"  requiring  recognition  of  the 
more  elongate,  shallow-bodied  species  as  Rosaura  indica  (Brauer, 
1901).  The  deeper-bodied  species  is  Gigantura  chuni  Brauer, 
1901  (other  species  have  been  described  but  the  characters  used 
to  distinguish  them  do  not  work,  nor  has  other  evidence  been 
found  to  support  the  hypothesis  of  more  than  two  species).  Of 
the  two,  Walters  (1961,  1 964)  argued  for  the  more  apomorphous 
condition  of  Gigantura  but  his  characters  need  to  be  re-exam- 
ined in  light  of  outgroup  comparisons  and  in  conjunction  with 
other  characters. 

Vanous  authors  have  allied  giganturids  with  such  disparate 
groups  as  Stylephoridae,  Saccopharyngiformes  and  "...  a  line 
[leading]  from  a  subiniomous  group  such  as  the  esocoids  toward 
the  synodontoid  inioms,  and  this  line  later  may  have  given  rise 
to  the  Cetunculi  .  .  ."  (Walters,  1961).  Rosen  (1973:438-441) 
has  offered  evidence  that  the  original  placement  by  Regan  (1925: 
57)  of  giganturids  with  synodontoids  was  correct.  Rosen  calls 
particular  attention  to  similarities  in  upper  jaw  and  infraorbital 
configuration  with  synodontoids  and  the  presence  of  a  retractor 
dorsalis  (=RAB  in  Rosen,  1973;  see  Winterbottom,  1 974b)  mus- 
cle configration  state  characteristic  of  the  synodontoid/alepi- 
sauroid  line  (Johnson,  1982:85,  95).  An  important  character 
(Johnson,  1982:71;  Okiyama,  this  volume)  uniting  synodon- 
toids with  alepisauroids  is  the  presence  in  larvae  of  multiple  (3 
or  more)  peritoneal  pigment  sections.  Uniting  synodontids  and 
harpadontids  (sensu  Sulak,  1977)  is  the  fact  that  in  larvae  of 
these  fishes  the  sections  are  paired  .  .  .  and  not  connected  over 
the  gut.  The  condition  in  "Rosaura"  is  that  seen  in  aulopids, 
chlorophthalmids,  primitive  scopelarchids,  and  ipnopids,  viz.  a 
single  section  situated  over  the  gut.  This  is  the  state  thought 
primitive  for  inioms.  Also  distinguishing  the  giganturids  is  a 
unique  conformation  of  the  gut.  In  larvae  the  gut  arises  from 
the  pylorus,  descends  round  the  left  margin  of  the  abdominal 
cavity,  crosses  transversely  midventrally,  reascends  the  right 
side,  turns  abruptly  mediad,  then  turns  again,  descending  abruptly 
and  obliquely  to  the  vent.  In  adults  the  intestine  arises  mid- 
ventrally, makes  a  few  small  twists,  ascends  the  right  side,  and 
passes  posteriad  above  the  dorsal  contour  of  the  expanded  stom- 
ach, only  descending  to  the  vent  posterior  to  the  terminus  of 
the  stomach.  In  all  the  inioms  I  have  examined  the  intestine 
arises  midventrally  and  passes  essentially  straight  back  to  the 
vent  along  the  midventral  wall  of  the  abdominal  cavity.  For  the 
time  being,  the  available  evidence  suggests  that  the  giganturids 
are  neoteleosts  (retractor  dorsalis  muscle),  allied  with  the  inioms 
(discrete  peritoneal  pigment  section),  diverging  early  from  the 
rest  and  acquiring  characters  making  them  among  the  most 
specialized  and  distinctive  of  teleosts. 

Field  Museum  of  Natural  History,  Roosevelt  Road  at 
Lake  Shore  Drive,  Chicago,  Illinois  60605. 


Basal  Euteleosts:  Relationships 
W.  L.  Fink 


AS  mentioned  in  the  introduction  to  this  section  of  the  sym- 
posium, the  order  Salmoniformes  has  had  a  history  of 
attrition,  such  that  today  I  would  recognize  it  as  coextensive 
with  the  Salmonidae.  Previously  included  taxa  are  now  scat- 
tered, primarily  as  unresolved  lineages  at  or  near  the  base  of 
the  Euteleostei.  What  follows  is  a  preliminary  analysis,  a  sketch 
of  alternative  hypotheses  of  interrelationships  of  the  basal  eu- 
teleosts. Fully  resolving  these  problems  will  take  more  time  and 
more  material  than  I  have  had  available  to  me,  and  I  hope  that 
work  stimulated  by  this  symposium  will  provide  insights  which 
have  not  been  forthcoming  using  traditional  material  and  char- 
acters. 

Unfortunately,  very  little  information  of  a  comparative  nature 
is  available  on  the  larvae  of  basal  euteleosts,  and  when  these 
larvae  have  been  discussed,  only  rarely  have  characters  or  char- 
acter transformations  useful  at  large  clade  levels  been  men- 
tioned. Since  adult  specimens  are  more  easily  available  in  most 
collections,  that  is  what  I  have  relied  on,  with  examination  of 
larvae  when  possible. 

Results 

The  Euteleostei  is  a  large  group  of  modem  teleosts  which  is 
poorly  diagnosed  in  terms  of  unique  traits,  and  most  more  phy- 
logenetically  advanced  members  lack  some  of  the  diagnostic 
characters.  Patterson  and  Rosen  (1977)  considered  the  following 
as  euteleostean  traits:  1)  an  adipose  fin,  2)  nuptial  tubercles,  and 
3)  an  anterior  membraneous  component  to  the  first  uroneural. 

Near  the  "base"  of  the  Euteleostei,  Fink  and  Weitzman(1982) 
recognized  several  lineages,  including  the  Esocoidei,  Ostario- 
physi,  Argentinoidei,  Osmeroidei,  Salmonidae,  and  Neoteleos- 
tei.  All  were  considered  monophyletic,  but  the  interrelations  of 
these  large  clades  were  left  unresolved  (Fig.  106).  Below  is  a 
review  of  each  of  the  groups,  with  new  information  included 
when  possible. 

Esocoidei  or  Esocae.—The^  fishes  have  been  a  continuing 
problem  for  ichthyologists.  They  are  considered  as  euteleosts 
on  the  basis  of  an  anterior  membraneous  component  to  the  first 
uroneural,  although  it  is  not  extensive.  No  esocoids  can  have 
an  adipose  fin  as  the  dorsal  fin  is  posteriorly  situated.  Neither 
do  they  have  breeding  tubercles.  Rosen  (1974)  provided  diag- 
nostic characters  documenting  monophyly  of  the  group.  Fink 
and  Weitzman  (1982)  suggested  that  esocoids  could  be  the  sister 
group  of  all  other  euteleosts  based  on  the  lack  in  the  latter  of  a 
toothplate  on  the  4th  basibranchial,  a  bone  which  is  present  in 
esocoids  and  other  primitive  teleosts  (see  those  authors  for  a 
discussion  of  the  distribution  of  this  character).  Wilson  and 
Veilleux  ( 1 982)  have  recently  reviewed  interrelationships  in  the 
Umbridae,  and  they  place  Umbra  and  Dallia  as  sister  taxa,  with 
Novumbra  as  their  sister  group;  all  these  together  are  placed  as 
the  sister  group  of  Esox.  This  corroborates  the  hypothesis  of 
Nelson  (1972). 

Rosen  (1974)  considered  Lepidogalaxias  to  be  a  member  of 
this  assemblage,  which  he  termed  the  Esocae.  Fink  and  Weitz- 


man (1982)  questioned  that  hypothesis,  leaving  the  genus  un- 
placed. I  have  further  comments  and  a  new  hypothesis  of  its 
relationships  below.  I  have  nothing  to  add  to  what  Fink  and 
Weitzman  (1982)  did  with  esocoids  sensu  stricto.  and  until  more 
is  forthcoming,  consider  them  the  likely  sister  group  to  other 
euteleosts. 

Ostariophysi— In  terms  of  numbers  of  species  and  morpholog- 
ical diversity,  this  is  the  dominant  basal  euteleostean  group. 
Fink  and  Weitzman  (1982)  did  not  consider  the  relations  of 
these  fishes  to  other  euteleosts,  primarily  because  their  survey 
was  intended  to  establish  the  placement  of  stomiiforms,  and 
there  was  no  evidence  suggesting  relationship  between  the  two 
groups.  No  phylogenetic  examination  of  ostariophysan  rela- 
tionships to  other  teleosts  has  been  done  since  Rosen  and  Green- 
wood (1970)  expanded  traditional  concepts  of  the  group  by 
adding  the  previously  protacanthopterygian  gonorynchifonns. 
Fink  and  Fink  ( 1 98 1 )  examined  relationships  within  the  group, 
placing  siluroids  and  gymnotoids  as  sister  taxa  (order  Siluri- 
formes),  these  the  sister  taxon  of  characiforms,  and  these  to- 
gether the  sister  group  of  cypriniforms  (the  Otophysi,  inclusive); 
sister  group  relationship  of  the  gonorynchiforms  to  the  Otophysi 
was  corroborated.  This  entire  assemblage  was  considered  mono- 
phyletic on  the  basis  of  numerous  characters,  including  lack  of 
a  dermopalatine,  unique  gasbladder  morphology,  specializa- 
tions of  the  vertebrae,  and  adductor  mandibulae  anatomy. 

Argentinoidei.— Cvetn'wood  and  Rosen  (1971)  combined  the 
alepocephaloid  and  argentinoid  fishes  into  an  expanded  Argen- 
tinoidei, in  the  Salmoniformes.  Fink  and  Weitzman  (1982)  agreed 
with  the  combination  of  the  two  groups  and  used  the  formal 
subordinal  name  to  include  both  subgroups.  However,  Fink  and 
Weitzman  (1982)  were  unable  to  provide  evidence  bearing  on 
relationships  of  these  fishes,  even  though  their  cladogram  (Fig. 
23,  Fig.  106  herein)  showed  them  as  the  sister  group  of  the 
osmeroids.  I  have  similarly  been  unable  to  place  them,  in  part 
because  of  lack  of  adequate  material. 

Osmeroidei— Thii  group,  which  includes  the  northern  and 
southern  smelts,  galaxiids  (here  including  Lovettia  and  Aplo- 
chiton),  Plecoglossus,  and  salangids,  can  be  diagnosed  as  mono- 
phyletic based  on  several  characters,  including  presence  of  one 
or  more  rows  of  teeth  near  the  medial  border  of  the  mesopter- 
ygoid,  loss  or  appearance  late  in  ontogeny  of  the  articular  bone, 
and  presence  of  a  foramen  in  the  posterior  plate  of  the  pelvic 
bone.  Some  subgroups  of  osmeroids  have  lost  various  of  these 
diagnostic  characters,  but  the  patterns  of  loss  allow  other  fea- 
tures to  provide  evidence  of  relationship  in  the  group. 

Nevertheless,  relationships  within  the  suborder  remain  prob- 
lematical. The  following  review  is  based  upon  examination  of 
specimens,  the  literature,  and  the  contributions  to  this  sym- 
posium. Incidentally.  I  have  not  attempted  to  diagnose  the  var- 
ious genera,  but  McDowall's  comments  (this  volume  and  1 969) 
indicate  that  such  needs  to  be  done.  The  phylogenetic  hypoth- 


202 


FINK:  BASAL  EUTELEOSTS 


203 


ESOCAE 


OSTARIOPHYSI 


ARGENTINOIDEI 


OSMEROIDEI 


NEOTELEOSTEI 


SALMONIDAE 

STOMIIFORMES 

AULOPIFORMES 

MYCTOPHIFORMES 

ACANTHOMORPHA 

Fig.  106.    The  hypothesis  of  relationships  suggested  by  Fink  and 
Weitzman  ( 1 982)  for  the  basal  euteleosts. 


eses  and  data  are  included  in  Fig.  107  and  its  caption.  The  data 
used  in  this  analysis  were  chosen  partly  because  they  have  been 
used  traditionally  in  osmeroid  systematics  but  I  have  little  con- 
fidence in  some  of  them;  as  a  result  this  analysis  represents  a 
preliminary  sketch  of  a  more  detailed  study. 

The  most  striking  thing  about  osmeroid  systematics  is  that 
we  still  have  questions  about  some  very  basic  things,  such  as 
the  status  of  the  Osmeridae.  As  noted  by  Nelson  (1970)  and 
Rosen  (1974),  no  evidence  has  ever  been  presented  that  the 
family  is  a  monophyletic  group.  Indeed,  it  seems  quite  possible 
that  Plecoglossus  could  be  more  closely  related  to  some  "os- 
merids"  than  to  others,  and  this  would  render  the  family  para- 
phyletic.  A  minimal  requirement  of  any  future  work  on  system- 
atics of  the  group  should  be  documentation  of  whether  it  is 
natural. 


Fig.  107.  Alternate  cladograms  of  relationships  within  the  Osme- 
roidei.  The  bottom  figure  represents  the  hypothesis  supported  when  all 
characters  are  given  equal  weight  and  paedomorphic  traits  are  consid- 
ered homologous.  The  top  figure  represents  the  hypothesis  which  con- 
siders the  paedomorphic  reductive  traits  of  salangids  and  galaxiids  as 
non-homologous.  For  discussion,  see  text. 

The  supporting  characters  are  listed  below,  with  the  derived  condition 
indicated  by  a  1 ,  the  primitive  by  a  0.  Each  character  number  is  indicated 
on  the  cladogram  where  it  is  in  the  derived  state.  Dark  squares  indicate 
unique  appearance  of  a  trait;  empty  squares  indicate  multiple  appear- 


18-20 


6ALAXIIDAE 

lovettia 

Aplochiton 

Retropinna 

Stokellia 

Prototroctes 

SALANGIDAE 

Plecoglossus 
"OSMERIDAE" 


2>29     -J-.GALAXIIDAE 

lovettia 

'Aplochiton 

■SALANGIDAE 

-Retropinna 

•Stokellia 

Prototroctes 

Plecoglossus 

'OSMERIDAE" 

ance  of  a  trait;  triangles  indicate  a  trait  that  is  reversed  at  a  lower  level 
of  generality;  and  circles  indicate  those  characters  in  the  reversed  state. 
1 .  Posterior  shaft  of  vomer  (0)  long  ( 1 )  shori.  2.  Articular  bone  (0) 
present  and  fused  with  angular  (1)  absent  or  greatly  reduced.  3.  Meso- 
pterygoid  teeth  (0)  over  much  of  bone  ventral  surface  (1)  restricted  to 
medial  border  of  ventral  surface  or  lacking.  4.  Pelvic  foramen  (0)  absent 
( 1 )  present.  5.  Anchor  membrane  of  egg  (0)  absent  ( 1 )  present.  6.  Caudal 
skeleton  fusion  patterns  (0)  none  or  rudimentary  neural  arches  fusing 
with  centrum  and  then,  if  at  all,  to  the  uroneural  ( 1 )  rudimentary  neural 
arch  fusing  with  uroneural  first,  then  these  to  the  centrum.  7.  Infraorbital 
sensory  canals  (0)  curved  posterodorsally  ( 1 )  curved  posteroventrally. 
8.  Mesocoracoid  (0)  present  ( 1 )  absent.  9.  Dorsal  fin  position  (0)  forward 
(1)  posterior.  10.  Principal  caudal  fin  rays  (0)  10/9(1)  9/9  or  fewer.  1  1. 
Palatine  teeth  (0)  present  (1)  absent.  12.  Ectopterygoid  bone  (0)  present 
( 1 )  absent.  1 3.  Extrascapular  (0)  present  ( I )  absent.  14.  Coracoid-cleith- 
rum  process  (0)  present  (1)  absent.  15.  Posterior  pubic  symphysis  (0) 
present  ( 1 )  absent.  1 6.  Scales  (0)  present  ( 1 )  absent.  1 7.  Vomerine  teeth 
(0)  present  (1)  absent.  18.  Posterior  border  of  bones  of  suspensorium 
(0)  smooth  (1)  deeply  incised  or  emarginate.  19.  Principal  caudal  fin 
rays  (0)  9/9(1)  8/8.  20.  Hypural  number  (0)  6  (1)  5.  21.  Infraorbital 
sensory  canals  (0)  not  extending  to  preopercle  (1)  extending  to  pre- 
opercle.  22.  Ceratohyal  ventral  border  (0)  more  or  less  straight,  bran- 
chiostegals  along  most  of  its  length  ( 1 )  deeply  concave  anterioriy,  bran- 
chiostegals  restricted  to  area  posterior  to  concavity.  23.  Homy  abdominal 
keel  (0)  not  present  ( 1 )  present.  24.  Ovaries  (0)  both  present  ( 1 )  left  only. 

25.  Ectopterygoid  bone  (0)  posterior  to  autopalatine  (1)  ventral  to  au- 
topalatine  (coded  as  present  in  Stokellia  based  on  McDowall,  1969). 

26.  Cucumber  odor  (0)  absent  (1)  present.  27.  Basioccipital  lateral  pegs 

(0)  none  (1)  present.  28.  Lateral  hyomandibular  spur  (0)  not  present  (1) 
present.  29.  Caudal  fin  posterior  border  (0)  deeply  forked  (1)  rounded 
or  emarginate.  30.  Adipose  fin  (0)  present  ( 1 )  absent.  3 1 .  Mesopterygoid 
teeth  (see  also  Character  3)  (0)  restricted  to  ventromedial  area  of  bone 

( 1 )  absent. 


204 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Salangids  have  been  associated  in  the  past  with  various  mem- 
bers of  the  osmeroid  assemblage,  but  even  this  was  questioned 
by  Nelson  (1970).  Rosen  (1974)  presented  evidence  from  the 
caudal  skeleton  which  shows  that  salangids  are  osmeroids,  but 
no  evidence  about  their  placement  within  the  group  has  been 
presented  to  date.  Fink  and  Weitzman  (1982)  agreed  with  Rosen 
and  placed  the  Salangidae  as  incertae  sedis  in  the  Osmeroidei. 

What  little  evidence  I  have  been  able  to  find  about  the  rela- 
tionships of  salangids  is  equivocal.  If  examined  by  a  standard 
parsimony  procedure,  as  represented  by  the  Wagner  analysis 
shown  in  Fig.  107  (bottom),  the  numerous  reductive  traits  of 
salangids  place  them  within  the  "southern  smelt"  plus  galaxiid 
assemblage.  On  the  other  hand,  salangids  share  with  Plecoglos- 
siis  and  the  "osmerids"  a  complex  caudal  skeleton  character 
involving  fusion  of  uroneural  1  to  a  compound  centrum  made 
upofPUl,  Ul,and  U2,  followed  ontogenetically  in  some  forms 
by  fusion  of  rudimentary  neural  arches  with  the  uroneural  por- 
tion of  the  complex.  This  latter  character  is  in  contrast  to  the 
autogenous  uroneurals  of  most  galaxiids,  the  "southern  smelts," 
and  other  primitive  teleosts.  Further,  when  uroneurals  and  ru- 
dimentary neural  arches  are  fused  in  galaxiids,  the  fusion  se- 
quence is  rudimentary  neural  arch  to  the  compound  centrum, 
followed  by  fusion  with  the  uroneural,  rather  than  the  reverse. 
The  hypothesis  that  emerges  from  these  observations  is  illus- 
trated in  Fig.  107  (top),  showing  salangids,  Plecoglossus.  and 
"osmerids"  in  an  unresolved  trichotomy.  For  further  discussion 
of  caudal  fin  morphology,  see  Greenwood  and  Rosen  (1971), 
Rosen  (1974),  and  Fink  and  Weitzman  (1982). 

Any  choice  of  these  alternate  hypotheses  of  salangid  relation- 
ships would  rest  on  whether  or  not  one  wished  to  accept  the 
numerous  reductive  traits  that  unite  the  salangids  with  the 
"southern  smelts"  and  galaxiids  as  homologues.  Such  choice  is 
based  on  criteria  which  cannot  be  discussed  in  detail  at  this 
point  due  to  space  restrictions,  but  I  have  commented  elsewhere 
(Fink,  1982)  on  hypothesis  choice  forced  by  confrontation  with 
apparent  paedomorphosis.  In  this  case,  for  example,  some  of 
the  general  morphological  attributes  that  salangids  share  with 
the  members  of  those  groups  differ  when  examined  in  detail. 
Although  this  lack  of  close  correspondence  in  similarity  is  cer- 
tainly no  guarantee  that  the  reductions  are  not  homologous,  it 
does  raise  the  issue.  Further,  the  highly  developed  caudal  skel- 
eton of  salangids  is  identical  to  that  of  "osmerids,"  and  thus 
more  differentiated  than  that  of  either  the  southern  smelts  or 
galaxiids.  This  incongruity  in  degree  of  morphological  differ- 
entiation suggests  that  in  this  case,  one  should  be  cautious  in 
assuming  homology  in  the  reductive  process  and  search  for 
other,  non-reductive  characters  to  resolve  possible  misplace- 
ments. 

The  family  Sundasalangidae  is  not  accepted  herein  because 
in  every  case  in  which  Roberts  (1981)  contrasted  sundasalangids 
and  salangids,  the  character  for  salangids  was  primitive.  I  suggest 
that  recognition  of  family  rank  for  Sundasalanx^^oviXd  probably 
render  the  Salangidae  paraphyletic  and  thus  defined  only  by  the 
absence  of  characters  present  in  Sundasalanx.  This  is  unac- 
ceptable both  because  it  forces  recognition  of  a  group  based  on 
characters  its  members  lack  and  because  it  artificially  breaks  up 
a  group  all  of  whose  members  share  a  unique  evolutionary  his- 
tory. 

Regarding  the  "southern  smelt  assemblage"  (including  gal- 
axiids, but  excluding  salangids),  I  am  less  pessimistic  than 
McDowall  (this  volume).  I  have  taken  the  liberty  of  using  the 
data  he  has  presented  and  combined  them  with  my  own  limited 


survey  of  specimens  and  the  literature  to  produce  the  hypotheses 
shown  in  Fig.  107.  The  group  can  be  diagnosed  by  presence  of 
a  posteroventral  deflection  of  the  infraorbital  sensory  canal  (Nel- 
son, 1972)  and  9/9  or  fewer  principal  caudal-fin  rays  (vs  a  pos- 
terodorsal  curvature  of  the  canal  and  10/9  rays  in  outgroups). 
Several  characters  support  the  placement  of  Retropinna  and 
Prototrocles  as  sister  taxa  including  presence  of  an  abdominal 
homy  keel,  loss  of  the  right  ovary,  and  ceratohyal  morphology. 
I  have  no  specimens  o( Stokellia  on  hand,  but  McDowall's  work 
( 1979)  clearly  shows  that  the  genus  is  diagnosable  and  that  it  is 
related  to  Retropinna  and  Prototrocles.  Unfortunately,  when 
contrasted  with  Stokellia,  it  is  not  clear  that  Retropinna  is  di- 
agnosable, since  the  latter  is  then  differentiated  by  primitive 
characters  present  in  other  taxa. 

Relationship  among  Aplochiton,  Lovettia  and  the  galaxiids  is 
supported  by  numerous  characters,  as  shown  in  Fig.  107.  I  have 
been  unable  to  find  any  features  that  link  the  former  two  genera 
together,  however,  and  more  work  needs  to  be  done  with  them. 
Galaxiids  themselves  can  be  shown  to  be  monophyletic  based 
on  such  characters  as  basioccipital  "pegs"  extending  lateral  to 
the  anterior  centrum  (McDowall,  1969,  Figs.  2B,  lOA,  but  note 
lack  of  "pegs"  in  G.  paucispondylus.  Fig.  1  OB). 

In  summary,  it  is  suggested  that  the  broad  outlines  of  rela- 
tionships among  the  osmeroids  are  beginning  to  emerge,  much 
as  suggested  by  Gosline  (1960a),  with  a  "southern  smelt"  as- 
semblage and  an  "osmerid"  assemblage.  Interrelationships  within 
these  groups  remain  problematical,  the  most  obvious  problems 
being  establishment  of  the  natural  groups  within  the  "osmerids" 
and  placement  of  the  salangids. 

Salmontds.  —  M.onox>\\y\y  of  this  group  is  based  primarily  on  a 
single  character,  apparent  polyploidy  of  the  karyotype  (Gold, 
1979).  Several  investigators  have  studied  interrelationships  of 
salmonids,  most  notably  Behnke  (1968)  and  Norden  (1961),  but 
these  works  were  not  phylogenetic  and  changes  can  be  expected. 
I  have  examined  phylogeny  within  the  group  only  to  establish 
polarities  for  characters  relevant  to  relationships  with  other  te- 
leosts. Regarding  the  latter  relationships,  there  have  been  several 
opinions,  with  most  workers  approaching  salmonids  with  an 
eye  to  finding  ancestors  ofother  groups  (see,  e.g.,  Gosline,  1960, 
Diagram  2).  The  only  phylogenetic  analysis  to  date  is  that  of 
Rosen  (1974),  which  was  discussed  by  Fink  and  Weitzman 
( 1 982).  The  latter  authors  presented  data  which  they  considered 
suggestive  of  neoteleostean  relationship  for  salmonids:  presence 
in  some  members  of  paired  cartilages  anterior  to  the  ethmoid 
region  (resembling  the  median  rostral  cartilage  of  neoteleosts) 
and  the  exoccipital  forming  part  of  the  occipital  condyle.  The 
anterior  cartilages  were  reported  by  Fink  and  Weitzman  (1982) 
to  be  prominent  in  Prosoplum,  an  observation  which  I  can 
confirm  from  additional  specimens.  In  addition,  examination 
ofsmall  juvenile  cichlids  shows  that  the  rostral  cartilage  appears 
to  develop  ontogenetically  from  bilateral  cartilage  bodies  which 
fuse  at  the  midline;  this  is  suggestive  of  corroboration  of  Fink 
and  Weitzman's  (1982)  hypothesis  that  the  rostral  cartilage 
evolved  from  paired  cartilages  anterior  to  the  ethmoid  region 
like  those  in  Prosoplum.  More  work  needs  to  be  done  on  the 
homology  of  "accessory"  ethmoid  cartilages,  using  double  stain- 
ing techniques  and  histology  on  a  wide  variety  of  teleosts. 

1  can  also  add  to  what  Fink  and  Weitzman  ( 1 982)  noted  about 
the  occipital  condyle.  1  have  confirmed  that  the  exoccipital  forms 
part  of  the  condyle  in  Thymallus  and  "salmonins."  This  mor- 
phology is  also  present  in  Prosoplum.  but  is  lacking  in  other 


FINK:  BASAL  EUTELEOSTS 


205 


coregonins.  In  a  number  of  features,  including  the  morphology 
of  the  nares,  Prosopium  stands  as  the  sister  group  of  other  cor- 
egonins, and  this,  plus  the  presence  in  the  outgroup  Salmoninae 
and  Thymallus  of  exoccipital  participation  in  the  condyle,  im- 
plies that  phylogenetically  derived  coregonins  have  secondarily 
lost  that  morphology.  As  noted  by  Fink  and  Weitzman  (1982), 
the  condyle  structure  as  found  in  salmonids  is  found  also  in 
neoteleosts.  It  is  also  present  in  Lepidogalaxias  (see  below)  and 
in  some  osteoglossomorphs.  I  do  not  wish  to  belabor  the  possible 
importance  of  this  character,  especially  since  more  careful  on- 
togenetic and  morphological  studies  need  to  be  done  and  other 
characteristics  evaluated. 

A  few  observations  from  my  survey  of  salmonids  may  be 
added  here.  I  have  found  but  two  characters  in  the  literature 
which  diagnose  the  coregonins;  one  of  these  needs  modification 
and  the  other  needs  to  be  more  concisely  put.  Lack  of  maxillary 
teeth  has  been  used  to  diagnose  the  group,  relative  to  other 
salmonids  (Norden,  1 96 1 ),  but  this  needs  to  be  emended  to  lack 
of  the  teeth  in  adults,  since  I  have  found  maxillary  teeth  in 
Prosopium  of  around  19  mm  SL.  I  have  not  yet  examined  spec- 
imens this  small  of  other  coregonins  so  do  not  know  the  gen- 
erality of  this  primitive  state.  The  other  character  is  reduction 
in  the  teeth  in  general;  this  needs  to  be  quantified  relative  to 
the  outgroups. 

The  salmonins  and  Thymallus  can  be  placed  together  based 
on  lack  of  ossification  of  the  supraethmoid  (hypethmoid  of  Nor- 
den, 1961;  Behnke,  1968),  and  apparently  on  yolk  character- 
istics, and  larval  size  (Kendall  and  Behnke,  this  volume).  Re- 
garding other  relationships  within  salmonids,  I  have  nothing  to 
add. 

Lepidogalaxias.— The  position  of  Lepidogalaxias  is  controver- 
sial. I  remain  unconvinced  by  Rosen's  (1974)  hypothesis  that 
the  genus  belongs  with  the  esocoids.  When  I  previously  dis- 
cussed this  genus  (Fink  and  Weitzman,  1982),  I  had  not  seen 
any  specimens,  but  R.  M.  McDowall  has  generously  made  sev- 
eral available  for  dissection  and  clearing  and  staining.  There  is 
no  question  that  this  little  fish  is  a  potpourii  of  contradictory 
and  reductive  characters  and  it  is  no  wonder  that  it  has  been  so 
difficult  to  place.  Pursuing  the  potential  of  relationship  of  this 
species  to  galaxiids,  extensive  comparisons  with  members  of 
that  group  have  been  made.  Lepidogalaxias  shares  a  host  of 
reductive  characters  with  galaxiids.  NVhile  these  may  indeed  be 
synapomorphous  traits,  in  cases  where  extensive  paedomor- 
phosis  is  suspected,  and  this  appears  to  be  so  in  the  morpho- 
logical similarities  involved,  one  hopes  to  find  some  innovative, 
non-reductive  characters  which  supply  evidence  for  grouping. 
I  have  found  two  such  characters  which  suggest  that  Lepido- 
galaxias is  related  to  neither  esocoids  nor  osmeroids,  but  rather 
may  be  the  sister  group  of  the  Neoteleostei,  as  diagnosed  by 
Rosen  (1973)  and  Fink  and  Weitzman  ( 1 982).  This  is  supported 
by  the  presence  in  Lepidogalaxias  of  two  non-reductive  traits, 
a  retractor  dorsalis  muscle  and  occipital  condyle  composed  of 
both  the  basioccipital  and  exoccipital  bones.  As  discussed  just 
above  and  by  Fink  and  Weitzman  ( 1 982),  the  latter  trait  is  also 
shared  with  salmonids.  Lepidogalaxias  lacks  a  rostral  cartilage 
or  its  homologue  and  type  4  teeth  (hinged  teeth  with  a  posterior 
axis  of  rotation.  Fink,  1981)  and  this  would  prevent  its  place- 
ment within  the  neoteleostean  assemblage.  Placing  Lepidoga- 
laxias as  the  neoteleostean  sister  group  and  leaving  salmonids 
as  their  sister  taxon  presumes  either  that  rostral  cartilage  homo- 
logues  in  the  salmonids  have  been  lost  in  Lepidogalaxias  or  are 


EUTELEOSTEI 


NEOTELEOSTEI 


EURYPTERYGII 


OSTEOGLOSSOMORPHA 
ELOPOMORPHA 
CLUPEOMORPHA 
ESOCOIDEI 
OSTARIOPHYSI 
ARGENTINOIDEI 
OSMEROIDEI 
SALMONIDAE 

LEPIDOGALAXIAS 

STOMIIFORMES 

AULOPIFORMES 

MYCTOPHIFORMES 

ACANTHOMORPHA 


Fig.  108.     Summary  cladogram  of  relationships  and  characters  dis- 
cussed in  the  text. 


not  homologues  after  all.  This  ambiguity  is  reflected  in  Fig.  108 
by  a  trichotomy.  Clearly,  more  work  remains  to  be  done  before 
we  can  be  really  confident  in  the  phylogenetic  placement  of  this 
intriguing  fish. 

Lepidogalaxias  can  be  diagnosed  by  a  number  of  characters, 
the  most  striking  of  which  is  fusion  of  the  frontal  bones  into  a 
single  ossification  (Rosen,  1974,  Fig.  40B).  In  their  comments 
on  this  species.  Fink  and  Weitzman  ( 1 982)  noted  that  there  was 
a  disagreement  about  whether  there  are  mesopterygoid  teeth 
present;  Rosen's  statement  that  teeth  are  lacking  is  cortect. 

Stomiiformes. —  Vmk  and  Weitzman  (1982)  recently  examined 
the  monophyly  and  relationships  of  stomiiforms  to  the  other 
basal  euteleosts  and  corroborated  Rosen's  (1973)  hypothesis 
that  they  are  the  sister  group  to  the  rest  of  the  Neoteleostei, 
removing  them  from  the  "salmoniforms."  This  placement  is 
supported  by  several  apomorphic  traits,  including  presence  of 
retractor  dorsalis  muscles  and  type  4  tooth  attachment,  as  well 
as  exoccipital  participation  in  the  cranial  condyle  and  a  rostral 
cartilage.  Weitzman  (1974)  presented  a  hypothesis  of  relation- 
ships at  the  "family"  level  within  the  stomiiforms,  as  well  as  a 
detailed  phylogeny  of  the  Stemoptychidae.  In  this  volume,  I 
present  a  generic-level  phylogeny  for  the  barbeled  stomiiforms 
(Family  Stomiidae)  and  some  brief  comments  on  the  "gonosto- 
matid-photichthyid"  genera.  Weitzman  is  currently  working  on 
relationships  of  the  latter  fishes  and  has  made  considerable  com- 
ments in  this  volume  (see  Ahlstrom,  Richards  and  Weitzman, 
this  volume). 


206 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Eurypterygii.  —  FinaWy,  a  few  comments  are  due  on  the  Myc- 
tophoidei  of  Greenwood  et  al.  (1966).  This  group  was  disman- 
tled by  Rosen  (1973),  and  divided  into  two  large  groups,  Au- 
lopiformes  and  Myctophiformes.  These  two  groups,  together 
with  the  Paracanlhopterygii  and  Acanthopterygii,  were  classified 
into  a  new  group,  Eurypterygii.  Aulopiformes  was  placed  as  the 
sister  group  of  all  other  eurypterygians,  and  myctophiforms  as 
the  sister  group  to  paracanthoptergyians  and  acanthopterygians. 
All  of  these,  together  with  stomiiforms,  form  the  Neoteleostei. 
Fink  and  Weitzman  (1982)  tentatively  accepted  monophyly  of 
the  Eurypterygii  based  on  the  presence  in  its  members  of  a 
toothplate  fused  with  the  third  epibranchial.  Aulopiformes  con- 
tains a  large  number  of  families,  including  the  Giganturidae, 
covered  in  this  portion  of  the  symposium.  About  the  latter 
family  I  have  little  to  say  except  that  my  own  dissections  cor- 
roborate Rosen's  placement  of  it. 

Summary 

A  summary  of  the  hypotheses  I  have  discussed  above  is  given 
in  Fig.  108.  The  most  striking  aspect  of  it  is  the  degree  of  un- 
certainty about  relationships  among  the  clades.  This  may  be  in 
part  due  to  the  limitations  of  my  study,  but  it  does  seem  to  me 
to  be  a  fair  summary  of  the  status  of  well  corroborated  hypoth- 
eses we  now  have  about  this  level  of  teleostean  phylogeny.  There 
are  certainly  other  arrangements  that  can  be  made,  depending 
on  which  characters  one  wishes  to  stress,  and  none  of  these 
should  be  discarded  out  of  hand.  As  examples,  I  will  cite  two 
characters  and  their  implications. 

First,  lack  of  the  posterior  shaft  of  the  vomer  suggests  that 
salmonids  and  osmeroids  are  sister  taxa.  Appropriate  outgroups 
have  the  shaft  ranging  from  "moderate"  (e.g.,  Chanos)  to  "elon- 
gate" (argentinoids).  My  own  opinion,  based  on  occipital  con- 
dyle structure  of  salmonids,  is  that  the  reduction  in  vomer  length 
has  occurred  independently  in  the  two  lineages  (it  has  also  been 
reversed  within  both);  the  ultimate  value  of  the  occipital  char- 
acter remains  to  be  seen. 

The  second  character,  presence  of  breeding  tubercles,  is  now 
considered  a  euteleostean  trait.  Note,  however,  that  tubercles 


are  lacking  in  esocoids  and  argentinoids  but  are  present  in  os- 
tariophysans,  osmeroids,  and  salmonids,  indicating  that  these 
three  clades  form  a  monophyletic  group.  Again,  there  are  char- 
acters that  contradict  this  grouping,  but  it  nevertheless  is  worthy 
of  consideration. 

It  is  always  frustrating  when  one  sets  out  to  solve  a  particular 
problem  and  then  comes  to  the  end  of  the  allotted  time  without 
a  resolution.  Although  I  have  been  able  to  shed  some  light  on 
several  problems  relevant  to  the  goals  of  this  part  of  the  sym- 
posium, I  have  not  been  able  to  unravel  the  interrelationships 
among  the  major  basal  euteleostean  clades.  Clearly  more  work 
is  needed,  especially  with  character  suites  which  have  been  tra- 
ditionally neglected.  Almost  all  of  our  concepts  of  relationships 
at  this  level  are  based  on  features  of  the  adult  caudal  skeleton 
and  branchial  basket.  Some  work  on  soft  anatomy,  particularly 
the  muscles  of  the  head,  has  been  informative  at  these  levels 
and  one  hopes  that  other  parts  of  the  soft  anatomy  will  be  equally 
profitable.  One  area  virtually  untouched  is  larval  anatomy.  It 
might  be  expected  that  not  many  important  features  will  be 
found  because  of  the  preponderance  of  primitive  characters  in 
larvae.  But  larval  characters  have  proven  useful,  as  is  shown  by 
the  ontogenetic  transformation  in  tooth  types  in  stomiiforms 
(from  type  4  to  type  3;  see  Fink,  1981)  as  well  as  the  specialized 
fin  traits  discussed  by  Ahlstrom  et  al.  (this  volume)  for  argen- 
tinoids. It  is  in  both  these  areas,  ontogenetic  character  trans- 
formations and  presence  of  specializations  for  larval  life,  that 
study  of  larval  fishes  promises  rewards.  The  inclusion  of  larval 
morphology  in  studies  of  higher  level  relationships  should  pro- 
vide a  richer  data  base  than  we  currently  have  and  perhaps  will 
reveal  some  crucial  characters  for  resolving  the  basic  questions 
I  have  addressed  above.  This  symposium  has  already  stimulated 
in  a  major  way  the  examination  of  larvae  for  phylogenetic  anal- 
yses, and  I  predict  that  it,  combined  with  the  new  ways  now 
emerging  of  analyzing  ontogenetic  information,  will  mark  a  new 
phase  in  the  modem  study  of  fish  classification. 

Museum  of  Zoology,  University  of  Michigan,  Ann  Arbor, 
Michigan  48109. 


Myctophiformes:  Development 
M.  Okiyama 


MYCTOPHIFORMES  is  currently  adopted  as  a  distinct 
order  with  intermediate  affinity  between  the  lower  and 
higher  teleost  groups,  whereas  no  one  feature  would  satisfac- 
torily separate  all  of  them  from  all  Salmoniformes  (Gosline  et 
al.,  1966).  Except  Rosen  (1973),  recent  workers  agree  well  with 
the  familial  composition  of  this  order  despite  slight  differences 
in  the  familial  or  subordinal  definition. 

Table  56  shows  the  recent  classification  given  by  Johnson 
(1982)  based  on  the  most  comprehensive  knowledge  now  avail- 
able. Important  points  of  this  scheme  are  the  exclusion  of  Sco- 
pelarchidae  from  Alepisauroidei  and  Pseudotrichonotidae  from 
Myctophiformes.  Further  details  in  this  connection  will  be  men- 
tioned again  in  my  paper  on  relationships  (this  volume). 


Exploitation  of  the  vast  hydrosphere  covering  the  pelagic  as 
well  as  benthic  habitat  between  the  surface  and  abyssal  or  ul- 
traabyssal  plain  by  diversified  members  of  this  group  is  doubt- 
lessly the  important  aspect  in  discussing  the  ontogenetic  prob- 
lems of  the  myctophiform  lineage.  Of  the  five  suborders, 
Myctophoidei  and  Alepisauroidei  are  exclusively  pelagic  and 
the  remaining  are  demersal  including  secondary  pelagic  genera 
such  as  Parasudis  and  Harpadon.  Synchronous  hermaphrodit- 
ism is  common  to  the  deep-water  and  offshore  forms  belonging 
to  Chlorophthalmoidei  and  Alepisauroidei  with  the  single  ex- 
ception of  Bathysauridae  in  Synodontoidei  (Table  56). 

In  general,  the  systematics  of  this  order  are  rather  well  under- 
stood except  for  several  families  or  genera.  As  is  clearly  shown 


OKIYAMA:  MYCTOPHIFORMES 


207 


Table  56.    Systematic  Status  and  the  Current  Knowledge  on  Early  Life  Stages  in  Myctophiformes. 


Suborder  and  family 


No. 

Reproduc-  _ 
tion'' 

Information 

species 

Eggs             Larvae 

Main  sources 

7  + 

G 

_(.C                         +_|.d 

Okiyama  (1974b) 

Aulopoidei 
Aulopidae 

Myctophoidei 
Myctophidae" 
Neoscopelidae 


Chlorophthalmoidei 
Chlorophthalmidae 


Ipnopidae 


Notosudidae 

Scopelarchidae" 
Synodontoidei 
Balhysauridae 
Harpadontidae 

Synodontidae 

Alepisauroidei 
Alcpisauridae 
Anotopteridae 
Evermannellidae" 
Omosudidae 
Paralepididae 


Aulopus 


Diaphus.  etc. 
Neoscopelus 
Scopelengys 
Solivomer 


Chlorophthalmus 

Parasudis 

Bathysauropsis 

Ipnops 

Bathytyphlops 

Bathymicrops 

Bathypterois 

Ahliesaurus 

Scopelosaurus 

Luciosudis 

Scopelarchus,  etc. 

Bathysaurus 
Harpadon 
Saurida 
Synodus 

Trachinocephalus 

Alepisaurus 

Anotopterus 

Evermannella.  etc. 

Omosudis 

Paralepis 

Notolepis 

Mautichthys 

Lestidium 

Lestidiops 

Unasudis 

Lestrolepis 

Stemonosudis 

Macroparalepis 

Dolichosudis 

Sudis 


Ca.  300 
3 
2 
1 


18  + 
2 
3 
3 

2 
2 

18 

■> 

13 
1 

17 

2 

4 

15 

Ca.  30 

1 


2 
1 
7 
1 
5 
3 
1 
4 

20 
4 
3 

13 
7 
1 
2 


G 
G 

7 

7 


H 
H 

7 

H 
H 
H 
H 
H 
H 
H 
H 


H 
G 
G 
G 
G 


H 
H 
H 
H 
H 
H 
H 
H 
H 
H 
H 
H 
H 
H 
H 


+ 
(+) 


(  +  ) 
(  +  ) 
(  +  ) 
(  +  ) 
(  +  ) 

(  +  ) 


+ 
+ 
+ 
+ 


+  +         Moser  and  Ahlstrom  (1970,  1974) 
+  Okiyama  (1974b) 

+  Okiyama  ( 1974b),  Butler  and  Ahlstrom  ( 1 976) 


+ 
+  + 

+ 

+  + 
+  + 
+  + 
+  + 
+  + 

+ 

+ 

+  + 
+  + 

+  + 


+  + 
+  + 
+  + 
+  + 
+  + 
+  + 

+  + 
+  + 

+ 

+  + 
+  + 
+  + 

+  + 


Tuning  (1918) 


Okiyama  (1981) 

Okiyama  (1972),  Parin  and  Belyamna  (1972) 

Okiyama  (this  study) 

Sanzo  (1938b).  Okiyama  ( 1 974b) 

Bertelsenet  al.  (1976),  Ozawa  (1978) 

Bertelsen  et  al.  (1976),  Ozawa  (1978) 

Bertelsen  et  al.  (1976) 

Johnson  (1974b,  1982) 

Marshall  (1961),  Rosen  (1971).  Johnson  (1974a) 
Okiyama  (1979b) 

Mito  (1961a),  Okiyama  (1974b).  Ozawa  (1983) 
Gibbs(1959),  Okiyama  (1974b),  Ozawa  (1983) 
Okiyama  (1974b) 

Rofen  (1966b) 

Okiyama  (this  study) 

Johnson  (1982) 

Ege  (1958).  Rofen  (1966b),  Belyanina  (1981) 

Ege  (1930,  1957),  Rofen  (1966a) 

Rofen  (1966a) 

Rofen  (1966a) 
Rofen  (1966a) 
Rofen  (1966a) 
Rofen  (1966a) 
Rofen  (1966a) 
Rofen  (1966a) 

Sanzo  (1917).  Rofen  (1966a),  Shores  (1969), 
Belyamna  (1981) 


■  For  the  details,  see  relevant  section.  ""G:  gonochonsm;  H:  hermaphroditism, 
early  developmental  stages  is  available  at  least  for  a  single  species. 


Parentheses  indicate  information  available  for  transparent  ovanan  eggs.  '^  Double  crosses  mean  that  a  series  of 


in  Table  56.  information  on  the  reproduction  and  development 
is  abundant  even  for  the  deep-water  species  contrary  to  the 
situation  of  about  20  years  ago  (Gosline  et  al.,  1966).  General 
larval  characteristics  of  this  order  were  summarized  by  Ahl- 
strom and  Moser  ( 1 976).  Selected  meristic  characters  including 
many  original  data  are  given  in  Table  57. 

Aulopidae  (Fig.  I09A-B).— This  bottom-fish  family  is  generally 
considered  the  most  primitive  representative  of  the  order.  Its 
systematics  are  inadequately  known;  at  least  seven  nominal  and 
two  undescribed  species  (Yamakawa,  pers.  comm.)  occur  in  the 
warm  waters  of  the  world  except  for  the  Indian  Ocean. 

Complete  early  life  history  series  including  egg  stages  are  known 
only  for  Aulopus  japonicus  (Okiyama.  1974b,  1980).  Fragmen- 
tary larval  accounts  are  also  available  for  some  unidentifiable 


species.  Suggested  dichotomy  in  the  larval  morphology  in  this 
family  (Okiyama,  1974b)  is  apparently  wrong  due  to  the  erro- 
neous identification  of  the  early  stages  oi  " Aulopus  filamento- 
sus"  in  Sanzo  ( 1 938b)  and  TSning  (1918),  which  are  now  ascribed 
to  Bathypterois  of  the  Ipnopidae. 

Eggs  of  .4.  japonicus  are  spherical  (1.18-1.14  mm  in  diame- 
ter), pelagic,  transparent,  without  an  oil  globule,  and  with  ir- 
regularly raised  meshes  on  the  chorion  surface.  Similar  features 
are  not  present  in  the  matured  ovarian  eggs  of  A.  filamentosus 
( 1.36-1 .44  mm  in  diameter)  with  numerous  oil  globules  (Sanzo, 
1938b).  The  known  larvae  differ  in  gut  structure,  size  of  the 
prominent  pigment  section  and  relative  width  of  the  slightly 
narrow  eyes.  However,  the  followmg  features  are  shared  in  com- 
mon: single  prominent  peritoneal  pigment  section  located  at  the 
middle  or  slightly  anterior  region  of  the  body;  gently  curved 


208 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  57.    Selected  Meristic  Characters  of  Myctophiform  Genera. 


Suborder  and  family 

Genus 

Dorsal 

Anal 

Pectoral 

Pelvic 

Branchiostegals 

Vertebrae 

Aulopoidei 

Aulopidae 

Aulopus 

14-22 

8-14 

11-14 

9 

10-17 

36-53 

Myctophoidei 

Mytophidae* 

Diaphus,  etc. 

10-26 

12-27 

0-22 

8 

6-12 

28-45 

Neoscopelidae 

Neoscopelus 

11-13 

10-13 

15-19 

8-9 

8-9 

30-31 

Scopelengys 

11-13 

12-14 

12-17 

7-8 

8 

29-35 

Solivomer 

12-14 

9-11 

14-16 

8 

9-11 

40-41 

Chlorophthalmoidei 

Chlorophthalmidae 

Chlorophthatmus 

9-13 

7-11 

15-19 

8-9 

8 

40-50 

Parasudis 

10 

8-9 

17 

9 

8 

38-39 

Balhysauropsis 

10-12 

10-11 

17-24 

9 

8-9 

44-56 

Ipnopidae 

Ipnops 

8-11 

11-19 

12-16 

8 

9-12 

54-61 

Balhytyphtops 

11-13 

13-17 

12-15 

8 

14-17 

62-66 

Balhymicrops 

8-10 

9-15 

9-10 

7-8 

8-10 

65-80 

Bathyplerois 

12-16 

7-13 

13-22 

8-9 

10-14 

49-65 

Notosudidae 

Ahliesurus 

9-11 

17-21 

10-12 

9 

10 

42-50 

Scopelosaurus 

9-13 

15-21 

10-15 

9-10 

10 

53-67 

Luciosudis 

10-13 

17-20 

12-14 

9-10 

10 

57-59 

Scopelarchidae* 

Scopelarchus.  etc. 

5-10 

17-39 

18-28 

9 

8 

40-65 

Synodontoidei 

Bathysauridae 

Bathysaurus 

15-18 

11-14 

15-17 

7-8 

8-12 

50-63 

Harpadontidae 

Harpadon 

10-15 

11-15 

11-13 

9 

16-26 

39-56 

Sauhda 

10-13 

9-13 

11-16 

9 

13-16 

43-67 

Synodontidae 

Synodus 

10-15 

8-15 

10-15 

8 

12-18 

49-65 

Trachinocephalus 

11-13 

14-16 

11-13 

8 

14 

54-58 

Alepisauroidei 

Alepisauridae 

Alepisaurus 

29-49 

11-19 

12-16 

7-10 

7 

47-51 

Anotopteridae 

Anotoplerns 

0 

14-16 

12-15 

9-11 

8 

78-83 

Evermannellidae* 

Evermannelta.  etc. 

10-13 

26-37 

11-13 

9 

8 

45-54 

Omosudidae 

Omosudis 

9-12 

14-16 

11-13 

8 

8 

39-41 

Paralepididae 

Paralepis 

9-12 

20-26 

14-17 

8 

8 

60-77 

Notolepis 

8-11 

23-34 

9-13 

8-9 

8 

74-90 

Maulichlhys 

10-12 

22-24 

15-17 

9 

8 

64-65 

Lestidium 

9-11 

26-33 

11-13 

9 

8 

75-91 

Lestidiops 

8-13 

25-35 

10-13 

6? 

8 

75-100 

Uncisudis 

10-11 

25-31 

11-13 

9 

8 

75-79 

Lestrolepis 

9-11 

31-44 

10-12 

8 

8 

82-98 

Stemonosudis 

7-12 

29-50 

10-13 

8-9 

8 

84-121 

Macroparalepis 

11-14 

21-32 

10-12 

9 

8 

80-110 

Dolichosudis 

10 

36-37 

11-12 

9 

8 

101 

Sudis 

12-16 

21-24 

13-15 

9 

8 

52-61 

•  For  the  details,  see  relevant  section. 


head  profile;  short  fins;  anus  far  fiDrward  with  wide  preanai 
interspace;  anteriorly  placed  dorsal  and  pelvic  fins.  A  size  series 
of  A.  japonicus  reveals  the  gradual  and  direct  development,  with 
scant  pigmentation  throughout  the  pelagic  stages;  melanophores 
are  restricted  to  the  eyes  and  the  caudal  and  postanal  regions, 
other  than  the  peritoneal  section  which  increases  in  size  in  older 
larvae.  Sequence  of  fin  formation  is  C-D-A-P,-?,.  Full  ray  com- 
plements are  visible  at  about  13.3  mm,  but  vertebral  ossification 
is  delayed  until  about  20  mm,  the  smallest  bottom  specimen 
available  in  my  collection.  Ontogeny  of  the  upper  jaw  bones  is 
remarkable  in  possessing  maxillary  teeth  (1-3)  in  larvae  smaller 
than  1 1  mm.  Two  supramaxillaries,  peculiar  to  this  family,  are 
ossifying  in  metamorphosed  juveniles. 

Myctophidae  (see  Moser,  Ahlstrom.  Paxton,  this  volume). 

Neoscopelidae  (Fig.  709C-D^. — Systematics  of  this  deep-sea  pe- 
lagic and  benthopelagic  family  are  well  understood  (Butler  and 


Ahlstrom,  1976;  Nafpaktitis,  1977),  except  for  5o/;vc)Wfr  which 
is  restricted  to  the  tropical  Western  Pacific.  The  remaining  two 
genera  are  known  from  the  world  oceans.  Developing  eggs  are 
unknown.  Mature  ovarian  eggs  of  Neoscopelus  macrolepidotus 
(0.83-0.98  mm  in  diameter)  contain  a  large  single  oil  globule 
of  0.39-0.61  mm  (Maruyama,  1970).  Advanced  larval  stages 
have  been  described  and  illustrated  for  Neoscopelus  sp.  (Oki- 
yama,  1974b)  and  two  species  of  Scopelengys  (Butler  and  Ahl- 
strom, 1976).  They  are  characterized  by  large  fan-shaped  pec- 
toral fins,  large  head  with  blunt  snout  tip,  small  round  eyes, 
laterally  compressed  deep  body,  and  an  oval  patch  of  melano- 
phores in  the  peritoneum,  distinct  from  the  solid  peritoneal 
pigment  sections  of  most  other  myctophiforms.  All  fins  differ- 
entiate rapidly  with  the  possible  sequence  as  P,-D-A-C-P,,  full 
counts  being  attained  at  a  small  size  (less  than  10  mm).  Pig- 
mentation is  clearly  difl^erent  between  the  two  genera.  Scope- 
lengys lacks  the  pigment  patch  lying  along  the  dorsum  of  the 
rectum  in  Neoscopelus.  Scopelengys  uniquely  develops  a  hori- 


OKIYAMA:  MYCTOPHIFORMES 

Table  58.    Comparison  of  the  Larval  Characters  Among  Four  Genera  of  the  Ipnopidae. 


209 


Characters 

Ipnops 

Bathytyphlops 

Bathymicrops 

Bathypterois 

Head  profile 

slung  down;  flat  top 

slightly  slung  down; 
flat  top 

slung  down;  flat 
top 

slung  down;  flat  top 

Pectoral  fin 

bilobed;  rays  long 

elongated;  fan-shaped 

elongated 

elongated;  fan-shaped 

Gut  size 

short 

short 

long 

long 

Anus  position;  close  to 

pelvic  fin 

pelvic  fin 

pelvic  fin; 
slightly 

anal  fin 

Anus-anal  fin  space 

wide 

wide 

wide 

narrow 

Peritoneal  pigment  section 

absent 

single 

absent 

*numerous  (12-20)  or 
absent 

Body  pigment  (melanophores) 

scant 

scant 

abundant 

scant 

Possible  sequence  of  fin 

P,  C-A-D-Pj 

P,  C-A-D-P, 

P,-C-A-D-P, 

P,-C-A  D-P, 

formation 

Transformation  complete 

ca.  42  mm  SL 

43-93  mm  SL 

70-90  mm  SL 

ca.  42-43  mm  SL 

'  Details  are  mentioned  in  the  text. 


zontal  pigment  bar  across  the  head.  Small  preopercular  spines 
are  known  only  in  Neoscopelus  whereas  a  long  snout  is  peculiar 
to  Scopelengys. 

Chlorophthalmidae  (Fig.  ]09E-F).— Of  Ihe  three  genera  of  this 
benthic  family,  the  cosmopolitan  Chlorophthalmus  is  particu- 
larly diverse  and  abundant.  Extensive  revision  of  this  genus  is 
needed,  since  there  are  many  undescribed  species  from  the  West- 
em  Pacific  and  the  known  species  can  be  divided  into  two  dis- 
tinct groups,  each  warranting  generic  status  (Doi  and  Okamura, 
1983). 

Eggs  are  not  known.  Despite  the  abundance  of  adults,  few 
larvae  have  been  reported.  Complete  developmental  series  are 
available  for  only  C.  agassizi  (Tamng.  1918).  Known  larvae  of 
other  species  such  as  C.  mento,  C.  prondens  and  Chlorophthal- 
mus spp.  (Pertseva-Ostroumova  and  Rass,  1973;  Miller  et  al., 
1979;  Okiyama,  unpubl.)  show  close  resemblance  to  C.  agassizi 
having  the  extremely  short  gut  with  large  preanal  interspace,  a 
similar  pigment  pattern  composed  of  a  single  peritoneal  pigment 
section  lying  at  the  pectoral  fin  base  and  a  melanophore  at  the 
hypural  complex,  short  fins  and  anteriorly  placed  dorsal  and 
pelvic  fins  (as  in  Aulopidae).  There  are  possible  specific  differ- 
ences in  the  size  at  appearance  of  the  peritoneal  pigment  section 
(ca.  7  mm  in  C.  prondens  vs  5-6.6  mm  in  C.  mento)  and  in  the 
arrangement  of  the  few  small  melanophores  on  the  dorsal  and 
ventral  margin  of  the  tail  near  the  notochord  tip  in  early  larvae. 
Meristic  characters  are  useful  in  discriminating  the  particular 
species  or  species  groups,  although  early  developmental  stages 
are  usually  very  difficult  to  identify  to  species. 

Larval  osteology  was  studied  in  detail  for  C.  agassizi  (Rosen, 
1971)  but  the  sequence  of  fin  formation  is  not  clear  except  that 
the  pectoral  fin  develops  early.  Principal  changes  during  the 
gradual  metamorphosis  include  the  rotation  of  the  eyes  dorsally 
which  takes  place  at  sizes  less  than  40  mm  (Ahlstrom,  1972a). 

Unusual  larvae  with  a  pigmentation  pattern  similar  to  the 
above  described  forms  are  found  in  ORl  collections  from  the 
Kuroshio  area  (Fig.  1 09 A).  These  are  distinct  in  that  the  head 
is  markedly  depressed,  bowed  with  duckbilled  appearance,  and 
a  single  peritoneal  pigment  section  is  large  enough  to  cover  the 
dorsal  half  of  the  short  gut.  Their  meristic  characters  (ca.  42 
myomeres  and  ca.  1 7  pectoral  rays)  suggest  a  possible  affinity 
with  Chlorophthalmus  (sensu  lato).  These  two  larval  types  seem 
to  substantiate  the  suggested  dichotomy  of  this  genus.  No  in- 
formation is  available  for  larvae  of  the  other  two  genera  (Par- 
asiidis  and  Bathysauropsis). 


Ipnopidae  (Fig.  1 1 OA-E).  — Four  benthic  genera  compose  this 
family  which  has  been  variously  classified  (e.g..  Nielsen.  1966; 
Sulak.  1977).  Despite  their  deep-sea  mode  of  existence,  larval 
stages  of  all  genera  have  been  mostly  obtained  from  the  surface 
waters.  Developing  eggs  are  not  known.  Mature  ovarian  eggs 
are  known  for  all  genera  with  virtually  identical  features  such 
as  a  spherical  shape,  diameter  of  about  1.0-1.2  mm,  and  the 
presence  of  a  single  large  oil  globule  (Nielsen,  1966;  Sulak,  1977 
and  pers.  comm.;  Merrett,  1980).  Although  intergeneric  differ- 
ences of  the  early  larval  stages  are  remarkable  (Table  58).  they 
share  several  conspicuous  characters  including  the  more  or  less 
hung-down  head  profile  and  the  elongated  precocious  pectoral 
fins.  At  metamorphosis  these  become  less  prominent  in  asso- 
ciation with  the  drastic  change  in  the  mouth  size  from  moderate 
to  huge  and  the  appearance  of  heavy  body  pigmentation. 

Two  larvae  (13.9,  10.6  mm)  are  known  for  Ipnops:  the  larger 
specimen  referred  to  /.  agassizi  was  described  in  considerable 
detail  and  illustrated  (Okiyama,  1981).  The  smaller  one  may  be 
/.  meadi  in  view  of  its  higher  anal  ray  count  (ca.  13).  A  divided 
pectoral  fin  with  elongated  upper  rays  is  peculiar  to  this  genus 
(Table  58).  Principal  changes  at  metamorphosis  include  the  de- 
velopment of  the  unique  eye  plaque,  a  depressed  head  with 
straight  profile,  and  the  disappearance  of  the  peculiar  feature  of 
the  pectoral  fins  along  with  the  loss  of  several  rays.  Metamor- 
phosis may  be  rapid,  but  the  smallest  benthic  juvenile  of  40 
mm  still  bears  the  immature  eye  plaque  (Sulak,  1977). 

Bathytyphlops  includes  only  two  species,  B.  sewelli  and  S. 
marionae  (Merrett.  1980).  A  larva  of  this  genus  was  first  de- 
scribed under  the  name  Macristiella  perlucens  of  uncertain  af- 
finity (Berry  and  Robins,  1967).  The  known  "Macristiella"  (19 
specimens,  7-43  mm)  are  all  referable  to  B.  marionae  except 
for  the  37  mm  larva  from  the  Indian  Ocean  and  the  smallest 
specimen  (Parin  and  Belyanina,  1972).  The  Indian  Ocean  spec- 
imen may  be  identified  as  B.  sewelli  on  the  basis  of  the  higher 
anal  ray  count  (18),  a  unique  character  for  this  species. 

Early  stage  larvae  have  little  melanistic  pigmentation,  but 
some  bluish  or  violet  coloration  is  present  on  the  fins  and  var- 
ious body  parts  in  living  specimens  (Berry  and  Robins,  1967). 
Preserved  individuals  sometimes  retain  this  feature,  usually  on 
the  large  pectoral  or  pelvic  fins.  Reduction  of  the  relative  size 
of  eyes,  and  the  loss  or  replacement  of  the  teeth  as  well  as  gill 
rakers  are  among  the  major  changes  at  metamorphosis,  in  ad- 
dition to  those  common  to  the  family.  Otherwise,  larval  de- 
velopment is  rather  direct  and  the  relative  position  of  the  fins 
and  the  anus  changes  little  throughout  ontogeny.  The  osteology 


210 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


OKIYAMA:  MYCTOPH I  FORMES 


211 


Fig.  109.  (A)  Aulopus  japonicus.  1  1.5  mm  SL,  from  Okiyama  (1974b);  (B)  Aulopus  sp.,  12.3  mm,  from  Okiyama  (1974b);  (C)  Neoscopelus 
sp.,  7.9  mm,  from  southwestern  Japan,  Ocean  Research  Institute  (ORI)  collection;  (D)  Scopelengys  dispar,  6.3  mm,  from  Okiyama  (1974b);  (E) 
Chlorophlhalmus  sp.,  17.1  mm,  from  Indian  Ocean,  ORI  collection;  (F)  Chlorophthalnms  (?)  sp.,  7.5  mm,  from  Kuroshio  waters  off  Japan,  ORI 
collection. 


of  both  larvae  and  adults  is  well  known  (Okiyama.  1972;  Parin 
and  Belyanina,  1972;  Sulak,  1977). 

Bathymicrops  represents  the  deepest  living  myctophiform. 
Two  species.  B.  regis  and  B.  brevianalis,  are  known  from  ex- 
tremely limited  material  from  4225-5900  m  (Nielsen,  1966; 
Merrett  and  Marshall,  1981).  Pelagic  eggs  are  unknown.  A  total 
of  five  larvae  and  juveniles  (13.0-70.0  mm)  are  available;  the 
smallest  two  larvae  (13.0,  14.7  mm)  from  Hawaiian  waters  are 
unidentifiable;  a  20  mm  larva  from  the  North  Atlantic  (=Sto- 
miatella  B  in  Roule  and  Angel,  1930:  PI.  1,  Fig.  7)  is  ascribed 
to  B.  regis;  the  largest  two  juveniles  (62.5,  70.0  mm)  from  the 
tropical  Pacific  are  tentatively  identified  as  B.  brevianalis. 

Despite  conspicuous  variation  among  specimens,  scattered 
melanophore  patches  and  an  extremely  slender  body  are  diag- 
nostic for  this  genus.  The  precocious  pectoral  fins  are  greatly 
elongated  even  in  the  smallest  larva,  but  the  raised  bases  of  the 
dorsal  and  anal  fins  and  the  prominent  finfolds  are  peculiar  to 
the  advanced  stages,  which  also  have  reduced  eye  size  and  a 
slightly  shorter  gut.  Size  at  metamorphosis  is  unusually  large, 
attaining  70-90  mm. 

Bathypterois  is  the  most  speciose  genus  in  this  family.  Three 
subgenera  (Benthosaurus,  Bathypterois  and  Bathycygnus)  and 
18  species  are  currently  included  (Sulak,  1977).  Known  bathy- 
metric  ranges  are  250-5,990  m.  Published  information  of  the 
developmental  stages  is  scant.  Pelagic  eggs  are  not  known.  A 
single  larva  of  14. 1  mm  (Okiyama,  1974b)  was  identified  as  B. 
{Bathycygnus)  longipes  by  Sulak  (1977).  As  stated  before,  the 
known  early  stages  of  "Aulopus  filamentosus"  are  all  referable 
to  those  of  Bathypterois.  probably  B.  (Bathypterois)  mediter- 
raneus  in  view  of  their  localities.  Complete  series  of  early  stages 
are  confined  to  this  species,  but  at  least  three  additional  larval 
forms  are  now  available.  These  known  larvae  share  the  distinct 
forward  shift  of  the  ventral  hypural  elements  in  addition  to  the 
features  given  in  Table  58. 

Known  larvae  are  provisionally  divided  into  two  groups  on 
the  basis  of  the  peritoneal  pigment  sections,  those  with  many 
sections  and  those  which  lack  peritoneal  pigment.  Except  for 
two  larvae,  B.  (B.)  longipes  and  B.  (Benthosaurus)  viridensis 
(33.1  mm)  from  the  Atlantic  (Fahay,  1983),  all  specimens  have 


the  former  character  state.  The  number  of  peritoneal  pigment 
sections  can  be  a  useful  tool  in  discriminating  the  lai^ae,  but 
ranges  of  variation  often  overlap  among  species.  A  western  Pa- 
cific form  with  12-18  pigment  sections  bears  close  resemblance 
to  B.  (B.)  mediterraneus  larvae  whereas  decidedly  lower  myo- 
mere counts  of  the  former  (45-48)  readily  separate  these  two. 
B.  viridensis  larvae  have,  in  addition  to  the  complete  absence 
of  the  peritoneal  pigment  sections,  several  peculiar  features  such 
as  a  slightly  telescopic  eye,  a  protruding  gut,  and  a  long  anal  fin 
and  short  tail.  Comparison  with  the  smallest  demersal  specimen 
(43  mm)  of  the  same  species  (Sulak,  1977)  indicates  that  prin- 
cipal metamorphic  changes  include  the  absorption  of  the  pro- 
duced gut,  lengthening  of  the  posterior  body  and  fin  shrinkage. 
This  may  represent  the  most  pronounced  metamorphosis  in  this 
genus,  since  less  remarkable  transformation  predominated  in 
the  other  species.  Identification  of  the  other  larval  types  remains 
to  be  determined. 


Notosudidae  (Fig.  1 J  lA-B).  —  Bertelsen  et  al.  ( 1 976)  extensively 
revised  this  oceanic  midwater  family,  including  information  on 
early  developmental  stages  of  all  species  (except  Scopelosaurus 
cradockei').  Supplemental  information  on  the  early  stages  is 
available  in  Ozawa  (1978).  Pelagic  eggs  are  unknown.  Maturing 
ovarian  eggs  of  Ahliesaurus  (ca.  0.3  mm  in  diameter)  and  Lu- 
ciosudis  (0.4-0.5  mm)  suggest  that  pelagic  eggs  are  uncommonly 
small  for  this  order. 

General  characteristics  of  these  larvae  are  extremely  similar 
throughout  the  family:  long,  slender  subcylindrical  body,  be- 
coming increasingly  compressed  toward  the  tail;  markedly  de- 
pressed head  with  wedge-like  snout;  posteriorly  protruding  lobes 
in  corpus  cerebelli;  narrow  eye  with  longer  horizontal  axis;  a 
more  or  less  distinct  conical  mass  of  choroid  tissue  on  the  pos- 
terior part  of  slightly  stalked  eye;  anus  at  about  midbody  (except 
.4hliesaurus)  widely  separated  from  anal  fin  origin;  slight  in- 
crease of  gut  length  with  growth  during  the  early  larval  stages; 
absence  of  the  peritoneal  pigment.  Maxillary  teeth  peculiar  to 
larvae  help  diagnose  this  family  but  are  not  unique  (see,  Au- 
lopidae).  Possible  sequence  of  fin  formation  is  CA-D-P.-Pj, 


212 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES -AHLSTROM  SYMPOSIUM 


last  elements  being  rarely  visible  in  larvae  less  than  20  mm. 
Apart  from  the  length  at  metamorphosis  varying  between  25 
and  45  mm  among  species,  pigmentation  pattern  is  usually  the 
only  useful  character  for  specific  identification.  Once  established 
these  pigment  patterns,  mostly  restricted  to  the  tail,  are  retained 
throughout  the  larval  stages,  although  a  few  species  are  known 
to  be  unpigmented  throughout  all  or  part  of  the  larval  period. 

Scopelarchidae  (see  R.  K.  Johnson,  this  volume). 

Bathysauhdae  (Fig.  1 1 IC).— This  deep-water  benthic  family 
consists  of  two  species  of  synchronous  hermaphrodites,  Bathy- 
saurus  mollis  and  B.  ferox  (=B.  agassizi)  (Sulak,  pers.  comm.; 
Wenner,  1978). 

Pelagic  eggs  are  unknown.  Maximum  size  of  mature  ovarian 
eggs  in  B.  ferox  is  1.2  mm  in  diameter  (Wenner,  1978).  So- 
called  "Macristium"  forms  are  now  proved  to  be  larval  Bath- 
ysaunis  (Rosen,  1971;  Johnson,  1974a);  at  least  several  of  the 
five  known  "Macristium"  larvae  (20-83  mm)  are  positively 
identified  with  B.  mollis.  Morphology  and  osteology  of  these 
specimens  have  been  closely  studied,  revealing  many  charac- 
teristic features  such  as  unusually  elongated  fins,  anterior  place- 
ment of  dorsal  and  pelvic  fins,  raised  bases  of  dorsal  and  anal 
fins,  long  gut  (coiled  or  uncoiled)  terminating  just  in  front  of 
anal  fin  origin,  six  peritoneal  saddle-shaped  pigment  sections 
all  evenly  spaced,  and  development  of  a  pattern  of  lateral  bars 
in  some  specimens.  Besides  this  last  feature,  meristic  differences 
serve  to  distinguish  two  species  despite  considerable  variation. 

Metamorphosis  may  take  place  gradually  at  exceptionally  large 
sizes  (more  than  83  mm).  Accompanying  changes  include  short- 
ening of  fins,  expansion  of  the  gape  with  necessary  associated 
changes  in  head  bones  and  associated  anatomy,  backward  shift 
of  the  dorsal  fin  origin,  and  darkening  of  the  body  surface,  oral 
cavity  and  peritoneum. 

Harpadontidae  (Fig.  1 1 ID-E).  — Two  genera  are  recently  in- 
cluded here  (Sulak,  1977;  Johnson,  1982).  Harpadon  comprises 
at  least  four  species  living  in  nearshore  waters,  estuarine  and 
relatively  deep  continental  shelf  waters  of  the  Indo-Pacific.  Crit- 
ical systematic  revision  of  this  genus  is  now  in  progress  (Schmitz, 
pers.  comm.).  A  pelagic  egg  referred  to  H.  nehereus  in  Delsman 
( 1929c)  appears  invalid  (Delsman  and  Hardenberg,  1934).  Early 
developmental  stages  are  poorly  studied;  only  two  specimens 
of//,  nehereus  (25.2,  and  ca.  40  mm)  have  been  illustrated  and/ 
or  briefly  described  (Delsman  and  Hardenberg,  1934;  Okiyama, 
1979b).  A  juvenile  of  55  mm  is  the  smallest  specimen  of  the 
deep  water  congener,  //.  microchir,  available  in  ORI  collections. 
Early  stages  are  readily  discriminated  from  most  other  myc- 


tophiform  larvae  by  the  exceptionally  high  numbers  of  bran- 
chiostegal  rays  (16-27)  and  the  following  characters:  elongate 
compressed  body  with  large  head  and  mouth,  short  snout  (due 
to  the  forward  shift  of  eyes),  scant  pigmentation  except  seven 
pairs  of  peritoneal  pigment  sections,  the  last  two  closer  together 
than  the  others,  and  extension  of  the  lateral  line  scales  onto  the 
caudal  fin.  Of  these  rather  advanced  developmental  features, 
pigmentation  pattern  may  be  common  to  the  earlier  stages. 
Apparently,  long  pectoral  and  pelvic  fins  are  peculiar  to  //. 
nehereus.  Also,  //.  microchir  is  more  lightly  pigmented  than  //. 
nehereus  at  similar  lengths. 

Metamorphosis  seems  gradual.  If  the  occurrence  of  melano- 
phores  over  the  stomach  is  of  significance  in  defining  this  pro- 
cess, transformation  is  completed  by  35  mm  in  //.  nehereus. 

There  are  about  1 5  species  of  Saurida  with  highest  diversity 
in  the  Western  Pacific.  Planktonic  eggs  are  known  for  S.  elon- 
gata,  S.  wanieso.  and  S.  tumbil  besides  several  unidentifiable 
species  (Mito,  1961a;  Zvjagina,  1965a;  Venkataramanujan  and 
Ramanoorthi,  1981).  These  are  spherical,  1.0-1.3  mm  in  di- 
ameter, transparent,  without  oil  globules  and  with  a  narrow  per- 
ivitelline  space.  Hexagonal  sculpturing  on  the  chorion  (0.03- 
0.05  mm  in  mesh  size)  is  either  present  (S.  wanieso  and  S. 
tumbil)  or  absent  (S.  elongata).  Early  developmental  stages  are 
known  for  9  species.  Of  these,  complete  developmental  series 
are  available  for  at  least  4  Pacific  species,  S.  tumbil,  S.  elongata. 
S.  wanieso  and  S.  gracilis  (Dileep,  1977;  Ozawa,  1983)  and  the 
Atlantic  species,  5.  brasiliensis  (K\x(i.omtX]f.ma..  1980).  These  lar- 
vae are  extremely  similar  to  those  of  Harpadon.  except  for  the 
lower  numbers  of  branchiostegals  and  invariably  short  fins. 
Complete  absence  of  the  preanal  finfold  in  the  early  stages  is 
peculiar  to  this  genus  (Ozawa,  1983).  Except  for  S.  brasiliensis. 
however,  these  are  divided  into  two  types  on  the  basis  of  pig- 
mentation pattern.  One  of  these  consisting  of  S.  gracilis  and 
probably  some  Atlantic  congeners  is  characterized  by  evenly 
spaced  peritoneal  pigment  sections  of  similar  size  and  simul- 
taneous differentiation.  In  addition,  prominent  pigment  along 
the  anal  fin  base  and  on  the  caudal  fins  may  be  diagnostic  for 
this  type.  5.  gracilis  larvae  uniquely  develop  a  small  choroid 
mass  on  the  ventral  side  of  narrow  eyes  (Ozawa,  1983)  while 
nothing  is  mentioned  in  this  regard  for  Hawaiian  larvae  (Miller 
et  al.,  1979).  Remaining  larvae  belong  to  the  second  type  in 
which  the  terminal  pigment  section  is  smaller  and  later-ap- 
pearing than  the  anterior  sections.  Other  pigment  is  also  scarse 
or  absent  in  this  latter  type,  where  specific  differences  are  known 
in  the  size  of  pigment  sections  and  vertebral  numbers.  Meta- 
morphosis occurs  fairly  gradually  with  considerable  variation 
in  size  among  species,  but  is  usually  complete  before  40  mm 
(Gibbs,  1959). 


Fig.  110.  (A)  Ipnops  agassizi.  13.9  mm  SL,  from  Okiyama  (1981);  (B)  Balhytyphlops  manonae.  13.1  mm,  from  Okiyama  (1972);  (C) 
Bathymicrops  brevianalis.  70.0  mm,  from  tropical  central  Pacific,  ORI  collection;  (D)  Bathypterois  sp.  (pigmented  type),  from  northeast  of 
Australia,  Southwest  Fisheries  Center  (SWFC)  collection;  (E):  Bathypterois  viridensts  (unpigmented  type),  from  Fahay  (1983). 


Fig.  111.  (A)  Scopelosaurus  smilhii.  1 3.4  mm  SL,  from  southwestern  Pacific,  ORI  collection;  (B)  the  same,  dorsal  view  of  head;  (C)  Bathysaurus 
ferox.  33.0  mm,  from  Marshall  (1961);  (D)  Harpadon  nehereus.  25.2  mm,  from  East  China  Sea,  ORI  collection;  (E)  Saunda  undosquamis.  15.6 
mm,  from  Okiyama  (1974b);  (F)  Synodus  lucioceps.  10.5  mm,  from  California  current  region.  SWFC  collection;  (G)  Trachinocephatus  myops. 
21.3  mm,  from  Zvjagina  (1965a). 

Fig.  112.  (A)  Atepisaurus  brevirostris.  12.1  mm,  from  Rofen(1966b);(B)/l./era>:.  10.0  mm,  from  central  Pacific  near  Hawaii,  SWFC  collection; 
(C)  Anotopterus  pharao,  14.2  mm,  from  California  current  region,  SWFC  collection;  (D)  Omosudis  lowei  (central  western  Atlantic  specimen), 
11.8  mm,  from  Rofen  (1966b);  (E-F)  O.  lowei.  22.5  mm,  from  tropical  western  Pacific,  ORI  collection,  showing  dorsal  view  of  head. 


OKIYAMA:  MYCTOPHIFORMES 


213 


214 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


OKIYAMA:  MYCTOPHIFORMES 


215 


216 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Synodontidae  (Fig.  1 1 IF-G). — Synodus  includes  about  30  species 
and  has  a  circumglobal  distribution  with  distinctly  high  diversity 
in  the  Indo-Pacific.  Another  monotypic  genus  of  this  family 
(Trachinocephalus)  shows  world-wide  distribution.  A  recent  re- 
vision of  the  Indo-Pacific  Synodus  (Cressey,  1981),  including 
many  new  species,  critically  changed  its  systematic  status.  Thus, 
most  of  the  known  eggs  and  larvae  are  subject  to  nomenclatural 
revision.  Early  stages  of  this  family  can  be  separated  from  those 
of  the  previous  family  by  the  presence  of  the  preanal  finfold 
(Ozawa,  1983). 

Trachinocephalus  myops  larvae  are  distinct  in  possessing  six 
pairs  of  large  peritoneal  pigment  sections  of  uniform  size,  a 
rounded  head  with  short  snout,  and  additional  unique  pigmen- 
tation (Rudometkina,  1980;  Ozawa,  1983).  This  species  and 
most  species  of  Synodus  have  an  extremely  elongated  body.  An 
exception  is  the  eastern  Pacific  species,  S.  lucioceps.  which  has 
a  slightly  deeper  body.  A  complete  developmental  series  is  known 
only  for  this  species  in  Synodus;  eggs  are  spherical,  1.33-1.44 
mm  in  diameter,  without  an  oil  globule,  with  moderately  broad 
perivitelline  space  and  hexagonally  sculptured  chorion:  larvae 
are  characterized  by  7  evenly  spaced  pairs  of  pigment  sections 
formed  gradually,  a  ventral  melanophore  lying  at  the  midpoint 
of  tail,  and  one  near  the  notochord  tip. 

As  in  the  Harpadontidae,  meristic  characters  and  pigmenta- 
tion patterns  are  of  particular  aid  in  identifying  the  early  stages 
of  this  family.  If  established  pigmentation  patterns  are  retained 
in  the  metamorphosed  juveniles  or  adults,  numbers  of  the  per- 
itoneal pigment  sections  of  all  Indo-Pacific  species  of  Synodus 
(Cressey,  1981)  vary  between  0  and  17  with  a  maximum  range 
of  infraspecific  variation  of  0-3  in  5.  binotalus  and  14-17  in  S. 
usitatus;  some  species  appear  to  lack  this  pigment  (i.e.,  S.  kaian- 
us  and  S.  binotalus),  however  this  needs  to  be  documented  by 
complete  developmental  series.  Another  point  of  interest  is  the 
asymmetry  and  size  disparity  of  the  pigment  pairs  known  in 
"S.  variegatus"  of  Okiyama  (1974b). 

Size  at  metamorphosis  and  sequence  of  fin  formation  of  this 
family  appear  to  be  identical  to  those  in  Harpadontidae.  Ozawa 
(1983)  revealed  the  following  pattern  of  fin  formation:  C-A-D- 

P,P2. 

Alepisauridae  (Fig.  //2,4-BA— This  widely  distributed  bathy- 
pelagic  family  includes  only  two  species,  .Hepisaurus  fero.x  and 
A.  brevirostris.  with  slightly  different  ranges:  the  latter  is  appar- 
ently absent  from  the  North  Pacific  (Francis,  1981).  Eggs  are 
unknown.  A  series  of  early  developmental  stages  of  Alepisaurus 
sp.  (6.9-17.2  mm)  has  been  described  and  illustrated  (Rofen, 
1966b).  In  addition,  three  larvae  (9.6-16.5  mm)  from  the  col- 
lection of  the  Southwest  Fisheries  Center,  La  Jolla  have  different 
features.  They  share  with  previous  specimens  a  large  head  and 
mouth,  prominent  canine  teeth  on  the  dentary,  small  fins  in- 
cluding pigmented  pectorals  of  moderate  size,  gently  curved 
head  profile  and  short  gut  with  heavy  pigmentation.  The  peri- 
toneal pigment  section  is  indistinct.  This  new  material  is  unique 
in  having  4  small  preopercular  spines,  pigment  patches  at  the 
anal  fin  origin,  and  distinct  bony  ridges  dorsally  on  the  head. 
Judging  from  the  locality  of  these  specimens,  near  Hawaii  in 
the  North  Pacific,  Alepisaurus  sp.  larvae  of  Rofen  (1966b)  can 
be  identified  with  A.  brevirostris.  and  these  with  A.ferox. 

Metamorphosis  may  be  gradual  with  possible  sequence  of  fin 
formation  P,  C-D-A-Pj. 

Anotopteridae  (Fig.  112C).— One  world-wide  species,  Anotop- 
terus  pharao,  constitutes  this  open  ocean  family,  uniquely  lack- 


ing the  dorsal  fin.  Eggs  are  not  known.  A  larva  (ca.  1 5  mm)  has 
been  briefly  described  without  illustration  (Nybelin,  1948):  this 
specimen  is  unavailable  now  (Thulin,  pers.  comm.).  Another 
larva  of  similar  size  (14.2  mm)  is  available  from  the  collection 
of  the  Southwest  Fisheries  Center,  La  Jolla.  It  is  characterized 
by  a  slender  thin  body,  absence  of  peritoneal  pigment  sections, 
large  head  with  pointed  snout,  a  fleshy  prolongation  at  the  tips 
of  both  jaws,  two  large  canine  teeth  on  each  palatine,  and  a 
fairly  long  gut  extending  beyond  midbody.  Pigmentation  is  scat- 
tered on  various  parts  of  body  including  the  snout,  jaw  tips, 
dorsal  midline  of  body,  near  the  tail  tip,  and  peritoneum  (par- 
ticularly along  the  dorsum  of  gut).  Except  for  the  pectoral  fin, 
fin  aniages  are  lacking.  A  juvenile  of  about  50  mm  illustrated 
in  Rofen  ( 1 966c)  is  similar  to  the  described  larva,  except  all  fins 
are  differentiated  including  the  adipose  fin:  body  pigmentation 
is  remarkable  in  this  juvenile.  Perhaps,  this  species  has  the  most 
direct  pattern  of  early  development  in  this  order. 

Evermannellidae  (see  R.  K.  Johnson,  this  volume). 

Omosudidae  (Fig.  ]  12D-F).—A  single  mesopelagic  species, 
Omosudis  lowei.  constitutes  this  cosmopolitan  family.  Pelagic 
eggs  are  not  known.  Excellent  developmental  series  have  been 
described  and  illustrated,  chiefly  based  on  Atlantic  material 
ranging  from  5.7  to  75.2  mm  (Ege,  1958:  Rofen,  1966b).  Re- 
cently, a  larva  (11.5  mm)  with  different  features  was  briefly 
described  and  illustrated  (Belyanina,  1982b).  Its  locality  in  the 
tropical  western  Pacific  is  peculiar  and  additional  specimens  are 
available  in  ORI  collections  (pers.  obs.). 

These  have  in  common  a  very  large  head  and  mouth,  stubby 
body,  long  pointed  snout,  straight  head  profile,  small  fins,  par- 
ticularly the  pectoral,  large  canine  teeth  on  denlary  and  palatine, 
and  several  closely  spaced  peritoneal  pigment  sections.  How- 
ever, trenchant  morphological  differences  between  the  Atlantic 
and  Pacific  specimens  are  known:  head  smooth  vs  armed  (along 
edge  of  preopercle  and  dorsum  of  head):  pigmentation  light  vs 
dense  at  a  similar  size:  pigmented  band  above  posterior  part  of 
anal  fin  absent  vs  present.  For  this  first  character,  there  is  a 
possibility  that  the  minute  preopercular  spines  have  been  over- 
looked in  the  Atlantic  larvae. 

Sequence  of  fin  formation  known  in  the  Atlantic  specimens 
is  C-DA-Pj-P,.  Metamorphosis  is  gradual  with  possible  dif- 
ferences in  the  size  of  completion  between  the  two  types  as 
suggested  above.  The  presence  of  two  larval  types  is  in  sharp 
contrast  with  the  current  concept  of  a  monotypic  family.  In  this 
connection,  Ege's  comments  ( 1 958)  on  the  significant  differences 
in  dorsal  ray  numbers  between  the  populations  from  the  South 
China  Sea  and  north  Atlantic  are  of  particular  interest. 

Paralepididae  (Fig.  1 13. 4-G}.— This  oceanic  pelagic  family  in- 
cludes about  1 1  genera  and  50  species  and  constitutes  the  second 
largest  group  in  the  order  after  Myctophidae.  Some  genera  are 
still  in  need  of  critical  revision,  while  the  two  established 
subfamilies  seem  valid.  Paralepidiinae  includes  two  tribes,  the 
Paralepidiini  (3  genera)  and  Lestidiini  (7  genera),  and  Sudinae 
has  I  genus  (Sudis).  Ege  (1930)  and  Rofen  (1966a)  mcluded 
early  larval  stages  in  their  extensive  studies  of  this  family.  Eggs 
are  not  known  but  developmental  stages  are  known  for  9  out 
of  1 1  genera.  Larval  development  of  Sudis  has  been  closely 
studied  for  5.  hyalina  and  S.  a/ro.v  (Sanzo,  1917:  Shores,  1969: 
Belyanina,  1981).  These  unusual  larvae  are  readily  discrimi- 
nated from  those  of  the  other  subfamily  by  the  relatively  short 
body  with  large  head,  long  pectoral  fins,  long  gut  and  early 


OKIYAMA:  MYCTOPHIFORMES 


217 


Fig.  113.  (A)  Paralepis  elongata.  16.7  mm  SL,  from  Rofen  (1966a);  (B)  Notolepis  coatsi.  60.5  mm,  from  Efremenko  (1983);  (C)  Leslidiops 
ringens.  9.4  mm,  from  California  current  region.  SWFC  collection;  (D)  the  same,  28.5  mm;  (E)  Stemonosudis  macrura.  1 1.2  mm,  from  Ege(1957); 
(F)  Sudis  hyalina.  16.1  mm,  from  Shores  (1969);  (G)  5.  alrox.  21.5  mm,  from  Berry  and  Perkins  (1966). 


established  complement  of  peritoneal  pigment  sections,  spine- 
tipped  flanges  on  ventral  region  of  preoperculum.  over  eye,  and 
snout.  5".  alrox  has  a  spine-tipped  flange  along  lower  jaw.  and 
the  large  spine  at  the  preopercular  angle  is  serrated  only  in  5. 
airo.x.  The  precocious  pectoral  fin  is  relatively  short  until  about 
1 5  mm  in  S.  atrox  whereas  it  is  very  long  even  in  8.0  mm  larvae 
of  5.  hyalina.  The  number  of  peritoneal  pigment  sections  is  6 


(5  in  early  larvae)  in  S.  atrox  vs  7-8  in  S.  hyalina.  Trunk 
pigment  is  evenly  distributed  in  S.  atrox  \s  patchy  in  S.  hyalina. 
Except  for  this  genus,  the  developmental  features  of  this  fam- 
ily are  remarkably  cohesive.  Known  lar\ae  have  a  very  long 
compressed  body,  a  short  trunk  in  early  larvae,  large  head  in 
advanced  larvae,  elongated  pointed  snout  with  straight  head 
profile,  various  numbers  of  peritoneal  pigment  sections  sequen- 


218 


ONTOGENY  AND  SYSTEMATICS  OF  HSHES-AHLSTROM  SYMPOSIUM 


tially  formed  with  gradual  lengthening  of  gut,  well  developed 
preanal  finfolds  and  apparently  precocious  anal  fin  rays.  Ad- 
ditionally, during  ontogeny  eye  shape  changes  from  ovoid  to 
round,  and  body  pigmentation  changes  from  light  to  dense. 
These  larvae  are  too  similar  in  general  appearances  to  determine 
trenchant  characters  that  define  genera  or  tribes.  Peritoneal  pig- 
ment sections,  are  of  prime  importance  in  identifying  early  stages, 
but  show  extreme  variability  with  respect  to  their  number  and 
sequential  development.  Of  particular  interest  in  this  connection 
is  Notolepis.  N.  rtssoi  develops  1 2  pigment  sections,  the  largest 
number  in  the  family  except  Stemonosudis  (3 1 ),  whereas  the 
Antarctic  congener,  TV.  coalsi,  has  only  a  single  section  which 
increases  in  size  with  growth  (Efremenko,  1978,  1983a).  Among 
the  various  genera  the  primary  section  develops  at  5-10  mm 
and  full  complements  are  formed  variously  by  the  species  be- 
tween 15-45  mm.  Usually,  metamorphosis  takes  place  around 
this  size  accompanied  by  the  development  of  a  black  perito- 
neum. 

In  addition  to  the  exceptionally  higher  number  of  pigment 
sections,  Stemonosudis  is  peculiar  in  having  a  filamentous  pro- 


jection on  the  lower  jaw  tip  (in  larvae  of  5.  macrura  and  in 
juveniles  and  adults  of  S.  intermedia  and  5.  elongatd).  Likewise, 
Uncisudis  (=Pontosudis)  uniquely  develops  an  elongated  pelvic 
fin. 

Patterns  of  melanophores  are  extremely  diverse  but  of  use  in 
identifying  species  or  species  groups;  pigment  patches  on  the 
caudal  peduncle,  dorsum  of  body,  and  caudal  and  pectoral  fins 
are  particularly  important.  Rofen  (1966a)  suggested  that  the 
single  larval  character  discriminating  the  two  tribes  in  Parale- 
pidiinae,  i.e.,  Paralepidiini  and  Lestidiini,  is  whether  the  rear- 
ward shift  of  the  anus  occurs  early  or  late  in  ontogeny. 

Incertae  ce^w.  —  Peculiar  eggs  described  by  Delsman  (1938)  and 
Mito  (1961a)  are  currently  considered  to  be  those  of  mycto- 
phiform  fishes  other  than  Myctophidae  (Moser  and  Ahlstrom, 
1970).  These  eggs  are  spherical,  1.12-1.37  mm  in  diameter,  with 
a  single  oil  globule  and  bear  numerous  short  appendages  on  the 
chorion.  Two  types  are  known  only  from  Asian  waters. 


Ocean  Research  Institute,  University  of  Tokyo, 
MiNAMiDAi,  Nakano-ku,  Tokyo  164,  Japan. 


1-15-1, 


Myctophidae:  Development 
H.  G.  Moser,  E.  H.  Ahlstrom  and  J.  R.  Paxton 


LANTERNFISHES  of  the  family  Myctophidae  are  found  in 
all  oceans  of  the  world.  Some  230-250  species  are  arranged 
in  36  generic/subgeneric  taxa  (Table  59).  All  nominal  species 
are  listed  in  Paxton  (1979).  Characteristic  of  the  family  is  the 
presence  of  light  organs  or  photophores  on  the  head  and  body 
(Fig.  1 1 4).  The  different  patterns  of  photophores  have  been  used, 
along  with  meristics  (Table  60),  in  species  diagnoses  and  as  a 
basis  for  classification  within  the  family  since  the  late  1800's. 
Most  authors  have  placed  the  Myctophidae  and  closely  related 
Neoscopelidae  with  the  families  Aulopidae,  Chlorophthalmidae 
and  related  families  in  an  order  or  suborder  variously  named 
the  Iniomi,  Myctophoidea  or  Myctophiformes  (Gosline  et  al., 
1966;  Greenwood  et  al.,  1966;  Nelson,  1976;  Johnson,  1982), 
although  Rosen  ( 1973)  separated  the  Myctophidae  and  Neosco- 
pelidae as  a  restricted  order  Myctophiformes.  Moser  and  Ahl- 
strom (1970,  1972,  1974),  Ahlstrom  et  al.  (1976)  and  Paxton 
(1972)  are  the  most  recent  papers  considering  relationships  with- 
in the  family;  characteristics  of  larvae  and  bones  and  photo- 
phores of  adults  were  primarily  utilized  in  the  respective  studies. 
Paxton's  (1972)  classification,  including  genera  recognized  sub- 
sequently, is  as  follows: 


Subfamily  Myctophinae 
Tribe  Electronini 

Genera:   Protomyctophum. 
Metelectrona- 


Krefftichlhys',  Elect rona. 


Tribe  Myctophini 

Genera:  Benthosema,  Diogenichlhys,  Hygophum,  Myc- 
tophum.  Symbolophorus 
Tribe  Gonichthyini 
Genera:  Loweina,  Tarletonbeania,  Gonichthys,  Centra- 
branch  us 
Subfamily  Lampanyctinae 
Tribe  Notolychnini 

Genus  Notolychnus 
Tribe  Lampanyctini 

Genera:  Taaningichthys,  Lampadena,  Bolinichthys.  Lep- 
idophanes,  Ceratoscopelus.  Stenobrachius,  Lampan- 
yctus,  Triphoturus,  Parvilux^ 
Tribe  Diaphini 

Genera:  Lobianchia,  Diaphus,  Idiolychnus* 
Tribe  Gymnoscopelini 

Genera:  Lampanyctodes,  Gymnoscopelus,  Notoscopelus, 
Lampichthys,  Scopelopsis,  Hintonia 

There  has  not  been  a  family  revision  at  the  species  level  since 
Fraser-Brunner's  (1949)  study.  A  large  number  of  more  recent 
generic  revisions  and  regional  studies  are  currently  the  primary 
sources  for  species  identifications;  most  of  these  have  been  uti- 
lized in  compiling  the  generic  distribution  limits  (Table  59).  The 
most  recent  zoogeographic  studies  are  those  of  Backus  et  al. 


Hulley  (1981). 
■  Wisner(1963). 


'  Hubbs  and  Wisner  (1964). 

"  Nafpaktitus  and  Paxton  (1978). 


MOSER  ET  AL.:  MYCTOPHIDAE 


219 


Table  59.    Geographic  Distribution  of  the  Genera  and  Subgenera  of  Myctophidae.  References  marked  *  are  useful  for  the  identification 
of  species.  The  division  of  the  Atlantic  and  Indian  Oceans  is  arbitrarily  taken  at  20°E,  the  Indian-Pacific  Ocean  boundary  at  130°E. 


No,  of 
species 


Lai.  extremes 


Krefftichthys 

I           Atlantic 
Indian 
Pacific 

34°S-60°S 
43°S-66°S 
34°S-72°S 

Protomyclophum 
(Protomyctophum) 

7           Atlantic 
Indian 
Pacific 

34°S-60°S 
44°S-65°S 
40'^70°S 

Protomyctophum 
(Hierops) 

7           Atlantic 
Indian 
Pacific 

70°N-56°S 
35°S-52°S 
57°N-67°S 

Electrona 

5           Atlantic 
Indian 
Pacific 

55°N-70°S 

2°N-68°S 

42°N-70°S 

Metelectrona 

2           Atlantic 
Indian 
Pacific 

35''S-5I°S 
35°S-47°S 
33°S-55°S 

Benthosema 

5           Atlantic 

80°N-38°S 

Indian 
Pacific 

2rN-35°S 
7I''N-42°S 

Diogenichthys 

3           Atlantic 
Indian 
Pacific 

50°N-48°S 
18°N-45°S 
37°N-41°S 

Hygophum 

9-1 1       Atlantic 
Indian 
Pacific 

49°N-48°S 
20''N-42°S 
39"'N-46°S 

Symbolophorus 

7-9         Atlantic 
Indian 
Pacific 

59''N-51°S 
2I°N-41°S 
50°N-59°S 

Myctophum 

13-14       Atlantic 
Indian 

65''N-40°S 
20°N-34°S 

Pacific 

42''N-42°S 

Loweina 

3-4         Atlantic 
Indian 
Pacific 

44°N-38°S 

IO°S-40'^ 

32''N-40°S 

Tarletonheania 

1-2         Atlantic 
Indian 
Pacific 

50°N-30°N 

Gonichthys 

3-4         Atlantic 
Indian 
Pacific 

47°N-40°S 
25°S-39'>S 
3I''N-42°S 

Cenlrobranchus 

3-4         Atlantic 
Indian 
Pacific 

46°N-35°S 
15°N-33°S 
37°N-37°S 

Nololychnus 

I           Atlantic 
Indian 
Pacific 

56''N-38°S 
1  rN-40°S 

34'>N-44°S 

Lobianchia 

2          Atlantic 
Indian 
Pacific 

6rN-5l°S 

2'>N-40°S 

32°N-47°S 

Diaphus 

65-75       Atlantic 
Indian 
Pacific 

62''N-52'^ 
23''N-48°S 
55°N-58°S 

Idiolychnus 

1           Atlantic 
Indian 
Pacific 

13°S-24°S 
2rN 

*Hulley  (1981:12) 

*Hulley  (1972:217);  Andriashev  (1962:224) 
Andnashev  (1962:225):  McGinnis  (1982:1 1) 

*Hulley  (1981:29,  19) 
Hulley  (1972:218);  *McGinnis  (1982:17) 
♦Andriashev  (1962);  *McGinnis  (1982:16,  17) 

Nafpaktitis  et  al.  (1977:31);  *Hulley  (1981:36) 
•Nafpaklitis  and  Nafpaktitis  ( 1 969:7);  'McGinnis  (1982:18) 
♦Wisner  ( 1 976:20);  'McGinnis  ( 1 982: 1 8) 

*Hulley  (1981:40,  46);  *McGinnis  (1982:21) 
Nafpaktitis  and  Nafpaktitis  (1969:10);  *McGinnis  (1982:21) 
•Andriashev  ( 1 962);  Ebeling  ( 1 962: 1 40);  *McGinnis  (1982:21) 

•Hulley  (1981:53) 

•McGinnis  (1982:25) 

•Bussing  (1965:200);  •McGinnis  (1982:25) 

•Nafpaktitis  et  al.  (1977:52);  Hulley  (1972:220);  (the  specimen  from 
55°S  is  possibly  mislabeled,  McGinnis,  (1982:26,  29)) 
Kotthaus  (1972:18);  *Nafpaktitis  and  Nafpaktitis  ( 1969:1 1) 
•Wisner  (1976);  Nafpaktitis  et  al.  (1977:52);  Robertson  et  al.  (1978:302) 

Nafpaktitis  et  al.  (1977:58);  Hulley  (1981:58) 
•Nafpaktitis  and  Nafpaktitis  (1969:15) 
•Wisner  (1976:49);  Rass  (1960:149) 

•Bekker(1965);  •Nafpaktitis  et  al.  (1977:38);  •Hulley  (1981:61) 

•Bekker  (1965:80);  Hulley  (1972:222) 

•Wisner  (1976);  •Bekker  (1965:94);  McGinnis  (1982:30) 

•Hulley  (1981:101) 
Kotthaus  (1972:27);  *Nafpaktitis  and  Nafpaktitis  (1969:29) 
•Wisner  (1976);  Frost  and  McCrone  (1979:755);  •McGinnis  (1982:33) 

•Nafpaktitis  et  al.  (1977:62);  'Hulley  (1981:87) 
Nafpaktitis  and  Nafpaktitis  (1969);  •Bekker  and  Borodulina 

(1978:1 20);  McGinnis  ( 1 982:34) 
•Kawaguchi  and  Aioi  (1972);  •Wisner  (1976);  Kawaguchi  et  al. 

(1972:27);  Paxton  and  Nafpaktitis  (ms) 

•Nafpaktitis  et  al.  (1977:85) 

•Bekker  (1964:23);  •Nafpaktitis  and  Nafpaktitis  (1969:31) 

•Wisner  (1976);  •Bekker  (1964:23);  McGinnis  (1982:37) 


•Bekker  (1963:160);  'Wisner  (1976:82) 

Nafpaktitis  et  al.  (1977:88);  Hulley  (1981:107) 
•Bekker  (1964:38) 
•Bekker  (1964);  •Wisner  (1976:86);  McGinnis  (1982:36) 

•Nafpaktitis  et  al.  (1977:91) 
•Bekker  (1 964:5 1,  58) 
•Bekker  (1964:58) 

•Nafpaktitis  et  al.  (1977:94);  'Hulley  (1972:222) 
Kotthaus  (1972:30);  McGinnis  (1982:37) 
Ebeling  (1962:141);  McGinnis  (1982:37) 

•Nafpaktitis  el  al.  (1977);  Bekker  (1967:98);  McGinnis  (1982:51) 
•Nafpaktitis  (1978:7);  McGinnis  (1982:51) 
'Wisner  (1976:96);  McGinnis  (1982:51) 

'Nafpaktitis  et  al.  (1977:158);  McGinnis  (1982:52) 

'Nafpaktitis  (1978:62,  78) 

'Nafpaktitis  (1978:62);  McGinnis  (1982:52) 

'Nafpaktitis  and  Paxton  (1978:495) 
'Nafpaktitis  and  Paxton  (1978:495-496) 


220 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES -AHLSTROM  SYMPOSIUM 


Table  59.    CoNTrNUED. 


No  of 

species  Ocean 


Lai.  exlremes 


References 


Lampanyctodes 

1           Atlantic 
Indian 
Pacific 

I9°S-34°S 

35''S 
34°S-5I°S 

Gymnoscopelus 
(Gymnoscopelus) 

4           Atlantic 
Indian 
Pacific 

34°S-66°S 
60°S-65°S 
40°S-72°S 

Gymnoscopelus 
(Nasolychnus) 

4-5         Atlantic 

Indian 
Pacific 

34°S-57°S 

24°S-65°S 
40°S-70°S 

Scopelopsis 

1           Atlantic 
Indian 
Pacific 

11°S-48°S 

9»S-40°S 

15°S-35°S 

Lampichthys 

1           Atlantic 
Indian 
Pacific 

30°S-48°S 

35'^-40°S 

7°S-49°S 

Notoscopelus 

(Noloscopelus) 

5          Atlantic 
Indian 

65°N-60°S 
8''S-36'^ 

Pacific 

50°N-37°S 

Notoscopelus 
(Parieophus) 

1           Atlantic 
Indian 
Pacific 

50°N-2rN 

Hintonia 

1           Atlantic 
Indian 
Pacific 

39°S-48''S 
47<«-51°S 
40°S-50°S 

Lampadena 
(Lampadena) 

8-9         Atlantic 
Indian 
Pacific 

65°N-48°S 

6''N-49°S 

4rN-49°S 

Lampadena 
(Dorsadena) 

I           Atlantic 
Indian 
Pacific 

45°N 

Taaningichthys 

3           Atlantic 
Indian 
Pacific 

43°N-44°S 

8°N-30°S 

4rN-68°S 

Ceratoscopelus 

3           Atlantic 
Indian 

52°N-45°S 
20''N-43''S 

Pacific 

43°N-42°S 

Lepidophanes 

2           Atlantic 
Indian 
Pacific 

43''N-48°S 

Bolinichthys 

7           Atlantic 
Indian 
Pacific 

53°N-41°S 
21°N-44°S 
31°N-43°S 

Triphoturus 

3-4         Atlantic 
Indian 
Pacific 

8°N-I4°S 
38°N-35°S 

Stenohrachius 

2           Atlantic 
Indian 
Pacific 

57<'N-30°N 

Parvilux 

2           Atlantic 
Indian 
Pacific 

40''N-14°S 

Lampanyctus 

40           Atlantic 
Indian 

65°N-60°S 
16°N-60°S 

Pacific 

59''N-72°S 

♦Ahlstrom  et  al.  (1976:146);  Grindley  and  Pennth  (1965:283) 

Paxton  and  Nafpaktitis  (in  prep.) 
*Wisner  (1976: 1 58-1 59);  McGinnis  (1982:55) 

*Hulley  (1981:254);  •McGinnis  (1982:59) 
*Andriashev  (1962:267);  •McGinnis  (1982:59) 
♦McGinnis  (1982:61,  58) 

♦Hulley  (1981:261);  (03°S,  Fraser-Brunner  (1931:224)  presumably  a 

waiO 
Smith  (1 933a: 1 26);  *McGinnis  (1982:64) 
♦Andriashev  (1962);  McGinnis  (1982:64) 

*Hulley  (1981:241) 
Legand  (1967:49);  McGinnis  (1982:57) 
*Wisner  (1976:222);  Paxton  and  Nafpaktitis  (in  prep.) 

Hulley  (1981:242) 
McGinnis  (1982:57) 
*Wisner  (1976:215);  McGinnis  (1982:57) 

♦Nafpaktitis  et  al.  (1977:254)  Andriashev  (1962:278) 
Nafpaktitis  and  Nafpaktitis  (1969:35);  Grindley  and  Penrith 

(1965:283) 
*Fujkii  and  Uyeno  (1976);  Frost  and  McCrone  (1979:755);  Collins 

and  Baron  (1981:11) 

•Nafpaktitis  et  al.  (1977:257) 


•Hulley  (1981:239) 

McGinnis  (1982:55) 
•Wisner  (1976:220);  McGinnis  (1982:55) 


•Kreflt  (1970:285);  Hulley  (1981:180) 
•Nafpaktitis  and  Paxton  (1968:20,  21) 
•Nafpaktitis  and  Paxton  (1968:20,  21) 


•Coleman  and  Nafpaktitis  (1972:2) 

•Hulley  (1981:167);  'Davy  (1972) 

•Nafpaktitis  and  Nafpaktitis  (1969:40) 

•Davy  (1972:72);  •Nafpaktitis  et  al.  ( 1 977: 1 9 1 ) 

•Nafpaktitis  et  al.  (1977:243);  Hulley  (1981:237) 

•Bekker  and  Borodulina  (1968:792);  •Nafpaktitis  and  Nafpaktitis 

(1969:65) 
•Wisner  (1976:207);  Robertson  et  al.  (1978:302) 

•Nafpaktitis  et  al.  (1977:225);  •Hulley  (1981:223) 


•Nafpaktitis  et  al.  (1977:240);  •Hulley  (1981:229) 

Kotthaus  (1972:18):  'Nafpaktitis  and  Nafpaktitis  (1969:60) 
•Johnson  (1975:58);  Nafpaktitis  et  al.  (1977:234) 

Hulley  (1981:205) 
•Nafpaktitis  and  Nafpaktitis  (1969:51) 
•Wisner  (1976:165) 


•Wisner  (1976:160) 


•Wisner  (1976:163,  164) 

•Nafpaktitis  et  al.  (1977:196);  •Hulley  (1981:183);  Zahuranec  (1980) 
•Nafpaktitis  and  Nafpaktitis  (1969);  Kotthaus  (1972:35);  •McGinnis  (1982:42); 

Zahuranec  (1980) 
•Wisner  (1976:191);  McGinnis  (1982:42);  Zahuranec  (1980) 


MOSER  ET  AL.:  MYCTOPHIDAE 


221 


Table  60. 

Meristics  of  the  Genera 

AND  Subgenera  of 

Myctophidae. 

Fin  rays 

Branchio- 

Dorsal 

Anal 

Pectoral 

Pelvic 

Procurrenl  caudal 

Vertebrae 

stegals 

Gill  rakers 

Krefflichthys 

11-14 

17-19 

14-16 

8-9 

8-9  +  7-9 

36-39 

6-8  +    19-23 

Protomyctomphum 

10-14 

21*-27 

14-17 

8-9 

7-9  +  6-9 

35-41 

8-10 

4-7  +  14-21 

P.  Hierops 

11-13 

20-27 

15-18 

8 

7-1 1  +  6-9 

36-42 

9-10 

3-5  +  13-18 

Electrona 

12-16 

18-22 

11-17 

8 

6-10  +  6-9 

33-41 

7-8 

3-10  +  12-25 

Metelectrona 

13-15 

19-22 

14-16 

8 

10  +  9 

35-38 

8 

4-7  +  16-20 

Benthoscma 

11-15 

16-22 

10-17 

8-9 

7-9  +  7-9 

31-37 

9 

3-10  +  10-21 

Dwgenichlhys 

10-13 

14-18 

10-14 

7-8 

7-9  +  7-9 

29-34 

7 

2-4  +  10-12 

Hygophum 

10-15 

18-25 

12-17 

8-9 

6-9  +  6-9 

34-40 

9 

3-6  +  12-16 

Myclophum 

11-15 

16-27 

12-22 

7-8 

7-9  +  7-9 

35-46 

8-9 

4-8  +  10-21 

Symhotophorus 

12-16 

18-24 

12-20 

8 

8-10  +  7-9 

36-42 

9 

4-7  +  12-19 

Loweina 

10-13 

13-17 

9-12 

7-9 

6-7  +  6-7 

37-39 

9 

2-3  +  5-10 

Tarlelonheania 

11-15 

16-20 

11-16 

8 

5-8  +  5-8 

40-42 

8 

4-6  +  10-12 

Gonichlhys 

10-13 

17-24 

11-18 

6-8 

5-6  +  5-6 

38-41 

9 

3-6  +  7-12 

Cenlrobranchus 

9-12 

16-20 

11-17 

8 

5-7  +  5-7 

35-40 

7-8 

0 

Nololychnus 

10-12 

12-15 

11-15 

6-7 

7-9  +  7-9 

27-31 

9-10 

2  +  8-9 

Lnlnanchta 

15-18 

13-15 

11-15 

8 

5-7  +  5-6 

33-35 

9 

4-6  +  11-16 

Diaphus 

10-19 

11-19 

9-14 

8 

5-8  +  5-8 

31-37 

8-9 

4-11  +  9-21 

Idwlychnus 

14-15 

14-16 

13-15 

8 

34 

6-7  +  14-15 

Lampanyctodes 

13-14 

14-17 

12-14 

8 

8-10  +  9-10 

36-39 

9-11 

10-11  +  20-23 

Gymnoscopelus 

14-21 

16-22 

12-16 

8-9 

10-12  +  11-15 

41-45 

10 

6-12  +  14-26 

G.  Nasolychnus 

16-20 

16-20 

12-15 

8 

8-13  +  10-15 

41-45 

10-11 

7-12  +  17-25 

Scopelopsis 

20-24 

23-27 

10-12 

7-8 

9-11  +  11-12 

38-39 

9-10 

7-9  +  16-18 

Lainpkhlhys 

16-18 

21-23 

11-15 

8 

10  +  12 

40-41 

9 

4-6  +  13-16 

Noloscopelus 

21-27 

18-21 

11-14 

8-9 

10-14  +  10-15 

35-40 

10 

4-10  +  9-22 

N.  Parieophus 

23-26 

18-20 

12-14 

37-38 

8-10  +  18-20 

Hmtonia 

14-16 

12-14 

13-15 

8 

10-11  +  13 

37-39 

9 

6-7  +  11-14 

Lampadcna 

13-16 

12-15 

13-18 

8 

8  +  8-9 

35-40 

9 

3-8  +  9-18 

L.  Dorsadena 

14-15 

12-14 

15-16 

8-9 

4-5  +  12 

Taaningichthys 

11-14 

11-14 

12-17 

8 

7-10  +  6-10 

34-41 

8-9 

2-5  +  6-14 

Ceraloscppelus 

13-15 

13-16 

12-15 

8 

6-7  +  6-7 

35-38 

9 

3-5  +  9-16 

Lepidophanes 

11-15 

13-16 

11-14 

8-9 

6-7  +  6-8 

33-37 

9 

3-4  +  8-1 1 

Bolinichthys 

11-15 

11-15 

11-15 

8 

7  +  7-8 

33-36 

9 

3-7  +  11-17 

Triphoturus 

12-16 

13-18 

8-10 

8 

5-7  +  6-7 

30-36 

10-11 

2-4  +  8-11 

Slenobrachius 

12-15 

14-16 

8-10 

8 

6-8  +  7-9 

35-38 

9-10 

5-6  +  12-14 

Panilux 

14-17 

15-18 

10-13 

8 

8  +  8-9 

35-38 

10-11 

4-6  +  11-15 

Lampanyctus 

10-19 

14-21 

0-17 

8 

6-8  +  6-8 

30-40 

8-11 

3-8  +  9-19 

•  Incorrectly  15-27  in  Paxlon,  1972 


(1977)  and  Hulley  (1981)  on  Atlantic  species  and  McGinnis 
(1982)  on  Southern  Ocean  species. 

Most  lantemfishes  make  extensive  vertical  migrations  from 
mesopelagic  depths  to  the  upper  waters  at  night,  some  reaching 
the  surface  (Paxton,  1 967).  The  fisheries  potential  of  myctophids 
and  other  mesopelagic  fishes  has  recently  been  reviewed  (Gjo- 
saeter  and  Kawaguchi,  1980).  Adults  range  in  size  from  20-300 
mm  (Kreflt,  1974)  and  have  a  life  span  of  from  one  year  in 
some  tropical  species  (Clarke,  1973)  to  more  than  five  years  in 
the  few  temperate  species  that  have  been  studied  (Smoker  and 
Pearcy,  1970;  Gjosaeter,  1973:  Kawaguchi  and  Mauchhne,  1982). 

Eggs 

Myctophids  are  oviparous  and  presumably  all  produce  plank- 
tonic  eggs  although  such  have  been  reported  for  only  two  species. 
Sanzo  (1939a)  indicated  that  mature  ovarian  eggs  of  E.  rissoi 
have  the  following  characteristics:  round  shape;  0.80-0.84  mm 
diameter;  segmented  yolk;  single  oil  globule,  ca.  0.28  mm  di- 
ameter; smooth  chorion.  He  illustrated  a  planktonic  egg  with 
similar  characteristics  and  tentatively  identified  it  as  that  o(  E. 
rissoi.  Robertson  (1977)  described  the  planktonic  egg  of  Lam- 
panyctodes hectoris  as  follows:  weakly  oval;  long  axis  0.74-0.83 


mm,  short  axis  0.65-0.72  mm;  strongly  segmented  yolk;  single 
oil  droplet,  0.21-0.23  mm  diameter:  narrow  perivitelline  space; 
chorion  smooth  and  delicate.  He  based  his  identification  on  the 
similarity  of  these  eggs  and  mature  ovarian  eggs  of  running  ripe 
L.  hectoris  captured  at  the  same  time  by  trawl. 

We  have  observed  planktonic  eggs  similar  to  those  described 
by  Robertson  (1977)  but  have  not  found  them  with  advanced 
embryos  that  could  be  matched  with  co-occurring  yolk-sac  myc- 
tophid  larvae.  The  fact  that  these  and  other  types  of  eggs  ten- 
tatively identified  as  myctophids  occur  in  relatively  low  abun- 
dance compared  with  myctophid  larvae  led  Moserand  Ahlstrom 
(1970)  to  suggest  that  the  fragile  chorion  breaks  in  contact  with 
plankton  nets  and  the  embryo  is  extruded  through  the  mesh. 

Larvae 

Moser  and  Ahlstrom  (1970)  reviewed  the  literature  on  myc- 
tophid larvae:  however,  numerous  recent  contributions  have 
advanced  our  knowledge  of  the  group  and  are  listed  in  Table 
61.  Of  the  32  recognized  genera  of  myctophids,  larvae  have 
been  described  for  all  but  Hintonia.  The  larval  stages  of  myc- 
tophids provide  sets  of  characters  that  are  useful  at  levels  of 
systematic  analysis  from  species  separation  to  hypotheses  of 


222 


ONTOGENY  AND  SYSTEM ATICS  OF  HSHES-AHLSTROM  SYMPOSIUM 


Table  61.    Summary 
abbreviated  as  follows: 


OF  Literature  Containing  Illustrations  of  Developmental  Stages  of  Myctophids.  Frequently  cited  authors  are 
Ahlstrom  (A),  Belyanina  and  Kovalevskaya  (B  +  K),  Dekhnik  and  Sinyukova  (D  +  S),  Moser  and  Ahlstrom  (M  +  A), 
Pertseva-Ostroumova  (P-O),  Shiganova  (S),  Tuning  (T). 


Species 


Single  larval  stage 


Multiple  larval  stages 


Transforming  stage 


Juvenile  stage 


Benlhosema 
fibulatum 
glaciale 

panamense 

pterota 
suborbitale 

Bolinichthys 

dislofax 
pyrsobolus 

Centrobranchus 

andrae 
breviroslris 
choerocephalus 
nigroocellatus 

Ceratoscopelus 
maderensis 

townsendi 
warming! 

Diaphus 

agassizii 

holli 

malayanus 

melapoclampus 

mollis 

pacificus 

rafinesquei 

Iheta 

Diogenichthys 
atlanticus 


laternalus 
panurgiis 

Electrona 
antarctica 
carhbergi 
rissoi 

subaspera 
Gonichthys 
coccoi 

tenuiculus 
Gymnoscopelus 
bolini 
braueri 

fraseri 
mcholsi 

opislhoplerus 
Hygophum 
atraium 
henoiti 
brunni 
hanseni 
hygomi 


M  +  A,  1974 
Holt,  1898;  S,  1977 


M  +  A,  1974;  P-O,  1974 
P-O,  1964;  M  +  A,  1974 


M  +  A,  1974 
P-O, 1964 

P-O, 1974 
P-O,  1964 

M  +  A,  1974 
P-O,  1974 

M  +  A,  1972;  S,  1977 
M  +  A, 1974 
Miller  etal.,  1979; 
Belyanina,  1982b 


D  +  S,  1966 

M  +  A, 1974 

P-O,  1964;  M  +  A,  1974 

P-O,  1964 


M  +  A, 1974 
M  +  A,  1974 


M  -I-  A,  1974 

M  -I-  A,  1974 

P-O,  1964 
P-O,  1964 


P-O.  1974 

T,  1918;  Sparta,  1951; 

M  +  A, 1974 
M  +  A, 1970 
Tsokur,  1981 
P-O,  1974;  Badcockand 

Merrett,  1976;  S,  1977 


Holt,  1898;  T,  1918 
Sparta,  1951 

M  +  A, 1970 

P-O,  1974;  S,  1977 


P-O,  1974 


P-O,  1974 
M  +  A,  1970 

- 

T,  1918;  D  -1-  S,  1966 

T,  1918 

S,  1977 

S,  1977 

P-O,  1975 
T,  1918 
Tsokur,  1975 
Sparta,  1952 
S,  1977 

P-O, 1975 
T,  1918 
Tsokur,  1975 
Sparta,  1952 
S,  1977 

T,  1918 


T,  1918;  A,  1965; 

M  +  A,  1970;  P-O,  1974; 

S,  1977 
A,  1965;  M  +  A,  1970 
P-O,  1974 


T,  1918 


T,  1918;  M  +  A,  1970; 
S,  1977 

M  +  A,  1970 
P-O,  1974 


P-O,  1967;  B  +  K,  1979         - 
B  +  K,  1979 

T,  1918;  Sanzo,  1939a;       Sanzo,  1939a 
D  +  S,  1966;  M  -I-  A,  1970 


M  +  A.  1974 
M  -I-  A,  1974 


T,  1918;  S,  1977; 

D  +  S.  1966 
M  +  A,  1970 

S,  1977 

P-O,  1977;  B  +  K,  1979 

P-O, 1977 

M  +  A,  1972;  P-O,  1977; 

B  +  K,  1979 
Yefremenko,  1977 

M  +  A,  1970 
T,  1918;  S,  1974 

S,  1977 

T,  1918;  P-O,  1974; 
S,  1977 


M  +  A,  1970 
S,  1977 

M  +  A,  1972 


M  +  A,  1970 
T,  1918;  S.  1974 

S,  1977 

T,  1918;  P-O,  1974; 
S,  1977 


Holt,  1898;  T,  1918; 

Sparta,  1951 
M  +  A, 1970 
Tsokur,  1981 
S.  1977 


M  +  A. 1970 


T,  1918;  S,  1977 


P-O,  1975 
T,  1918 
Tsokur,  1975 
Sparta,  1952 
S,  1977 

T,  1918 


T.  1918;  M  +  A,  1970; 
S,  1977 

M  +  A,  1970 


T,  1918;  Sanzo,  1939a; 
M  +  A, 1970 


T,  1918;  S,  1977 


S,  1977 


M  +  A, 1970 
T,  1918;  S,  1974 

S,  1977 

T,  1918;  S,  1977 


MOSER  ET  AL.:  MYCTOPHIDAE 
Table  6 1 .    Continued. 


223 


Species 


Single  larval  stage 


Mullipte  larval  stages 


Transforming  stage 


Juvenile  stage 


macrochir 
pro.ximum 

rcinhardli 
taanmgi 

Idiolychnus 
urolampus 

Kretflichthys 
anderssoni 

Lampadena 
luminosa 

urophaos 
Lainpanyctodes 
hectoris 

Lampanyclus 
achirus 
crocodilus 
jordani 
nohilis 
pusillus 
regalis 
ritleri 

Lampichthys 

procerus 
Lepidophanes 

gaussi 

guerjtheri 

Lohianchia 


M  +  A, 1974 

M  +  A,  1974;  Miller 

et  al.,  1979 
M  +  A,  1974 
M  +  A,  1974 


M  +  A. 1974 


M  +  A,  1974 


M  +  A,  1974;  Miller 
etal.,  1979 


M  +  A.  1974 

P-O,  1964 
Miller  etal.,  1979 

M  +  A,  1974 
M  +  A,  1974 


M  +  A, 1974 
M  +  A,  1972 


S.  1975 
P-O,  1974 

M  +  A,  1970;  S,  1977 


Yefremenko,  1976; 
B  +  K,  1979 


M  +  A,  1972 


Ahlstrom  et  al., 
1976 


T,  1918;  D  +  S,  1966 

T,  1918;  D  +  S,  1966 
A, 1965 

M  +  A, 1972 
S,  1977 


S,  1975 
P-O,  1974 

M  +  A,  1970;  S,  1977 


Yefremenko,  1976 


M  +  A,  1972 

Ahlstrom  et  al., 
1976 


T,  1918 


T,  1918 
Bolin,  1939b 


M  +  A,  1972 


M  +  A,  1972;  S,  1977 


S,  1975 


M  +  A,  1970;  S,  1977 


Yefremenko,  1976 


Ahlstrom  et  al., 
1976 


T,  1918 


T, 1918 


do/Ieini 

M  +  A,  1974 

T,  1918;  D  +  S,  1966; 

T,  1918;  S,  1977 

T,  1918;  S,  1977 

S,  1977 

gemellari 

Sanzo,  1931c;  P-O, 
1964;  M  +  A,  1974 

T,  1918 

T,  1918 

T,  1918 

Loweina 

rara 

M  +  A.  1974 

M  +  A,  1970;  P-O,  1974 

M  +  A,  1970 

M  -1-  A,  1970 

lerminata 

Belyanina,  1982b 

- 

— 

— 

Meleleclrona 

ventralis 

M  +  A, 1974      - 

— 

— 

— 

Myctophum 

asperum 

P-O.  1964;  M  -1-  A,  1974 

Imai,  1958;  P-O,  1974 



Imai,  1958;  P-O,  1974 

aurolaternatum 

M  +  A,  1974 

— 



__ 

brachygnathum 

M  -1- A,  1974 

_ 





lychnobium 

M  +  A,  1974;  P-O,  1974 

— 

P-O,  1974 



nilidulum 

M  +  A,  1974 

M  +  A.  1970;  P-O,  1974 



M  +  A,  1970 

oblusirostre 

M  +  A,  1974 

_ 

— 



punctalum 

M  +  A, 1974 

Sanzo,  1915b;  T,  1918; 
S.  1977 

Sanzo,  1915b; 
S,  1977 

T, 

1918; 

T,  1918;  S,  1977 

selenops 

M  +  A, 1974 

— 

_ 



spinosum 

M  +  A, 1974 

P-O,  1974 

P-O,  1974 

P-O,  1974 

Nololychnus 

valdiviae 

P-O,  1964;  M  +  A,  1974 

T,  1918 

T.  1918 

T,  1918 

Notoscopelus 

caudispinosus 

Belyanina,  1982b 

_ 

— 



elongatus 

— 

T,  1918 

T,  1918 

T, 1918 

resplendens 

M  +  A,  1974 

M  +  A.  1972;  Badcock  and 
Merrett,  1976;  S,  1977 

M  +  A,  1972; 

s. 

1977 

224 


ONTOGENfY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 

Table  61.    Continued. 


Species 

Single  larval  stage 

Multiple  larval  stages 

Transforming  stage 

Juvenile  stage 

Parvilux 

ingens 

M  +  A,  1974 

- 

- 

- 

Prolomyclophum 

arcticum 

— 

T, 1918 

T, 1918 

T,  1918 

boHni 

_ 

P-O,  1967;  B  +  K,  1979 

— 



chilensis 

M  +  A. 1974 

— 

_ 

— 

crockeri 

— 

M  +  A, 1970 

— 

M  -1- A,  1970 

normani 

P-O.  1967;  M  +  A, 

1974 

_ 

_ 

P-O,  1967 

parallelum 

— 

P-O,  1967;  B  -1-  K,  1979 

— 

_ 

subparallelum 

M  +  A,  1974 

— 

— 

— 

tenisoni 

M  +  A, 1974 

— 

— 

_ 

ihompsom 

P-O,  1964 

P-O,  1967;  M  +  A,  1970 

- 

M  -1-  A,  1970 

Scopelopsis 

nndlipiinctatus 

— 

M  +  A,  1972;  P-O,  1972 

M  +  A,  1972;  P-O,  1972; 
M-F  A,  1974 

M  +  A, 1972 

Stenobrachius 

leucopsarus 

P-O,  1964;  M  -1-  A, 

1974 

Fast,  1960;  A,  1965; 
A, 1972b 

Fast,  1960 

Fast,  1960 

Symbolophours 

hoops 

— 

P-O, 1974 

— 

— 

californiense 

P-O,  1964;  M  -1-  A, 

1974 

A,  1965;  M  +  A,  1970; 
P-O,  1974 

M  +  A,  1970;  P-O,  1974 

— 

evermanni 

P-O,  1964 

P-O,  1974 

P-O,  1974 

P-O,  1974 

veranyi 

— 

Sanzo,  1915b;  T,  1918; 
D-l-S,  1966 

Sanzo,  1 9 1 5b;  T,  1918 

Sanzo,  1 9 1 5b,  T,  1918 

Taaningichthys 

minimus 

- 

M  +  A, 1972 

- 

- 

Tarletonbeania 

crenularis 

P-O,  1964;  M  ^  A, 
P-O,  1974 

1974; 

A,  1965;  M  -i-  A,  1970 

Bolin,  1939b;  M  +  A,  1970 

M  -1-  A,  1970 

Tripholurus 

mexicanus 

M  +  A,  1974 

A,  1965;  A,  1972b 

_ 

_ 

nigrescens 

Moser,  1981 

- 

- 

- 

Op 


po  XXvo 


Fig.  1 14.     Hypothetical  myctophid  showing  photophore  terminology,  from  Paxton  (1972). 


MOSER  ET  AL.:  MYCTOPHIDAE 


225 


Fig.  1 15.  Larvae  of  Electronini.  (A)  Krefftichlhys  anderssoni.  15.7  mm;  (B)  Protomyctophum  normani.  15.2  mm;  (C)  P.  Heirops  ihompsom, 
13.8  mm;  (D)  Elcclrona  rissoi.  7.9  mm;  (E)  £.  antarclica.  12.7  mm;  (F)  Melelectrona  ventralis,  10.3  mm.  A,  B,  E,  F  from  Moser  and  Ahlstrom 
(1974);  C  and  D  from  Moser  and  Ahlstrom  (1970). 


226 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  62.    Sequence  of  Formation  of  Photophores  which  Appear  in  Fourteen  Genera  of  Myctophidae.  The  Bfj  appear  first  in  all  genera 

listed.  Parentheses  indicate  photophores  appear  late  in  larval  period. 


PO, 


PO, 


PVO,       PLO 


VO,        AOa,       AOa, 


Benthosema 
suborbitale 
glaciate 
pterola 
fibulalum 

Diogenichthys 
lalernalus 
atlanticus 

Myclophum 
spinosum 
lychiwbium 
asperum 
brachygnalhum 
obtusirosire 
selenops 

Lobianchia 
Diaphus 

theta 

pacificus 

Gymnoscopelus 

Lampanyctodes 

Scopelopsis 

Lainpichthys 

Notoscopelus 

Lampadena 

Ceratoscopelus 

Lepidophanes 

Bolinichlhys 


22--2  1  1  333------33 

-         -         -         -        (1)       (1)       (1)       (1)       (1)       (1)        -------         - 

--1-4  6---2  3  5--5-6- 

--1--3  5--2---6--46 

_________         1         ________ 


(5) 


1 
2 
2 
2 

3 

1 

1 

(1) 


(4) 
(3) 


(7)       (8) 
(5)        - 


(4) 


(1) 


-  (9)       (3)       (6) 

-  -        (6)        - 


3 
3 
1 

2 

1 

(1) 


ordinal  relationships.  One  set  is  the  size  at  various  develop- 
mental milestones.  Myctophid  larvae  hatch  at  about  2  mm  length 
with  a  yolk-sac  remnant.  Notochord  flexion  occurs  in  a  narrow 
size  interval  (0.5-2.0  mm)  and  the  size  at  mid-flexion  is  typically 
about  half  the  maximum  larval  size.  Size  at  transformation  also 
occurs  within  a  short  length  interval,  usually  not  exceeding  2 
mm.  Most  myctophid  species  transform  in  the  length  range  of 
12-19  mm,  although  some  (e.g.,  Electrona  rissoi,  Notolychnus 
valdiviae)  are  as  small  as  9-10  mm  at  transformation  and  some 
species  of  Symbolophorus  reach  about  23  mm  before  transfor- 
mation. Gymnoscopelus  nicholsi  has  the  largest  larvae  recorded, 
up  to  28  mm. 

Head,  body,  and  gut  shape  are  distinctive  for  most  species 
and  within  most  genera  there  is  a  similarity  of  shape  (Figs.  1 1 5- 
124).  While  most  myctophid  larvae  are  moderately  slender, 
body  shape  can  range  from  highly  attenuate  (e.g.,  Hygophum 
reinhardti)  to  markedly  robust  (e.g.,  some  Myctophum  and 
Lampanyctus  species).  Some  are  deep-bodied  but  laterally  com- 
pressed (e.g.,  Gonichthyini).  Robust  larvae  and  deep-bodied, 
laterally  compressed  forms  tend  to  have  large  heads  and  jaws, 
while  attenuate  forms  have  flat  heads. 

The  eye  is  varied  in  size  and  shape  and  provides  numerous 


characters.  In  the  Myctophinae  the  eyes  are  elliptical  in  outline 
in  contrast  to  most  Lampanyctinae  which  have  rounded  eyes. 
Further  specializations  in  Myctophinae  are  the  presence  of  var- 
iously shaped  choroid  tissue  on  the  ventral  surface  of  the  eye 
in  most  genera  and  eye  stalks  in  several  genera.  Among  1am- 
panyctine  genera  eyes  are  sessile  and  only  Lobianchia  doflcini 
and  species  of  Triphoturus  have  markedly  narrowed  eyes  with 
choroid  tissue. 

The  gut  has  distinctive  transverse  rugae  and  ranges  from  short, 
to  elongate,  to  trailing  free  from  the  body.  In  most  myctophids 
it  extends  to  about  the  midpoint  of  the  body  and  is  slightly  S- 
shaped.  The  curvature  tends  to  be  more  pronounced  in  taxa 
with  short  guts.  In  two  myctophine  genera  (Metelectrona  and 
some  Hygophum  species)  the  anterior  section  of  the  gut  is  small 
in  diameter  and  opens  dorsally  into  the  relatively  larger  pos- 
terior section. 

In  most  myctophids,  ray  formation  and  ossification  of  fins 
proceeds  in  the  following  sequence:  caudal,  pectoral,  anal,  dor- 
sal, and  pelvic.  However,  in  some  Symbolophorus  species  the 
pelvic  fin  forms  early  and  ossification  of  rays  precedes  that  of 
the  anal  and  dorsal  fins.  In  most  species  the  pectoral  fin  is 
relatively  small,  but  deep-bodied  and  robust  forms  in  both 


Fig.  1 16.  Larvae  of  Myctophini.  (A)  Benthosema  glaciale.  10.5  mm;  (B)  B.  suborbitale.  9.2  mm;  (C)  B.  pterola.  8.5  mm;  (D)  B.  fibulatum. 
8.7  mm;  (E)  Diogenichthys  lalernalus.  1.1  mm;  (F)  D.  atlanticus.  8.8  mm.  A-D  from  Moser  and  Ahlstrom  (1974);  E  and  F  from  Moser  and 
Ahlstrom(1970). 


^^JmiJdJ^i^L 


*ss;;;^a 


228 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  1 17.  Larvae  of  Myctophini.  (A)  Hygophum  proximum.  8.9  mm;  (B)  H.  taaningi.  6.8  mm;  (C)  H.  reinhardti.  12.8  mm;  (D)  Symbolophorus 
californiense.  1 1.5  mm;  (E)  Myctophum  punclalum.  13.6  mm;  (F)  M.  aurolalernatum.  26.0  mm.  A.  B.  E,  F  from  Moser  and  Ahlstrom  (1974);  C 
and  D  from  Moser  and  Ahlstrom  (1970). 


MOSER  ET  AL.:  MYCTOPHIDAE 


229 


subfamilies  have  large  fins  and  fin  bases.  In  Symbolophorus  the 
fin  base  is  uniquely  shaped  and  in  Lobianchia  the  fin  blade  has 
a  unique  shape.  In  two  genera  (Loweina.  Tarletonbeania)  the 
lowermost  pectoral  ray  is  elongate  and  ornamented.  The  finfold 
is  enlarged  in  many  myctophine  genera  and  greatly  enlarged  in 
one  myctophine  tribe,  the  Gonichthyini. 

Myctophids,  with  the  exception  of  Notolychnus  and  Taan- 
ingichthys,  develop  the  middle  branch iostegal  photophore  (Br,) 
during  the  larval  period.  It  is  located  posteroventral  to  the  orbit 
but  during  transformation  assumes  a  position  beneath  the  orbit 
on  the  branchiostegal  membrane.  Three  myctophine  genera  and 
1 1  lampanyctine  genera  develop  additional  photophores  during 
the  larval  period;  however,  the  Br,  is  always  the  first  to  develop. 
The  larval  photophore  complements  and  the  sequence  of  ap- 
pearance of  constituent  photophores  are  useful  characters. 

Myctophid  species  have  distinct  melanophore  patterns,  with 
the  exception  of  the  large  genus  Diaphus,  for  which  only  a  few 
specific  patterns  have  been  identified.  Most  genera  may  be  sep- 
arated by  overall  similarity  of  pattern  among  their  species  and 
some  have  unique  melanophore  loci.  There  are  no  clear  patterns 
for  tribes  or  subfamilies  although  certain  pigment  loci  are  per- 
sistent in  some  tribes  (e.g.,  caudal  fin  base  spots  in  diaphines; 
dorsal  midline  series  in  gymnoscopelines). 

In  the  following  summary  of  key  larval  characters,  the  genera 
are  listed  for  convenience  as  in  Moser  and  Ahlstrom  (1970. 
1972,  1974)  and  the  sequence  does  not  necessarily  imply  rela- 
tionship. Likewise,  the  species  groups  serve  only  to  identify 
phenotypically  similar  larval  types.  Larvae  of  a  majority  of  myc- 
tophid genera  have  a  moderately  slender  body,  a  head  of  mod- 
erate size,  with  a  slightly  convex  dorsal  profile  and  a  pointed 
snout  of  moderate  length.  Body  and  head  shape  are  noted  only 
when  they  depart  from  this  morph.  In  Myctophinae  eye  shape 
is  noted  when  it  is  markedly  elliptical  and  size  is  noted  only 
when  larger  or  smaller  than  typical.  In  Lampanyctinae  eye  shape 
is  noted  only  when  it  departs  from  the  round  condition  and  eye 
size  only  when  larger  or  smaller  than  typical.  Choroid  tissue  is 
described  only  when  it  is  present.  Gut  length  and  shape  are 
described  only  if  there  is  a  departure  from  the  typical  morph  — 
a  slightly  S-shaped  gut  that  extends  to  about  midbody.  The  most 
persistent  pigment  locus  in  myctophid  larvae  is  above  or  to  the 
side  of  the  free  terminal  section  of  the  gut,  thus  only  the  lack 
of  this  pigment  is  noted.  Larval  photophores,  in  addition  to  the 
Br,,  and  their  sequence  of  appearance  are  shown  in  Table  62. 

Myctophinae 

Krefflichthys.  — Fig.  1 15 A;  head  small  with  short  snout;  conical 
choroid  tissue;  gut  straight,  extending  beyond  midbody;  dorsal 
fin  displaced  posteriad;  lateral  gut  and  postanal  median  ventral 
melanophore  series;  large  lateral  hypural  pigment  patch. 


Protomyctophum.  —  ¥\%.  1 15B,  C;  two  subgenera;  head  small  to 
moderate  in  size;  gut  short,  wide  space  between  anus  and  anal 
fin;  head  pigment  lacking  except  in  otic  region  of  P.  Heirops 
chilensis;  some  species  may  have  melanophores  on  lateral  gut, 
above  gut  on  trunk,  above  gas  bladder,  in  postanal  ventral  mid- 
line series,  prominent  pigment  on  lateral  hypural  region.  P. 
Heirops:  Fig.  1 1 5C;  characters  similar  to  P.  Protomyctophum 
except  eye  narrower. 

Eleclrona— Fig.  1 15D,  E;  body  moderately  slender  to  moder- 
atey  deep;  head  moderately  large;  snout  blunt  or  pointed;  gut 
short,  somewhat  saccular,  strongly  S-shaped;  space  between  anus 
and  anal  fin  not  as  large  as  in  Protomyctophum;  three  morphs. 
E.  subaspera-E.  carlsbergi:  eye  slightly  elliptical,  small  lunate 
choroid  mass  in  E.  carlsbergi;  pigment  above  gut;  E.  subaspera 
has  pigment  lateral  to  cleithrum.  E.  rissoi:  Fig.  1 1 5D;  head  large, 
broad;  eye  very  narrow;  pigment  at  lower  jaw  symphysis,  on 
pectoral  fin  blade.  E.  antarctica:  Fig.  1 1 5E;  body  and  head  lat- 
erally compressed;  gut  mass  protrudes  ventrally  from  body  pro- 
file; eye  small,  narrow,  with  bicolored  elongate  conical  choroid 
mass;  pigment  on  upper  jaw,  pectoral  fin  blade,  lateral  gut, 
lateral  hypural  region. 

Metelectrona.  — Fig.  1 1 5F;  body  and  head  laterally  compressed; 
dorsal  finfold  enlarged  with  fin  base  initially  separated  from 
body;  lunate  choroid  mass;  anterior  gut  section  with  small  di- 
ameter, opening  dorsally  into  somewhat  saccular  posterior  sec- 
tion; pigment  below  lower  jaw  and  on  isthmus. 

Benthosema.  —  Fig.  1 16A-D;  two  morphs;  photophores  (Table 
62).  B.  glaciale-B.  sitborbitale:  Fig.  1 16A,  B;  eyes  narrow,  with 
small  lunate  choroid  mass;  gut  moderately  short  in  preflexion 
larvae  with  space  between  anus  and  anal  fin;  pigment  on  snout, 
lower  jaw,  hindbrain,  lateral  and  ventral  cleithral  region;  pig- 
ment above  gut  in  B.  glaciate.  B.  pterota-B.  fibulatum:  Fig.  1 I6C. 
D;  eyes  less  narrow  than  in  above  morph,  with  sliver  of  choroid 
tissue  or  none;  gut  extends  to  about  midbody  with  no  space 
between  anus  and  anal  fin;  preflexion  larvae  with  melanophore 
series  on  lateral  gut  and  on  postanal  ventral  midline,  coalescing 
to  a  single  melanophore;  lateral  cleithral  pigment;  lower  jaw 
pigment  in  B.  pterota. 

Diogenichthys.  — Fig.  I16E,  F;  eyes  very  narrow  in  preflexion 
stage,  less  so  in  postflexion;  photophores  (Table  62);  pigment 
series  on  lateral  gut  and  on  postanal  ventral  midline,  increasing 
with  development;  spot  at  caudal  fin  base;  pigment  on  tip  of 
lower  jaw  in  D.  laternatus;  D.  atlanticus  has  spot  on  trunk  above 
terminal  gut  flexure  and  pigment  on  symphyseal  barbel. 


Fig.  1 18.  Larvae  of  Myctophum.  (A)  M.  phengodes.  9.8  mm;  (B)  M.  asperum.  6.8  mm;  (C)  M.  brachygnathum.  7.5  mm;  (D)  M.  selenops.  7.8 
mm;  (E)  A/,  spinosum,  9.0  mm.  From  Moser  and  Ahlstrom  (1974). 

Fig.  1 19.  Larvae  of  Gonichthyini.  (A)  Loweina  rara.  17.6  mm;  (B)  Tarletonbeania  crenularis.  18.9  mm;  (C)  Gomchthys  tenutculus.  1.1  mm; 
(D)  Centrobranchus  choerocephalus.  7.3  mm.  From  Moser  and  Ahlstrom  (1970). 

Fig.  120.  Larvae  of  Lampanyctinae.  (A)  Notolychnus  valdiviae.  8.7  mm;  (B)  Lobianchia  dojleini.  8.2  mm;  (C)  L.  gemellari.  6.7  mm;  (D) 
Diaphus  theta.  6.9  mm;  (E)  D.  pacificus.  5.2  mm;  (F)  Gymnoscopelus  nicholsi.  23.5  mm.  A-E  from  Moser  and  Ahlstrom  (1974);  F  from  Moser 
and  Ahlstrom  (1972). 

Fig.  121.  Larvae  of  Lampanyctinae.  (A)  Lampanyctodes  hectoris.  1 3.0  mm;  (B)  Scopelopsis  muttipunctatus.  1 3.4  mm;  (C)  Lampichthys  procerus, 
1 4.5  mm;  (D)  Notoscopelus  resplendens.  11.2  mm;  (E)  Lampadena  lununosa.  1 2.8  mm;  (F)  Taanmgichthys  minimus.  1 4.4  mm.  A  from  Ahlstrom 
et  al.  (1976);  B,  C.  F  from  Moser  and  Ahlstrom  (1972);  D  and  E  from  Moser  and  Ahlstrom  (1974). 


230 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES -AHLSTROM  SYMPOSIUM 


^yig^i®^' 


^^^> 


MOSER  ET  AL.:  MYCTOPHIDAE 


231 


232 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


MOSER  ET  AL.:  MYCTOPHIDAE 


233 


234 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Hygophum.  — Fig.  1 17A-C;  diagnostic  pattern  of  melanophores 
at  the  cleithral  symphysis  and  isthmus  region  consisting  of  paired 
pigment  dashes  that  form  a  median  Hne  as  the  series  extends 
forward  on  the  isthmus;  Br,  photophore  forms  late  in  larval 
period;  three  morphs.  H.  proximum-H.  hygomi-H.  benoiti-H. 
hanseni-H.  brunni:  Fig.  1 1 7A;  eye  moderately  narrow  with  con- 
ical choroid  tissue;  pigment  sparse  in  most  species  with  some 
lateral  gut  spots  in  all  species;  some  species  may  have  pigment 
on  hypaxial  myosepta,  jaws,  lateral  cleithral  region,  base  of  cau- 
dal rays.  H.  atratuin-H.  reinhardti:  Fig.  1 1 7C;  body  very  slen- 
der; head  flat;  eyes  very  narrow,  on  short  stalks;  elongate  conical 
choroid  mass;  gut  almost  straight,  small  diameter;  pigment  se- 
ries along  lateral  gut  and  hypaxial  myosepta;  pigment  at  caudal 
fin  base;  pigment  on  lower  jaw  symphysis  in  H.  atratum.  H. 
macrochir-H.  taaningi:  Fig.  117B;  body  and  head  deep  and 
laterally  compressed;  eyes  large,  relatively  wide;  no  choroid  tis- 
sue; anterior  gut  section  narrow  in  diameter,  opening  dorsally 
into  somewhat  saccular  posterior  section;  H.  macrochir  has  pig- 
ment on  upper  and  lower  jaw  and  a  patch  of  melanophores  on 
posterior  gut  section;  H.  taaningi  has  pigment  on  gular  region 
and  lateral  surface  of  cleithrum. 

Symbolophorus.  — Fig.  1 17D;  head  broad,  somewhat  flat;  eyes 
slightly  stalked,  conical  choroid  mass;  pectoral  fin  large  with 
supernumerary  rays,  base  wing-shaped,  rays  ossify  early;  pelvic 
fin  large,  early-forming  in  some  species;  dorsal  finfold  well  de- 
veloped with  fin  base  forming  in  it;  pigment  series  on  lateral 
gut  and  postanal  ventral  midline  in  preflexion  larvae;  pigment 
on  snout,  hindbrain,  lateral  cleithral  region,  isthmus,  paired  fins. 

Myctophum.— Figs.  1 17E,  F  and  1 18A-E;  at  least  five  distinct 
morphs,  all  but  M.  aurolaternatum  with  enlarged  fan-shaped 
pectoral  fins,  some  with  supernumerary  rays  and  early  ossifi- 
cation; conical  choroid  mass.  M.  aurolaternatum:  Fig.  117F; 
body  very  slender;  head  somewhat  flat;  eyes  small,  on  elongate 
stalks;  gut  straight,  at  midbody  becomes  trailing,  extending  to 
well  beyond  caudal  fin;  dorsal  finfold  well  developed,  fin  base 
forms  at  its  margin;  pigment  series  on  lateral  gut,  evenly  dis- 
tributed on  trailing  section,  except  heavier  near  terminus;  pig- 
ment on  jaws,  isthmus,  opercle,  branchiostegal  membrane,  pec- 
toral fin,  anal  fin  base,  caudal  fin.  M.  nitidulum-M .  punctatum: 
Fig.  1 1 7E;  body  moderately  slender  to  slightly  deep;  head  broad, 
somewhat  flat  in  preflexion  stage;  eyes  on  short  stalks;  numerous 
small  melanophores  on  snout,  jaws,  brain,  isthmus,  branchio- 
stegal membrane;  two  rows  of  melanophores  on  ventral  surface 
of  gut;  opposing  melanophores  on  postanal  dorsal  and  ventral 
midline;  pigment  on  pectoral  fin  base  and  blade  and  at  base  of 
caudal  rays.  M.  phengodes:  Fig.  1 1 8 A;  body  and  head  moder- 
ately deep;  similar  to  M.  nitidulum,  except  pigment  sparse  and 
eyes  not  stalked;  pigment  at  base  of  pectoral  fin  rays.  M.  spi- 
nosum-M.  lychnohium:  Fig.  1 18E;  head  with  convex  dorsal  pro- 
file and  long  snout  giving  the  larva  a  fusiform  appearance;  long 
axis  of  eye  rotated  towards  horizontal;  photophores  (Table  62); 
head  heavily  pigmented  on  jaws,  brain,  postorbital  and  oper- 
cular regions;  pigment  above  gut  on  trunk,  embedded  in  my- 
osepta in  M.  spinosum;  opposing  dorsal  and  ventral  midline 
blotches,  larger  and  more  deeply  embedded  in  M.  spinosum 
with  embedded  myoseptal  pigment  along  horizontal  septum; 
blotch  at  base  of  caudal  rays.  M.  asperum-M .  brachygnathum- 
M.  obtusirostre-M.  selenops:  Fig.  118B-D;  body  deep,  robust; 
head  broad,  deep  with  convex  dorsal  profile  and  large  snout; 
eye  relatively  larger  than  in  other  morphs;  choroid  tissue  broadly 


conical,  except  in  M.  selenops  where  it  is  elongate  and  pigmented 
at  tip;  photophores  (Table  62);  head  pigment  similar  to  M. 
spinosum;  most  species  have  heavy  pigment  lateral  to  cleithra 
and  on  pectoral  fin  bases;  all  species  lack  trunk  and  tail  pigment, 
except  M.  asperum  which  has  extensive  embedded  myoseptal 
and  dorsal/ventral  midline  blotches. 

Loweina.  — Fig.  1 1 9 A;  body  and  head  moderately  deep,  laterally 
compressed;  dorsal  and  anal  fins  displaced  far  posteriad;  dorsal 
and  ventral  finfolds  greatly  enlarged  and  conspicuously  pig- 
mented to  produce  a  disc-shaped  profile;  eyes  large;  gut  with 
expanded  anterior  section  and  enlarged  terminal  section;  pec- 
toral fin  large  with  lower-most  ray  elongate,  ornamented  with 
pigmented  spatulations;  interorbital  pigment  band;  pigment  at 
lateral  cleithral  surface,  dorsal  fin  origin,  and  opposing  midline 
blotches  at  caudal  peduncle  region. 

Tarletonbeania.  — Fig.  1 198;  similar  to  Loweina.  except  median 
fins  displaced  less  posteriad;  eye  narrower  and  with  lunate  cho- 
roid mass;  four  melanophores  on  periphery  of  brain,  two  me- 
lanophore  series  on  ventrum  of  gut. 

Gonichthys.  — Fig.  1 1 9C;  body  and  head  deep  and  laterally  com- 
pressed, leaf-like;  snout  large,  angulate  in  profile;  eye  small  with 
elongate  conical  choroid  mass,  pigmented  at  tip;  enlarged  dorsal 
and  ventral  finfolds;  pectoral  fins  moderately  large;  pigment  on 
snout,  jaws,  midline  of  brain,  postorbital  and  opercular  regions; 
pigment  on  lateral  hindgut  and  on  trunk  above  gut;  series  of 
embedded  blotches  on  dorsal  midline  of  body,  opposing  blotch- 
es on  postanal  ventral  midline;  large  pigment  patch  on  lateral 
caudal  peduncle  region  in  G.  tenuiculus;  heavy  embedded  pig- 
ment streak  along  horizontal  septum  in  G.  coccoi. 

Centrobranchus.  — Fig.  119D;  morphology  similar  to  Gonich- 
thys except  snout  markedly  blunt  and  rounded  and  terminal  gut 
flexure  less  acute;  two  morphs.  C.  choerocephalus-C.  breviros- 
tris-C.  nigroocellatus:  Fig.  1 19D;  eye  very  narrow  with  unpig- 
mented  choroid  mass  that  exceeds  it  in  length;  pigment  sparse; 
some  at  postorbital-opercular  region,  branchiostegal  membrane, 
ventral  surface  of  liver.  C.  andrae.  eye  wider  than  in  above 
morph  and  with  short  conical  choroid  mass;  pigment  extensive, 
on  snout,  upper  jaw,  dorsal  brain,  opercle,  branchiostegal  mem- 
brane, lateral  hindgut,  ventral  surface  of  liver,  pectoral  fin  base; 
embedded  spots  along  dorsal  midline  with  opposing  spots  along 
postanal  ventral  midline;  embedded  spots  along  horizontal  sep- 
tum in  caudal  peduncle  region. 

Lampanyctinae 

Notolychnus.  —  Fig.  1 20A;  head  relatively  large  with  moderately 
elongate  snout;  eyes  usually  narrow,  often  irregular  in  shape;  gut 
short,  more  so  in  preflexion  stage;  no  photophores,  even  Br, 
lacking;  pigment  on  lateral  hindgut,  gas  bladder,  base  of  caudal 
rays;  a  persistent  but  sparse  postanal  ventral  midline  series. 

Lobianchia.  — Fig.  120B,  C;  body  deep,  robust;  head  broad  with 
large  snout;  pectoral  fins  large;  blade  wing-shaped  with  upper 
rays  longer  than  others;  photophores  (Table  62);  head  unpig- 
mented;  pigment  on  trunk,  on  gut  below  pectoral  fin  base,  on 
pectoral  fin  base  and  blade,  embedded  in  gut  region  anterior  to 
pectoral  fin  base,  along  anal  fin  base,  and  at  base  of  caudal  rays; 
embedded  melanophores  in  myosepta  above  pectoral  fin  be- 
coming extensive  in  postflexion  stage;  two  morphs.  L.  dofleini: 
Fig.  120B;  eye  small,  narrow,  with  lunate  to  squarish  choroid 


MOSER  ET  AL.:  MYCTOPHIDAE 


235 


Fig.  122.  Urvae  of  Lampanyctmae.  (A)  Ceraloscopehis  townsendi.  16.6  mm;  (B)  Lepidophanes  gaussi.  13.5  mm;  (C)  BoUmchthvs  distofax. 
9.4  mm;  (D)  Slenohrachius  leucopsarus.  10.4  mm;  (E)  Parvilux  ingens.  14.4  mm;  (F)  Triphoturus  mexicanus.  10.5  mm.  A-E  from  Moser  and 
Ahlstrom  (1974);  F  from  Ahlstrom  (1972b). 


236 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


mass;  gradual  transition  from  lower  pectoral  rays  to  longer  upper 
rays.  L.  gemellari:  Fig.  120C;  eye  large,  almost  round,  choroid 
mass  a  lunate  sliver;  abrupt  transition  between  lower  pectoral 
rays  and  long  upper  rays. 

Diaphus.  — Fig.  1 20D,  E;  pigment  lacking  on  head;  melanophore 
at  anteroventral  surface  of  liver,  one  or  more  at  midgut  region, 
one  or  more  at  base  of  caudal  rays;  gas  bladder  pigmented;  two 
morphs.  D.  theta:  Fig.  120D;  body  moderately  slender;  head 
moderate  in  size;  photophores  (Table  62);  numerous  melano- 
phores  in  postanal  ventral  midline  series,  persisting  into  post- 
flexion  stage.  D.  pacificus:  Fig.  120E;  body  moderately  deep, 
somewhat  robust;  head  moderately  large;  photophores  (Table 
62);  a  few  melanophores  in  postanal  ventral  midline  series,  usu- 
ally coalescing  to  one  before  flexion  stage. 

Gymnoscopelus.  — Fig.  120F;  photophores  (Table  62);  pigment 
above  brain,  at  lateral  cleithral  region,  above  midgut,  above  gas 
bladder;  postanal  ventral  midline  series  present  but,  in  some 
species,  restricted  to  caudal  peduncle  region;  melanophore  series 
on  each  side  of  dorsal  midline,  in  most  species  extending  be- 
tween caudal  and  dorsal  fins,  in  others  extending  forward  to 
dorsal  fin  origin,  and  in  others  restricted  to  caudal  penduncle 
region;  pigment  at  base  of  caudal  rays;  some  species  have  pig- 
ment on  lateral  hypural  region;  lateral  pigment  patch  at  caudal 
peduncle  in  G.  opisthopterus,  which  also  has  embedded  mela- 
nophores above  vertebral  column. 

Lampanyctodes.  — Fig.  121  A;  photophores  (Table  62);  pigment 
above  brain,  at  anteroventral  surface  of  liver,  above  gas  bladder; 
a  postanal  ventral  midline  series  and  a  series  on  each  side  of 
dorsal  midlme  between  dorsal  and  caudal  fins;  pigment  at  base 
of  caudal  rays  and  at  lateral  hypural  region. 

Scopelopsis.  — Fig.  121B;  photophores  (Table  62);  pigment  sim- 
ilar to  Lampanyctodes  except  additional  melanophores  on  hind- 
brain,  nape,  lateral  cleithral  region;  pigment  rows  along  dorsum 
irregular. 

Lampichthys.  — Fig.  121C;  photophores  (Table  62);  pigment 
similar  to  Scopelopsis  except  dorsal  rows  consist  of  large  closely- 
spaced  melanophores  which  at  maximal  development  extend 
from  caudal  fin  to  dorsal  fin  origin;  a  short  melanophore  series 
along  horizontal  septum  on  caudal  peduncle  in  late  postflexion 
stage. 

Notoscopelus.  — Fig.  12 ID;  photophores  (Table  62);  body  mod- 
erately deep;  head  moderately  large;  eye  large;  snout  becomes 
somewhat  bulbous  at  flexion  stage;  gut  short  in  early  preflexion 
stage,  elongates  to  about  midbody  by  late  preflexion;  pigment 
at  tips  of  jaws,  above  brain,  above  gas  bladder  and  at  lateral 
cleithral  region  in  early  postflexion  larvae;  additional  pigment 
develops  below  lower  jaw,  on  hindbrain  and  nape;  series  of 
melanophores  on  each  side  of  dorsal  midline,  beginning  at  mid- 
body  and  gradually  developing  along  entire  dorsum;  series  along 
horizontal  septum  and  along  anal  fin  base;  pigment  on  base  of 
caudal  rays  and  on  pelvic  and  anal  rays  in  some  species  at  late 


postflexion  stage;  extensive  embedded  myoseptal  pigment  on 
trunk  or  tail  in  postflexion  stages  of  some  species. 

Lampadena.  —  Fig.  1 2 1 E;  photophores  (Table  62);  pigment  above 
brain,  nape,  gut,  gas  bladder;  most  species  have  large  melano- 
phores along  dorsal  midline,  with  opposing  postanal  ventral 
midline  melanophores;  some  species  with  smaller,  more  nu- 
merous melanophores  in  dorsal  and  ventral  series;  embedded 
pigment  above  spinal  column  in  some  species. 

Taaningichthys.  — Fig.  121F;  body  slender;  lower  jaw  projects 
beyond  upper;  no  photophores,  even  Br,  lacking;  pigment  above 
brain,  in  otic  region,  one  to  several  opposing  melanophores  at 
postanal  dorsal  and  ventral  midline;  late  postflexion  larvae  may 
develop  minute  melanophores  along  each  side  of  dorsal  midline; 
pigment  at  base  of  caudal  rays;  series  of  embedded  melano- 
phores above  spinal  column. 

Ceratoscopelus.  — Fig.  122A;  eye  elliptical  in  early  larvae;  pho- 
tophores (Table  62);  pigment  above  gut;  postanal  ventral  mid- 
line series  in  early  larvae,  coalesces  to  a  single  spot  in  postflexion 
larvae;  C.  maderensis  has  short  series  at  dorsal  and  ventral 
midline  in  caudal  peduncle  region;  embedded  pigment  above 
posterior  region  of  spinal  column  in  some  species. 

Lepidophanes.  — Fig.  122B;  eye  small;  photophores  (Table  62); 
usually  two  melanophore  pairs  at  dorsal  midline  in  caudal  pe- 
duncle region  and  one  or  two  ventral  midline  melanophores;  L. 
gaiissi  has  median  melanophore  above  hindbrain  and  median 
ventral  melanophore  below  pectoral  fin  base. 

Bolinichthys.  — Fig.  1 22C;  moderately  deep-bodied;  snout  blunt; 
eye  large;  photophores  (Table  62);  sparse  pigment;  midline  spot 
above  brain,  embedded  otic  spot,  embedded  pigment  above  gut; 
some  species  with  a  sparse  postanal  median  ventral  series  that 
coalesces  to  a  single  melanophore;  B.  distofa.x  has  a  short  series 
on  horizontal  septum;  embedded  pigment  above  posterior  re- 
gion of  spinal  column  in  some  species. 

Triphoturus.  — Fig.  122F;  eye  elliptical  with  choroid  mass;  pig- 
ment at  tip  of  lower  jaw,  at  angular  region  of  jaw,  at  lateral 
cleithral  region;  early  preflexion  larvae  have  paired  lateral  gut 
spots  near  pectoral  fin  base  and  at  midgut;  anterior  pair  coalesces 
to  a  median  position  anteroventral  to  liver,  the  posterior  pair 
becomes  dorsal  to  gut;  pigment  above  gas  bladder;  early  pre- 
flexion larvae  have  postanal  median  ventral  series  that  coalesces 
to  one  or  two  spots;  pigment  along  margin  of  preanal  finfolds; 
a  single  dorsal  spot  at  adipose  fin  in  T.  mexicanus;  a  series  of 
pigment  dashes  on  horizontal  septum  in  T.  nigrescens. 

Stenobrachius.  —  Fig.  1 22D;  gut  melanophores  and  postanal  me- 
dian ventral  series  similar  to  Triphoturus;  pigment  above  brain 
and  nape  in  postflexion  stage;  late  postflexion  larvae  have 
embedded  melanophores  in  trunk  myosepta  and  melanophore 
series  on  each  side  of  dorsal  midline. 

Parviln.x.  — Fig.  122E;  head,  eyes  large;  tapered  body;  gut  short 


Fig.  123.  Larvae  oi  Lampanyclus.  (A)  L.  steinbecki.  6.6  mm;  CalCOH  Sla.  70.200;  (B)  L.  pusiUus.  1.1  mm;  redrawn  from  Taaning  (1918); 
(C)  L.  nobilis,  9.6  mm;  SEFC,  OR  II  7343  Sta.  98;  (D)  L.  par\icauda.  7.5  mm.  SWFC,  Eastropac  Op  Sta.  023;  (E)  L.  crocodilus.  11.5  mm,  redrawn 
from  Tining  (1918). 


MOSER  ET  AL.:  MYCTOPHIDAE 


237 


Fig.  124.  Larvae  of  Lampanyclus.  (A)  L.  rilleri.  10.1  mm;  (B)  L.  idostigma.  7.2  mm.  CalCOFI  6002  Sta.  133.45;  (C)  L.  regalis.  13.0  mm; 
(D)  Lampanyctus  sp.,  8.7  mm;  (E)  L.  achirus.  13.4  mm;  (F^  Lampanyclus  sp.,  9.4  mm.  A,  C,  D,  E  from  Moser  and  Ahlstrom  (1974);  F  from 
Moser(1981). 


MOSER  ET  AL.:  MYCTOPHIDAE 


239 


in  early  preflexion  stage,  elongates  to  midbody  by  flexion  stage; 
in  postflexion  stage  pigment  above  brain,  embedded  in  otic 
region,  lateral  to  cleithrum,  at  anteroventral  region  of  liver;  one 
to  several  dorsal  median  melanophores  and  one  ventral  median 
melanophore  at  caudal  peduncle. 

Lampanyclus.  — Figs.  123.  124;  body  slender;  head  deep;  gut 
short  in  early  preflexion  stage;  during  preflexion  stage  gut  length- 
ens to  midbody.  body  deepens  and  becomes  somewhat  robust 
in  most  species;  pigment  above  brain  in  most  species;  postflex- 
ion larvae  develop  trunk  myoseptal  pigment  that  increases  to 
cover  most  of  the  anterior  trunk  at  transformation;  at  least  6 
morphs.  L.  nohilis-L.  parvicaiida-L.  oinostigma-L,  crocodilus- 
L.  ritteh-L.  idostigma:  Figs.  123C-E.  124A,  B;  body  and  head 
moderately  deep;  eyes,  jaws,  pectoral  fins  moderate  in  size;  pig- 
ment may  be  present  at  snout,  lower  jaw.  opercle,  above  gut, 
anteroventral  surface  of  liver,  at  dorsal  or  ventral  midline  on 
tail.  L.  pusillus-L.  steinbecki:  Fig.  123 A,  B;  deep,  broad  body 
and  head,  very  robust;  snout  blunt;  eyes  large;  dorsal  and  anal 
fins  displaced  posteriad;  pectoral  fins  moderately  large;  L.  pus- 
illus  heavily  pigmented  on  head,  body,  pectoral  fin  base;  series 
along  horizontal  septum;  L.  steinbecki  with  pigment  below  lower 
jaw,  on  opercle.  pectoral  fin  base;  series  along  horizontal  septum 
and  embedded  pigment  on  tail  in  postflexion  larvae.  L.  regalis- 


L.  ater.  Fig.  1 24C;  deep,  large  head  and  body;  snout  elongate, 
jaws  large,  teeth  well  developed,  especially  at  tip  of  upper  jaw; 
preopercular  spines  in  some  species;  dorsal  and  anal  fins  dis- 
placed posteriad;  pectoral  fins  moderate  to  large;  pigment  may 
be  present  at  tips  of  jaws,  embedded  in  snout,  at  postorbital 
and  opercular  regions,  pectoral  and  pelvic  fins;  spot  at  adipose 
fin  in  L.  regalis;  one  or  two  dorsal  spots  in  L.  ater.  Information 
on  L.  ater  irom  H.  Zadoretsky  (Dept.  Zoology,  Univ.  of  Rhode 
Island,  pers.  comm.).  L.  achirus:  Fig.  1 24E;  body  moderately 
deep;  head  and  jaws  large  with  snout  produced  into  toothy  ros- 
trum; dorsal  and  anal  fins  displaced  posteriad;  pectoral  fins  mod- 
erately large;  pigment  on  tips  of  jaws,  embedded  in  snout,  and 
present  at  postorbital  and  opercular  regions.  L.  lineatus-L.  cu- 
prarius:  body  moderately  elongate;  snout  elongate,  jaws  large; 
head  pigment  as  in  L.  achirus;  L.  lineatus  pigment  consists  of 
numerous  melanophores  along  dorsum  and  ventrum  and  at  base 
of  caudal  rays;  L.  cuprarius  has  pigment  above  gut  and  an  ir- 
regular bar  below  dorsal  fin.  Information  from  H.  Zadoretsky 
(pers.  comm.). 

(H.G.M.)  National  Marine  Fisheries  Service,  Southwest 
Fisheries  Center,  P.O.  Box  271,  La  Jolla,  California 
92038;  (J.R.P.)  The  Australian  Museum,  6-8  College 
Street,  Sydney  2000,  Australia. 


Myctophidae:  Relationships 
J.  R.  Paxton,  E.  H.  Ahlstrom  and  H.  G.  Moser 


THE  family  Myctophidae  has  usually  been  placed  in  the  order 
Myctophiformes  (Iniomi.  Scopeliformes)  since  the  work 
of  Regan  (191  la),  who  recognized  two  suborders,  the  Mycto- 
phoidea  and  Alepisauroidea  (ateleopodids,  given  a  third  sub- 
order, are  currently  placed  elsewhere).  The  families  Myctophi- 
dae and  Neoscopelidae  have  long  been  considered  close  relatives; 
they  were  placed  in  one  family  until  1949  (Smith).  Although 
Greenwood  et  al.  (1966:371)  relegated  the  order  to  a  subordinal 
level  within  the  Salmoniformes,  they  pointed  out  that  mycto- 
phoids.  and  neoscopelids  in  particular,  possess  advanced  char- 
acters that  indicate  they  may  be  ancestral  to  the  paracanthop- 
terygian  radiation.  Paxton  (1972:54-55)considered  myctophids 
and  neoscopelids  most  closely  related  to  the  Chlorophthalmi- 
dae.  with  that  evolutionary  line  of  the  Myctophoidea  arising 
from  an  aulopid-like  ancestor.  Moser  and  Ahlstrom  (1970: 141- 
142)  described  the  larval  similarities  in  the  families  Chloroph- 
thalmidae,  Neoscopelidae  and  Myctophidae. 

Family  Relationships 

Rosen  (1973,  1982)  split  ofl"  the  Myctophidae  and  Neosco- 
pelidae as  a  restricted  order  Myctophiformes  which  he  consid- 
ered the  primitive  sister  group  of  both  the  Paracanthopterygii 
and  Acanthopterygii;  the  remaining  myctophiform  families  were 
placed  in  a  new  order  Aulopiformes.  Matsuoka  and  Iwai  (1983) 
found  cartilage  in  the  adipose  fin  of  only  the  Myctophidae  and 


Neoscopelidae  in  the  five  'iniomous'  families  they  studied.  Oki- 
yama  (1974b)  studied  the  relationships  of  the  suborder  Mycto- 
phoidea (sensu  Gosline  et  al.,  1966)  and  based  on  larval  peri- 
toneal pigment  spots  and  the  relationship  of  abdominal  to  caudal 
vertebrae,  three  familial  groups  were  recognized:  Aulopidae- 
Synodontidae-Bathysauridae,  Chlorophthalmidae-Ipnopidae 
and  Neoscopelidae-Myctophidae.  Sulak  (1977)  lumped  the 
Ipnopidae  and  Bathypteroidae  into  the  Chlorophthalmidae  and 
the  Harpadontidae  and  Bathysauridae  into  the  Synodontidae, 
considering  both  groups  arose  from  the  Aulopidae;  he  did  not 
consider  the  position  of  the  Myctophidae.  Schwarzhans  (1978) 
considered  myctophids  and  neoscopelids  most  closely  related 
and  distinct  from  Aulopiformes  on  the  basis  of  otolith  mor- 
phology. 

In  his  excellent  study  of  the  Evermannellidae.  Johnson  ( 1 982) 
presented  a  rigorous  analysis  of  5 1  characters  involving  mostly 
adult  but  some  larval  features.  He  concluded  that  neoscopelids 
and  myctophids  are  most  closely  related  to  each  other,  sharing 
eight  derived  character  slates,  but  that  they  were  the  sister  group 
of  four  families  (Notosudidae,  Scopelarchidae,  Chlorophthal- 
midae and  Ipnopidae)  constituting  a  chlorophthalmoid  group 
within  the  Myctophiformes.  However,  he  noted  only  a  single 
shared  derived  character  in  those  six  families,  and  it  is  shared 
with  part  of  another  line.  Johnson  (1982:95)  placed  the  Aulo- 
pidae in  a  second  line  and  all  remaining  families  in  the  third 


240 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  63.    Characters  of  the  Myctophidae.  (0)  =  plesiomorphic 
state,  (1)  =  apomorphic  state,  (2)  =  different  or  advanced  apomorphic 
state,  1  =  by  outgroup  comparison,  2  =  raised  photophore,  3  =  gener- 
alized larva,  *  =  discussed  in  text. 

Characters 

1.  Jaws  long  (0).  moderate  (1),  short  (2)—*. 

2.  Extrascapulars  2  (0),  1  from  fusion  (1),  1  from  loss  (2)  —  *. 

3.  Cleithral  shelf  absent  (0),  present  (1)—  1. 

4.  Pre  3-9  (0?),  1-2(1?)-*. 

5.  Larval  eyes  round  (0),  narrow  ( 1 )—  1 ,  3. 

6.  Dn  present  (0?),  absent  (1?)-*. 

7.  Moderately  or  strongly  hooked  teeth  in  posterior  dentary  absent 
(0),  present  (1)-1. 

8.  Procurrent  ventral  rays  5-10  (0),  9-15  (1)-1. 

9.  Supramaxillary  present  (0),  absent  (1)—  1,  *. 

10.  PO4  level  (0),  raised  (l)-2. 

1 1 .  Pubic  plate  narrow  (0),  wide  ( 1 )—  1 . 

12.  PO,  and  PO,  level  (0),  raised  (l)-2. 

13.  VO,  level  (0),  raised  (l)-2. 

14.  PVO  horizontal  (0).  angled  (1),  vertical  (2)-2. 

15.  Caudal  luminous  organs  present  (0),  absent  (1)—*. 

16.  AOa,  level  (0),  raised  (l)-2. 

17.  Pol  angled  (0),  horizontal  (l)-2,  *. 

18.  Enlarged  teeth  in  dentary  absent  (0),  present  (I)—  I. 

19.  Vertebrae  28-41  (0),  41-45,  (1)-1,  *. 

20.  VO,  level  (0),  elevated  (1)- 2. 

2 1 .  Enlarged  dentigerous  area  on  anterior  premaxillary  absent  (0),  pres- 
ent (1)-1. 

22.  Secondary  photophores  absent  (0).  present  ( 1 )—  1 . 

23.  Larval  gut  moderate  (0),  initially  short  (1),  long  (2)  — 3,  *. 

24.  Larval  trunk  myoseptal  pigment  absent  (0),  present  (1)— 1,  3. 

25.  Slightly  hooked  teeth  in  posterior  dentary  absent  (0),  present  ( 1  )— 
I. 

26.  Caudal  luminous  organs  not  sexually  dimorphic  (0),  sexually  di- 
morphic (1)—*. 

27.  Larval  photophores  (except  Br,)  absent  (0),  present  (1)—  1,  3.  *. 

28.  Hyomandibular  foramen  behind  anterior  head  (0),  in  anterior  head 
(1)-1. 

29.  Accessory  luminous  tissue  absent  (0),  present  (1)—  1. 

30.  Caudal  luminous  organs  any  other  state  (0),  homogeneous  and 
translucent  ( 1)—*. 

3 1 .  Procurrent  ventral  rays  without  hooks  (0),  with  hooks  ( 1 )—  1 . 

32.  Procurrent  dorsal  rays  without  hooks  (0),  with  hooks  (1)—  1. 

33.  Crescent  of  white  tissue  on  posterior  iris  absent  (0),  present  (1)  — 
1. 

34.  Pol  0(0),  1  (1),  2-3  (2)- 2,  *. 

35.  Dorsal  process  of  opercular  head  of  hyomandibula  absent  (0),  pres- 
ent (1)-1. 

36.  SAOs  weakly  angled  (0),  strongly  angled  (I)  — 2,  *. 

37.  Larval  eyes  moderate  (0),  very  large  (1)— I,  3. 

38.  PLO  level  with  PVO,  (0),  above  PVO,  (l)-2. 

39.  SAO  2,  close  to  VO  and  AO  series  (0),  2-3  above  VO  and  AO 
series  (1)— 2. 

40.  Larval  pectoral  fin  moderate  (0),  large  (1)— 3,  *. 

41.  Mouth  terminal  (0),  subtcrminal  (1)— 1. 

42.  Antorbital  broad  (0),  thin  (1)—  1. 

43.  Larval  fin  fold  small  (0),  extensive  (1)—  1,  3. 

44.  PLO  below  (0)  opposite  or  proximate  to  upper  pectoral  base  (1), 
far  above  upper  pectoral  base  (2)— 2. 

45.  Lower  pharyngeal  teeth  conical  (0),  pegs  or  plates  (1)—  1. 

46.  Nasal  trough-shaped  (0),  convex  (1)—  1. 

47.  Larval  lower  pectoral  ray  not  elongate  (0),  elongate  (1)—  1,  3. 

48.  Gill  rakers  lathe-like  (0),  as  tooth  plates  (1)— 1. 

49.  Dorsal  hypurals  4  (0),  3-2  (1).  1  (2)-  I. 

50.  Coracoid  fenestra  present  (0),  absent  (1)—  1. 

51.  Double  row  of  isthmus  pigment  in  larvae  absent  (0),  present  (1)  — 
1,  3. 

52.  Premaxillary  teeth  conical  (0),  flattened  (1)—  1. 

53.  Larval  pectoral  base  fan-shaped  (0),  wing  shaped  (1)—  1,  3. 

54.  Larval  head  pigment  present  (0),  absent  (1)—  1,  3. 


Table  63.    Continued. 


55.  Larval  choroid  tissue  absent  (0).  present  (1)— 1, 

56.  Larval  body  width  moderate  (0),  thin  (1)—  1,  3. 

57.  Larval  gut  uniform  (0),  bipartite  ( 1 )—  1 ,  3. 

58.  Ossified  distal  pectoral  radials  0  (0),  1-7  (1)— 1. 

59.  CO,  keel  or  ridge  absent  (0),  present  (1)—  1,  *. 


group  (the  alepisauroids  plus  synodontoids)  in  his  arrangement 
of  the  order.  We  do  not  have  further  evidence  to  present  in 
favour  of  any  of  the  above  hypotheses  (but  do  note  the  coiled 
gut  of  neoscopelid  lai-vae  resembles  the  condition  found  in  higher 
groups). 

Generic  Relationships 

Paxton  (1972)  analyzed  features  of  the  osteology  and  pho- 
tophore patterns  of  the  Myctophidae  and  presented  a  taxonomy 
outlining  his  views  of  evolutionary  relationships  that  included 
two  subfamilies  (Myctophinae  and  Lampanyctinae),  six  tribes 
(Myctophini,  Gonichthyini,  Notolychnini,  Lampanyctini,  Dia- 
phini  and  Gymnoscopelini),  28  genera  and  two  subgenera.  The 
Myctophinae  was  considered  the  more  primitive  of  the  subfam- 
ilies, while  the  monotypic  Notolychnini  was  provisionally  placed 
in  the  Lampanyctinae.  In  four  papers  Moser  and  Ahlstrom  ( 1 970, 
1972,  1974;  Ahlstrom  et  al.,  1976)  detailed  the  larval  charac- 
teristics of  all  but  two  genera  of  Myctophidae  and  translated 
their  findings  into  a  picture  of  evolutionary  relationships.  The 
relationships  proposed  by  Paxton  and  Moser  and  Ahlstrom  were 
strikingly  similar  overall  and  in  many  details.  The  larval  studies 
supported  the  recognition  of  two  subfamilies  composed  of  the 
same  genera  indicated  by  the  adult  analysis,  highlighted  the 
enigmatic  features  of  Notolychnus.  and  recognized  three  addi- 
tional tribes  in  the  Lampanyctinae.  Notable  differences  in  the 
conclusions  of  the  two  studies  included  consideration  of  the 
Lampanyctinae  as  the  most  primitive  subfamily  by  Moser  and 
Ahlstrom,  non-recognition  of  the  tribe  Gonichthyini  ( Tarleton- 
beama.  Loweina.  Gonichthys,  Ccntrohranchus)  as  a  monophy- 
letic  taxon  in  the  larval  study,  inclusion  of  the  genera  Taan- 
ingichthys.  Lampadena.  Bolinchthys,  Lepidophanes  and 
Ceratoscopelus  in  the  tribe  Gymnoscopelini  by  Moser  and  Ahl- 
strom and  the  tribe  Lampanyctini  by  Paxton,  and  recognition 
of  the  genera  Metelectrona  and  Parvilux  as  valid  genera  on  the 
basis  of  larval  characters,  which  Paxton  had  synonymized  with 
Electrona  and  Lampanyctus  respectively  on  the  basis  of  adult 
features.  Neither  study  restricted  characters  to  the  derived  state 
and  the  proposed  phylogenies  were  based  on  overall  similarities. 
The  present  work  will  attempt  an  analysis  of  derived  character 
states  and  re-examine  the  proposed  relationships  within  the 
family. 

We  have  used  as  character  states  (Table  63)  features  of  adult 
osteology  and  photophore  patterns  as  described  by  Paxton  (1972), 
and  features  of  larvae  as  described  by  Moser  and  Ahlstrom 
(1970,  1972,  1974)  and  Ahlstrom  et  al.  (1976)  summarized  in 
Moser  et  al.  (this  volume).  The  distribution  of  the 
character  states  among  the  genera  (we  have  not  considered  sub- 
genera in  this  analysis)  is  tabularized  (Table  64).  The  criteria 
for  determining  apomorphic  character  states  have  been  consid- 
ered by  many,  including  Marx  and  Rabb  (1972)  and  Zehren 
(1979:153).  We  have  used  three  criteria,  the  numbers  of  which 
are  listed  after  each  character  in  Table  63:  (1)  Outgroup  com- 


PAXTON  ET  AL.:  MYCTOPHIDAE 


241 


Lampanyctini 


Diaphini 


Triphoturus 

Parvilux 

Lampanyctus 

Stenobrachius 

Lampadena 

Taaningichthys 

Bolinichthys 

Ceratoscopelus 

Lepidophanes 

Idiolychnus 

Lobianchia 

Diaphus 

Notoscopelus 

Lampichthys 

Scopelopsis 

Gymnoscopelus 

Hintonia 

Lampanyctodes 

Notolychnus 


Fig.  125.  Phylogenetic  diagram  of  the  Myctophidae,  subfamily  Lam- 
panyctinae.  Numbers  refer  to  the  apomorphic  characters  described  in 
Table  63.  Numbers  in  the  middle  of  vertical  lines  (e.g.,  4,  6)  refer  to 
characters  for  which  the  apomorphic  state  is  unknown.  Underlined 
numbers  refer  to  apomorphic  states  unique  to  all  members  of  a  given 
lineage;  bracketed  numbers  (e.g.,  59)  refer  to  apomorphic  states  that 
have  secondanly  reversed  in  at  least  one  member  of  the  lineage;  non- 
bracketed,  non-underlined  numbers  refer  to  character  states  found  in 
all  members  of  a  given  lineage  but  also  by  convergence  in  at  least  one 
other  taxon  in  the  family. 


parison.  All  previous  workers  have  considered  the  Myctophidae 
and  Neoscopelidae  as  sister  groups;  we  have  taken  the  character 
state  in  the  Neoscopelidae  to  be  the  plesiomorphic  condition 
for  the  Myctophidae.  Paxton  (1972:57)  described  the  parallel 
evolutionary  trends  in  the  neoscopelids  and  myctophids,  with 
SoliYonier  similar  to  the  Lampanyctinae  and  Neoscopelus  sim- 
ilar to  the  Myctophinae.  We  have  largely  limited  our  analysis 
to  those  characters  which  display  only  one  state  in  the  Neosco- 
pelidae. Where  variation  occurs  within  the  family,  the  character 
is  discussed  individually  below.  (2)  Linear  photophores.  We 
have  considered  a  photophore  elevated  out  of  linear  series  to 
be  apomorphic.  One  line  of  support  for  this  decision  occurs  in 
the  ontogeny  of  those  myctophid  species  with  a  larval  PLO 
photophore,  which  develops  on  the  pectoral  base  (where  it  pre- 
sumably has  a  different  function  from  that  of  the  adult)  and 
moves  dorsally  during  development  (Ahlstrom  et  al.,  1 976:Fig. 
4).  Also  the  photophores  of  Neoscopelus.  the  only  luminous 
neoscopelid  genus,  are  largely  linear.  However  there  is  some 
question  of  the  homology  of  Neoscopelus  and  myctophid  pho- 
tophores. O'Day  (1972:71)  described  the  ultrastructure  of  myc- 
tophid photophores  and  ".  .  .  confirm(s)  Brauer's  ( 1 908)  original 
recognition  of  the  close  resemblance  of  photogenic  tissue  in  the 
Neoscopelidae  to  that  found  in  the  Myctophidae."  However 
Herring  and  Morin  (1978:318)  considered  photophores  of  Neo- 
scopelus and  the  myctophids  to  be  very  different,  on  the  basis 


Myctophini 


Gonichthyini 


41,42,43 


Notolychnus 

Krefftichthys 
Protomyctophum 
Electrona 
Metelectrona 

Symbolophorus 

Myctophum 
Benthosema 
s^  Diogenichthys 
Hygophum 

Loweina 

Tarletonbeania 

Gonichthys 

Centrobranchus 

48,49 

Fig.  126.     PhylogeneticdiagramoftheMyctophidae,  subfamily  Myc- 
tophinae. Numbers  are  defined  as  in  Fig.  125. 


of  Kuwabara's  (1954)  description  of  Neoscopelus  compared  to 
that  of  Brauer  (1908).  As  ventral  photophores  have  evolved 
independently  at  least  one  other  time  in  the  stomiiform  fishes 
(Fink  and  Weitzman  1982:71),  the  potential  for  such  evolution 
in  deeper  water  fishes  is  high  enough  that  one  cannot  consider 
their  mere  existence  a  case  for  homology.  A  study  of  the  ultra- 
structure  of  Neoscopelus  photophores  would  be  of  value.  (3) 
Generalized  larvae.  The  larvae  of  neoscopelids  are  highly  spe- 
cialized with  a  robust  body,  a  large  head  and  jaws  with  prom- 
inent teeth,  a  long  gut  that  may  be  coiled  and  large  pectoral  fins. 
We  do  not  think  these  features  were  present  in  the  ancestors  of 
the  two  families,  and  where  they  are  present  in  the  myctophids, 
consider  they  have  evolved  independently.  We  have  used  only 
one  such  feature,  large  pectoral  fins  (40,  Table  63)  in  our  anal- 
ysis. We  consider  the  generalized  larva  of  the  myctophid  ances- 
tor had  the  following  characters,  based  on  the  distribution  of 
larval  features  in  myctophids  and  other  teleosts:  body  moder- 
ately slender,  gut  slightly  S-shaped,  extending  to  about  midbody, 
head  moderate  in  size,  eyes  round  or  nearly  so,  without  stalks 
or  choroid  tissue,  small  or  moderate  finfold  and  fins  and  Br, 
the  only  larval  photophores  present. 

We  have  used  a  total  of  59  characters,  far  fewer  than  the  total 
described  in  the  previous  studies.  For  many  we  were  unable  to 
determine  a  derived  state,  as  they  displayed  two  or  more  states 
or  were  absent  in  the  neoscopelids.  In  the  osteological  descrip- 
tions small  shape  differences  or  classifications  of  a  continuum 
were  often  found  in  both  families  and  were  not  included.  A 
number  of  the  characters  utilized  require  comment  or  expla- 
nation: (I)  Jaws  are  long  in  Solhomer and  short  in  Neoscopelus, 
and  following  our  ground  rules  should  not  be  utilized.  However, 
they  appear  to  be  of  such  fundamental  importance,  affecting 
many  correlated  characters  and  appearing  to  represent  a  major 
subfamilial  difference  (Paxton,  1972),  that  they  are  included 
here.  Paxton  (1972:58)  considered  short  jaws  to  be  primitive, 
primarily  because  they  occurred  in  Protomyctophum,  thought 


242 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  64.    Character  States  in  the  Genera  of  Myctophidae.  The  59  characters  are  described  in  Table  1.0  =  plesiomorphic  state,  1  = 


apomoi 

-phic 

state,  2  - 

=  dif 

reren 

t  or 

adva 

need 

apo 

mori 

)hic  state 

9  = 

unk 

now: 

1  or 

30th 

slates. 

1 

2 

3 

4 

6 

7 

8 

9 

10 

11 

12 

13 

14 

15 

16 

17 

18 

19 

20 

21 

n 

23 

24 

25 

26 

27 

Krefftichthys 

2 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

0 

Protomyctophum 

2 

0 

0 

9 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

0 

0 

1 

0 

Electrona 

2 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

0 

0 

0 

0 

0 

0 

0 

0 

1 

0 

0 

1 

0 

Metelectrona 

2 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

0 

0 

0 

0 

0 

1 

0 

0 

0 

0 

0 

1 

0 

Benlhosema 

2 

0 

0 

9 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

0 

0 

9 

0 

0 

1 

1 

Diogenichthys 

2 

0 

0 

0 

1 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

0 

0 

0 

0 

0 

1 

1 

Hygophum 

2 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

0 

Myctophum 

2 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

0 

0 

0 

0 

0 

0 

0 

0 

9 

0 

0 

1 

9 

Symbolophorus 

2 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

0 

0 

0 

9 

0 

0 

0 

0 

0 

0 

0 

1 

0 

Loweina 

1 

2 

0 

0 

0 

0 

0 

0 

0 

0 

1 

0 

0 

0 

1 

0 

0 

0 

0 

2 

0 

0 

1 

0 

Tarlelonbeania 

1 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

0 

0 

0 

0 

0 

0 

0 

0 

2 

0 

0 

1 

0 

Gonichlhys 

1 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

0 

Centrobranchus 

: 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

0 

Nololychnus 

0 

9 

0 

9 

0 

0 

0 

9 

0 

0 

0 

1 

0 

0 

0 

0 

0 

1 

0 

0 

1 

0 

0 

1 

0 

Lobianchia 

0 

0 

9 

0 

0 

0 

1 

1 

0 

1 

1 

0 

0 

0 

0 

0 

1 

0 

0 

0 

1 

0 

1 

1 

Diaphus 

9 

0 

0 

0 

0 

0 

1 

1 

0 

1 

1 

1 

9 

0 

9 

0 

1 

0 

0 

0 

0 

0 

0 

1 

Idiolychnus 

0 

0 

9 

0 

0 

0 

1 

0 

0 

1 

1 

0 

1 

0 

0 

0 

1 

0 

0 

9 

9 

0 

1 

9 

Lampanyctodes 

0 

0 

0 

0 

0 

0 

0 

1 

1 

1 

0 

0 

0 

0 

0 

0 

1 

0 

0 

0 

0 

0 

0 

1 

Gym  noscopelus 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

2 

1 

1 

0 

0 

1 

0 

0 

0 

0 

0 

0 

0 

1 

Scopelopsis 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

2 

0 

0 

1 

0 

0 

9 

0 

1 

0 

0 

1 

1 

1 

Lampichlhys 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

2 

0 

0 

1 

0 

0 

0 

0 

1 

0 

0 

0 

0 

1 

Notoscopeliis 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

2 

0 

0 

1 

1 

0 

0 

1 

0 

1 

1 

1 

1 

1 

Hinloma 

0 

0 

9 

0 

0 

9 

0 

0 

0 

0 

2 

1 

0 

9 

1 

0 

1 

0 

1 

9 

9 

1 

0 

9 

Lampadena 

0 

0 

0 

0 

9 

0 

0 

0 

2 

0 

0 

0 

1 

0 

0 

1 

0 

0 

0 

0 

0 

1 

Taaningichthys 

1 

0 

0 

0 

0 

0 

0 

0 

2 

0 

0 

0 

0 

0 

0 

1 

0 

0 

0 

0 

0 

0 

Ceratoscopelus 

1 

0 

0 

0 

0 

0 

0 

0 

2 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

Lepidophanes 

0 

0 

0 

0 

0 

0 

0 

2 

0 

0 

0 

0 

0 

1 

0 

0 

0 

0 

0 

0 

1 

Bolinichlhys 

2 

0 

0 

0 

0 

0 

0 

2 

0 

0 

0 

0 

0 

1 

1 

9 

0 

0 

0 

0 

1 

Tripholurus 

0 

0 

1 

0 

0 

0 

0 

2 

0 

0 

0 

0 

0 

1 

0 

0 

0 

0 

0 

0 

0 

Stenobrachius 

0 

0 

0 

0 

0 

0 

0 

2 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

0 

0 

0 

Parvilux 

0 

0 

0 

0 

0 

0 

0 

2 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

Lampanyctus 

0 

0 

0 

0 

0 

0 

0 

2 

0 

0 

0 

0 

0 

9 

0 

9 

0 

1 

0 

0 

0 

Solivomer 

0 

0 

9 

9 

0 

0 

0 

9 

0 

9 

9 

9 

1 

9 

9 

0 

0 

9 

0 

0 

9 

9 

0 

0 

9 

Neoscopelus 

-> 

0 

9 

0 

0 

0 

0 

9 

0 

9 

9 

9 

1 

9 

9 

0 

0 

9 

0 

0 

0 

0 

0 

0 

0 

Scopeleng^'s 

0 

0 

9 

0 

0 

0 

0 

9 

0 

9 

9 

9 

1 

9 

9 

0 

0 

9 

0 

0 

0 

0 

0 

0 

0 

to  represent  the  most  primitive  myctophid  based  on  photophore 
pattern.  However  Myers  (1958)  has  shown  that  short  jaws  have 
arisen  from  the  long-jawed  condition  a  number  of  times  in 
teleost  ^  olution,  and  discussed  their  adaptive  advantages.  We 
consider  short  jaws  to  be  the  apomorphic  condition  within  both 
the  Myctophidae  and  Neoscopelidae,  and  moderate  jaws  also 
to  be  derived  from  long  jaws.  (2)  Extrascapulars  are  single  in 
neoscopelids;  therefore  two  extrascapulars  in  some  myctophids 
should  be  the  derived  condition.  However  Paxton  (1972:58) 
described  how  the  neoscopelid  extrascapular  differs  in  position 
and  shape  from  that  of  myctophids.  Following  Williston's  Rule 
we  consider  a  single  extrascapular  to  be  derived  from  the  fusion 
of  two  elements,  independently  attained  in  each  family.  In  Low- 
eina the  single  condition  has  arisen  through  the  loss  of  the  dorsal 
extrascapular.  (4)  With  no  outgroup  with  similar  photophores 
for  comparison,  we  are  unable  to  determine  whether  1-2  or  3- 
9  Prcs  is  the  apomorphic  state.  However  the  two  character  states 
follow  subfamilial  limits,  and  one  of  the  states  must  be  derived 
and  definitive  for  its  subfamily.  (6)  All  myctophids  have  at  least 
one  of  the  orbital  light  organs,  Dn  and  Vn,  and  most  have  both. 
We  are  not  sure  whether  the  presence  or  the  absence  of  a  Dn  is 
apomorphic,  but  one  of  those  stales  defines  a  major  line  within 
the  Lampanyctinae.  (9)  Although  the  Neoscopelidae  have  a  su- 
pramaxillary,  Paxton  (1972:62)  considered  the  supramaxillary 
of  some  Myctophidae  to  be  an  independently  derived  feature. 


due  to  a  difference  in  shape  and  its  required  loss  at  least  four 
times  within  the  family  if  considered  primitive.  However,  John- 
son (1974b:205,  1982:79)  has  shown  the  presence  of  supra- 
maxilla(e)  to  be  primitive  in  other  myctophiforms  (sensu  lato); 
the  absence  of  a  supramaxilla  in  myctophids  is  here  considered 
a  derived  state  through  loss.  (15)  Although  caudal  luminous 
organs  are  not  present  in  neoscopelids,  they  are  present  in  all 
but  three  myctophid  genera,  where  their  loss  is  here  considered 
derived.  No  other  characters  indicate  that  any  of  the  three  genera 
(Diaphus,  Gymnoscopelus,  Hintoma)  are  the  most  primitive  in 
the  family.  (17)  Two  or  three  horizontal  Pols  are  in  a  linear 
position  and  should  be  considered  the  plesiomorphic  condition. 
However  in  those  genera  with  horizontal  Pols  (Notoscopelus, 
Lampichlhys  and  Scopelopsis)  the  photophores  are  high,  close 
to  the  lateral  line.  We  consider  the  primitive  myctophid  state 
to  be  one  with  low  photophores  with  none  or  one  Pol  (character 
34).  We  therefore  consider  the  horizontal  position  of  Pols  to 
be  derived,  while  noting  the  state  in  Hintonia  is  intermediate 
between  angled  and  horizontal.  (19)  Although  Johnson  (1982: 
76)  considered  a  higher  number  of  vertebrae  (42-62)  plesio- 
morphic for  iniomous  fishes,  lower  numbers  of  vertebrae  in 
neoscopelids  and  almost  all  myctophids  indicate  the  higher 
number  in  Gymnoscopelus  is  a  secondary  specialization  in  these 
families.  (23)  The  larval  gut  of  some  neoscopelids  is  long  and 
coiled,  clearly  a  specialization  foreshadowing  the  condition  of 


PAXTON  ET  AL.:  MYCTOPHIDAE 

Table  64.    Extended. 


243 


28 

29 

30 

31 

32 

33 

34 

35 

36 

37   38   39   40 

41 

42 

43 

44 

45 

46 

47 

48 

49 

50 

51 

52 

53 

54 

55 

56 

57 

58 

59 

1 

0 

0 

0 

0 

0 

0 

0 

0 

0    0 

D   0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

I 

0 

0 

0 

0 

1 

0 

0 

0 

0 

0 

0 

0 

0 

0    0 

1   0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

0 

0 

0 

9 

0 

1 

0 

0 

0 

0 

0 

0 

0 

9 

0    1 

1   0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

9 

0 

0 

9 

0 

1 

0 

0 

0 

0 

0 

9 

0 

1 

0    1 

1   0 

0 

0 

1 

0 

0 

0 

0 

0 

0 

0 

1 

0 

0 

0 

0 

1 

1 

1 

0 

1 

0 

0 

0 

0 

0 

0 

9 

0    1 

1   0 

0 

0 

0 

2 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

9 

0 

0 

1 

0 

1 

0 

0 

0 

0 

0 

0 

0 

0    1 

1   0 

0 

0 

0 

2 

0 

0 

0 

0 

I 

0 

0 

1 

0 

0 

0 

0 

0 

1 

0 

1 

0 

0 

0 

0 

0 

2 

0 

9 

0    1 

0 

0 

0 

0 

2 

0 

0 

0 

0 

0 

0 

1 

0 

0 

0 

9 

9 

9 

1 

0 

1 

0 

0 

0 

0 

0 

0 

0 

0    1 

0 

0 

0 

9 

2 

9 

0 

0 

0 

9 

0 

0 

0 

0 

0 

1 

0 

0 

9 

0 

1 

0 

0 

0 

0 

0 

0 

1 

0   1 

0 

0 

0 

2 

0 

0 

0 

0 

0 

0 

0 

0 

1 

0 

1 

0 

0 

9 

0 

1 

0 

0 

0 

0 

0 

0 

0 

0   1 

1 

1 

1 

1 

0 

0 

1 

0 

1 

0 

0 

0 

0 

0 

0 

1 

0 

1 

0 

1 

0 

0 

0 

0 

0 

0 

0 

0   1 

1 

1 

1 

1 

0 

0 

1 

0 

1 

1 

0 

0 

0 

0 

1 

1 

0 

0 

0 

1 

0 

0 

0 

0 

0 

0 

0 

0   1 

1 

1 

1 

1 

1 

1 

0 

0 

1 

0 

0 

0 

0 

0 

1 

1 

0 

0 

0 

1 

0 

0 

0 

0 

0 

0 

0 

0   1 

1 

1 

1 

1 

1 

1 

0 

1 

2 

0 

0 

0 

0 

0 

1 

1 

0 

1 

0 

0 

0 

0 

0 

0 

0 

2 

0 

0 

0   0 

0 

0 

0 

0 

2 

0 

0 

0 

0 

2 

0 

0 

0 

0 

0 

0 

0 

0 

0 

9 

1 

0 

0 

0 

0 

0 

9 

0 

0 

0   1 

1 

0 

0 

0 

2 

0 

0 

0 

0 

9 

0 

0 

0 

0 

1 

9 

0 

0 

1 

1 

9 

9 

0 

0 

0 

0 

9 

0 

0 

0   1 

0 

0 

0 

0 

2 

0 

0 

0 

0 

9 

0 

0 

0 

0 

1 

0 

0 

0 

1 

9 

1 

0 

0 

0 

0 

0 

2 

0 

0 

9   1 

9 

0 

0 

9 

2 

0 

0 

9 

0 

9 

0 

9 

0 

9 

9 

9 

9 

9 

1 

9 

0 

1 

0 

0 

0 

0 

1 

0 

0 

0   1 

0 

0 

0 

0 

2 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

9 

1 

0 

0 

0 

0 

2 

9 

0 

0   1 

0 

0 

0 

0 

2 

0 

0 

0 

0 

9 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

1 

0 

0 

0 

0 

2 

1 

0 

0   1 

0 

0 

0 

0 

2 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

1 

0 

0 

0 

0 

2 

1 

0 

0   1 

0 

0 

0 

0 

2 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

1 

0 

0 

0 

0 

2 

1 

0 

0   1 

0 

0 

0 

0 

2 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

0 

0 

0 

0 

2 

0 

0 

9   1 

9 

0 

0 

9 

2 

0 

0 

9 

0 

0 

0 

9 

0 

9 

9 

9 

9 

9 

0 

0 

0 

9 

1 

0 

0 

0 

I 

0 

0 

0   1 

0 

0 

0 

0 

2 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

0 

0 

1 

1 

0 

0 

0   1 

0 

0 

0 

0 

2 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

1 

0 

1 

1 

0 

2 

0 

0 

0   1 

0 

0 

0 

0 

2 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

1 

0 

1 

0 

0 

2 

1 

1 

0   1 

0 

0 

0 

0 

2 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

0 

1 

1 

0 

0 

0 

1 

2 

0 

0 

1  1 

0 

0 

0 

0 

2 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

0 

0 

0 

0 

0 

0 

2 

0 

1 

0   1 

0 

0 

0 

0 

2 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

9 

0 

0   1 

0 

0 

0 

0 

2 

0 

0 

0 

0 

9 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

2 

1 

0 

1  1  1 

0 

0 

0 

0 

2 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

9 

0 

0 

0 

0 

2 

1 

1 

9   1   1 

9 

0 

0 

0 

2 

0 

0 

0 

0 

9 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

9 

0 

9 

9   9   5 

9 

0 

0 

9 

9 

0 

0 

9 

0 

0 

0 

9 

0 

9 

9 

9 

9 

9 

0 

0 

0 

0 

0 

0 

0 

0 

9 

0 

9 

0   9   9 

1 

0 

0 

0 

9 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

9 

0 

9 

0   9   9 

1 

0 

0 

0 

9 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

some  acanthopterygians.  Although  it  could  be  argued  that  the 
short  gut  that  lengthens  during  development  in  a  few  forms  of 
myctophids  represents  the  primitive  condition,  we  consider  the 
primitive  myctophid  condition  a  moderate— iengthed  gut,  with 
different  derived  states,  short  and  long.  (26)  Although  the  caudal 
luminous  organs  are  sexually  dimorphic  in  about  half  the  genera, 
we  assume  the  original  caudal  organs  were  not  sexually  dimor- 
phic. (27)  No  photophores  are  present  on  the  described  larvae 
of  Neoscopelus.  However  the  Br,  develops  in  all  larval  mycto- 
phids except  Taamngichthys  and  Notolychmis.  and  its  univer- 
sality indicates  it  was  present  in  the  ancestral  myctophid.  Other 
larval  photophores  however  are  present  in  fewer  than  half  of 
the  genera  and  we  consider  their  presence  derived.  (30)  The 
strongly  developed  caudal  luminous  organs  found  in  Lampa- 
dena  and  Taaningichthys  are  clearly  a  more  specialized  state 
than  the  relatively  unstructured  organs  found  in  many  other 
genera.  (34)  See  the  discussion  of  character  17.  (36)  Although  a 
strongly  angled  set  of  SAOs  represents  a  linear  position  for  the 
first  two  photophores,  we  consider  this  condition  developed  by 
the  SAO,  rising  from  a  lower  position  in  the  weakly  angled, 
plesiomorphic  position.  (59)  We  consider  the  absence  of  a  keel 
or  ridge  on  the  fifth  circumorbital  of  Hintonia  to  be  secondarily 
derived  through  loss.  This  is  the  only  character  state  we  have 
used  which  is  not  present  in  all  examined  members  of  the  line 
it  defines. 


We  have  thus  attempted  to  determine  polarity  for  25  osteo- 
logical,  17  larval  and  17  photophore  characters.  We  initially 
attempted  a  phylogenetic  analysis  utilizing  the  distribution  of 
23  larval  characters  at  the  species  level.  The  resulting  diagram 
split  some  genera  into  as  many  as  three  unrelated  lines.  We 
remain  convinced  that  the  myctophid  genera  as  currently  de- 
fined by  larval  morph,  photophore  pattern  and  osteology  rep- 
resent monophyletic  lines  (even  though  such  genera  as  Diaphus, 
Lampanyctus,  Myctophum  and  Hygophum  may  be  formally  di- 
vided as  subgenera  or  genera  by  future  work).  These  genera  we 
use  as  the  starting  point  in  the  present  study.  We  have  con- 
structed a  phylogenetic  tree  (Figs.  1 25,  1 26)  based  on  our  knowl- 
edge of  the  family  and  used  the  apomorphic  states  of  the  59 
characters  to  define  the  various  branching  points,  which  is  the 
basis  of  the  following  discussion. 

The  subfamily  Lampanyctinae  is  defined  by  two  apomorphies 
restricted  to  all  members  of  the  subfamily  (those  characters 
found  in  all  members  of  a  lineage  and  nowhere  else  in  the  family 
are  underlined  in  Figs.  125  and  126).  the  presence  of  a  cleithral 
shelf  and  a  single,  fused  extrascapular.  The  subfamily  Mycto- 
phinae  is  defined  by  two  apomorphies,  short  or  moderate  jaws 
and  narrow  larval  eyes,  but  these  features  are  also  found  in  a 
few  genera  of  the  Lampanyctinae.  The  number  of  Pre  photo- 
phores defines  all  members  of  one  of  the  subfamilies  (see  dis- 
cussion of  character  4  above). 


244 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Notolychmis  valdiviae.  here  considered  a  monotypic  tribe, 
could  not  be  placed  with  certainty  in  either  subfamily.  Moser 
and  Ahlslrom  ( 1 970: 1 38,  1 974:409)  and  Paxton  ( 1 972:6 1 )  dis- 
cussed the  characters  and  problems  of  this  enigmatic  species. 
With  long  jaws  and  the  lack  of  a  cleithral  shelf  both  considered 
plesiomorphies,  the  apomorphic  number  of  Pre  photophores 
unknown,  and  the  larval  eyes  variable  and  intermediate  in  shape, 
future  work  is  required  to  resolve  this  trichotomy. 

We  recognize  three  tribes  in  the  subfamily  Lampanyctmae 
(Fig.  125).  The  tribe  Lampanyctini,  with  nine  genera,  is  defined 
by  the  presence  of  a  row  of  moderately  to  strongly  hooked  teeth 
in  the  posterior  dentary;  the  only  other  genus  with  this  feature 
is  the  myctophine  Diogenichthys.  These  nine  genera  are  also  the 
only  lampanyctines  to  lack  a  Dn  orbital  photophore,  but  we  are 
unsure  if  this  is  a  derived  state  (see  discussion  of  character  6 
above).  Moser  and  Ahlstrom  (1972)  and  Ahlstrom  et  al.  (1976: 
148)  placed  five  of  these  genera  (Lampadena.  Taaningichthys. 
Bolinichthys.  Lepidophanes,  Ceratoscopelus)  in  the  tribe  Gym- 
noscopelini,  based  primarily  on  larval  photophore  pattern.  Pho- 
tophores which  appear  in  larvae  of  Lampanyctinae  are  essen- 
tially the  same  ones  which  develop  in  myctophine  larvae  (Moser 
et  al.,  this  volume)  and,  if  they  are  adaptive  as  Moser  (1981) 
has  suggested,  it  is  likely  that  they  have  appeared  in  these  typical 
sites  independently  in  a  number  of  lineages.  Moreover,  these 
photophores  develop  at  the  end  of  the  larval  period,  if  at  all,  in 
Bolinichthys  and  no  photophores  develop  in  Taaningichthys 
larvae.  Likewise,  the  larval  pigment  characters  do  not  support 
the  inclusion  of  these  five  genera  in  the  Gymnoscopelini. 

In  addition  to  the  distribution  of  hooked  dentary  teeth  and 
Dn  photophores,  other  features  influenced  our  decision  about 
these  five  genera.  The  ischial  ligament  is  medium  or  long  in  all 
Lampanyctini  except  Taaningichthys  (and  some  species  of  Dia- 
phus).  while  the  fifth  circumorbital  has  a  ridge  or  keel  in  all 
gymnoscopelines  (but  is  lacking  in  some  species  of  Diaphus)  and 
no  lampanyctines  except  Bolinichthys  (thus  the  brackets  around 
character  59  in  Fig.  125).  Finally  all  of  the  gymnoscopeline 
genera  except  Notoscopelus  are  restricted  to  the  southern  ocean 
(Moser  et  al.,  this  volume:  Table  59),  while  the  Lampanyctini 
are  found  both  north  and  south  (except  Stenobrachim)  of  the 
equator.  Placement  of  the  five  genera  in  the  Lampanyctini  re- 
quires fewer  character  reversals  and  parallelisms. 

Within  the  Lampanyctini,  the  development  of  larval  photo- 
phores in  addition  to  Br,  (character  27)  unites  the  five  genera 
discussed  above.  We  recognize  Dorsadena  as  a  subgenus  of 
Lampadena  until  specimens  other  than  the  types  are  available 
for  osteological  study  and  the  larvae  are  discovered.  We  have 
not  found  an  apomorphic  character  that  defines  the  line  in- 
cluding Stenobrachius.  Triphoturus.  Lampanyctus  and  Parvilux. 
We  are  recognizing  Parvilux  on  the  basis  of  a  weakly  angled 
SAO  and  larval  shape  and  pigmentation. 

We  consider  the  tribe  Diaphini  to  be  the  sister  group  of  the 
Gymnoscopelini.  The  relationships  among  the  three  genera  of 
Diaphini  are  not  clear.  One  of  us  (HGM)  has  re-examined  the 
specimens  on  which  the  larval  features  of  Idiolychmis  urolampus 
were  based  (see  Moser  and  Ahlstrom,  1974:405-406;  Nafpak- 
titis  and  Paxton,  1978),  and  now  thinks  they  could  represent 
Lobianchia  gemellari.  with  the  larvae  of  Idiolychnus  still  un- 


known. Two  characters  shared  by  Lobianchia  and  Idiolychnus. 
the  presence  of  caudal  organs  and  the  absence  of  a  luminous 
patch  above  the  pectoral  fin,  are  considered  plesiomorphic,  while 
the  absence  of  a  Vn  and  differences  of  photophore  positions  are 
not  clearly  apomorphic.  The  most  unequivocal  derived  state  is 
the  presence  of  a  wide  pubic  plate,  indicating  Lobianchia  and 
Diaphus  are  the  sister  group  pair. 

Within  the  Gymnoscopelini  the  proposed  generic  relation- 
ships are  based  almost  entirely  on  characters  of  the  photophores 
and  luminous  tissue.  No  consistent  osteological  or  larval  fea- 
tures define  generic  groupings.  Southern  ocean  larvae  require 
more  study.  The  larvae  of  Hintonia  are  unknown  and  not  enough 
species  of  Gymnoscopelus  have  been  studied  to  ascertain  if  the 
subgenus  Nasolychnus  can  be  defined  by  any  larval  characters. 
The  species  of  Notoscopelus  should  also  be  studied  to  find  sup- 
porting characters  of  the  subgenus  Parieophus. 

Within  the  subfamily  Myctophinae  (Fig.  126),  we  also  rec- 
ognize three  tribes,  the  Electronini,  Myctophini  and  Gonichthy- 
ini.  The  Gonichthyini  is  clearly  a  derived  lineage,  with  a  num- 
ber of  osteological,  photophore  and  larval  characters 
distinguishing  the  four  genera  from  the  rest  of  the  subfamily. 
We  think  the  larval  specializations  of  eyes  and  pectoral  fins  arose 
after  the  split  of  the  two  generic  pairs. 

Paxton  (1972)  was  unable  to  find  osteological  characters  to 
clearly  separate  the  remaining  genera  of  the  Myctophinae  into 
two  lineages.  We  have  utilized  photophores  to  distinguish  the 
Myctophini  from  the  Electronini,  while  recognizing  there  is  a 
mosaic  of  osteological  and  larval  characters  within  these  nine 
genera.  We  have  little  question  of  the  sister  group  relationship 
of  the  generic  pairs  Krefftichthys—Protomyctophum.  Mycto- 
phum  —  Symbolophorus  and  Benthosema  —  Diogenichthys. 
However  two  larval  features,  thin  head  and  body  and  a  bipartite 
gut,  are  shared  by  Metelectrona  and  some  species  of  Hygophum. 
Since  we  think  Hygophum  is  a  monophyletic  line,  we  consider 
these  shared  larval  features  parallelisms  that  do  not  indicate 
common  ancestry.  Paxton  (1972)  considered  Metelectrona  a 
synonym  of  Electrona.  The  description  of  a  second  species  of 
Metelectrona  (Hulley,  1981),  coupled  with  its  larval  and  pho- 
tophore characters,  convinced  us  to  recognize  the  genus. 

Of  the  59  derived  characters  utilized  in  our  analysis,  only  20 
are  restricted  to  members  of  the  lineage  they  define,  and  eight 
of  these  are  autapomorphic  at  the  generic  level.  The  remaining 
39  characters  are  not  found  in  the  apomorphic  state  in  any 
member  of  the  opposite  lineage  from  the  defined  branching 
point,  but  are  found  in  some  members  of  other  lineages  within 
the  family.  This  presumed  homoplasy  of  larval,  photophore  and 
even  osteological  characters  indicates  that  the  proposed  phy- 
logeny  was  arrived  at  with  some  difliculty.  Ten  of  our  proposed 
lineages  are  undefined  by  derived  characters.  We  think  that 
future  work  will  support  our  proposed  phylogeny,  although  some 
details  may  be  modified,  and  that  new,  less  plastic  characters 
and  better  definitions  of  polarity  will  help  resolve  the  problems. 

(J.  R.  P.)  The  Al'stralian  Museum,  6-8  College  Street, 
Sydney  2000,  Australia;  (H.G.M.)  National  Marine 
Fisheries  Service,  Southwest  Fisheries  Center,  P.O.  Box 
271,  La  Jolla,  California  92038. 


Scopelarchidae:  Development  and  Relationships 
R.  K.  Johnson 


THE  Scopelarchidae  has  traditionally  been  included  with  the 
primarily  oceanic  Alepisauroidei  (Marshall,  1955;  Gosline 
et  al.,  1966;  Rosen,  1973;  Johnson,  1974b,  the  most  recent 
complete  revision).  Johnson  (1982)  excludes  the  scopelarchids 
from  the  alepisauroids,  rejects  putative  sister-group  relationship 
with  the  Evermannellidae,  and  provisionally  allies  the  scope- 
larchids with  the  chlorophthalmoids.  All  scopelarchids  are 
oceanic  and  meso-  or  bathypelagic.  The  majority  of  known  adult 
specimens  were  taken  in  hauls  to  depths  between  500  and  1 ,000 
m.  For  most  species  there  exists  no  evidence  to  suggest  diel 
migration,  however,  Merrett  et  al.  ( 1 973:39-40)  present  limited 
evidence  for  diel  migration  ("considerably  dispersed  vertically") 
in  Benthalhella  infans.  Scopelarchids  are  relativedly  large-bod- 
ied (to  302  mm  SL;  Iwami  and  Abe,  1980).  All  Scopelarchidae 
are  tubular-eyed  predators  (see  Munk,  1966;  Locket,  1970; 
Muntz,  1976;  Johnson,  1982)  concentrating  most  frequently  on 
fish,  not  capable  of  engorgement  of  enormously  large  food  par- 
ticles (unlike  evermannellids,  Omosudis,  Alepisaurus,  Antop- 
terus  and  at  least  some  paralepidids).  Luminous  tissue  occurs 
in  Benthalhella  infans  (Merren  et  al.,  1973)  and  probably  occurs 
in  Scopelarchoides  kreffti  (Johnson,  1 974b).  The  family  contains 
1 7  species  arranged  in  four  genera  and  occurs  throughout  the 
world  ocean  except  that  no  scopelarchid  inhabits  the  Arctic 
Ocean  or  the  Mediterranean  Sea.  Among  iniomous  fishes,  the 
Scopelarchidae  is  distmguished  by  the  following  combmation 
of  characters:  ( 1 )  basihyal  short  to  elongate  but  well-ossified;  (2) 
lingual  teeth  strong,  straight  to  strongly  hooked,  invariably  pres- 
ent over  basihyal,  present  or  absent  over  basibranchials;  (3)  body 
and  postorbital  regions  of  head  completely  covered  with  cycloid 
scales;  (4)  lateral  line  scales  large,  differing  distinctively  in  exact 
conformation  between  all  species  (Johnson,  1974b:  Fig.  2);  (5) 
parietal  bones,  when  present,  small,  widely  separated  by  frontals 
and  supraoccipital;  (6)  coracoid  broadly  expanded;  (7)  two  post- 
cleithra,  widely  separated  in  vertical  dimension;  (8)  unossified 
gap  (filled  by  tube-like  structure  of  fibrous  connective  tissue) 
between  skull  and  first  vertebral  centrum  (see  Merrett  et  al., 
1973:17);  (9)  posttemporal  unforked;  (10)  no  basisphenoid,  or- 
bitosphenoid,  gill  rakers,  or  free  second  ural  centrum;  (11)  eyes 
tubular,  directed  straight  upward  (except  in  3  species  where 
directed  dorsoanteriad);  (12)  larvae  with  0,  1  or  3  peritoneal 
pigment  sections.  The  genera  and  species  are  distinguished  by 
gross  morphological,  meristic,  morphometric,  osteological,  pig- 
ment and  larval  characters  (Tables  65  and  66). 

Development 

Eggs  of  scopelarchids  are  unknown.  Larvae  are  known  for  all 
species  except  Scopelarchoides  kreffti  and  developmental  series 
have  been  illustrated  and  described  (Rosen,  1973;  Merrett  et 
al.,  1973;  Johnson,  1974b;  Belyanina,  1981,  1982a;Moser,  1981). 
Except  for  limited  information  on  Benthalhella  infans  in  Merrett 
et  al.  ( 1 973),  osteological  description  has  been  confined  to  adults. 
Except  in  Benthalhella.  development  is  direct,  adult  characters 
are  essentially  acquired  one  by  one,  with  completion  of  trans- 
formation at  30  to  more  than  80  mm  SL  depending  upon  the 


species.  Larvae  of  Benthalhella  undergo  very  rapid  (i.e.,  small 
size  increment)  transformation  after  a  prolonged  period  of  growth 
while  retaining  larval  form  (see  below).  Larvae  of  most  species 
are  known  from  hauls  within  the  top  100  m  and  the  larvae  of 
a  number  of  species  have  been  taken  in  the  top  50  m.  Con- 
trariwise the  larvae  of  one  species,  Benthalhella  dentata.  have 
not  been  taken  in  hauls  shallower  than  150  m  and  most  were 
taken  in  hauls  to  depths  in  excess  of  500  m.  Except  possibly  the 
cases  oi Benthalhella  elongata  and  B.  macropmna  (see  Johnson, 
1974b:228),  the  distributional  ranges  of  larvae  and  adults  are 
coextensive.  There  is  no  evidence  (the  data  are  quite  incomplete) 
for  seasonality  in  reproductive  effort.  Scopelarchids  are  syn- 
chronous hermaphrodites. 

The  following  paragraphs  describe  those  characters  most  ev- 
ident in  the  early  life  history  of  scopelarchids,  including  those 
of  value  in  distinguishing  genera  and  species. 

Gross  aspect  (Fig.  127).  — Larvae  range  from  extremely  elongate 
and  shallow  (Benthalhella)  to  quite  short  and  deep  (some  species 
of  Scopelarchus  and  Scopelarchoides).  Small  larvae  are  trans- 
lucent, scaleless,  colorless  (except  for  pentoneal  pigment  sec- 
tions, when  present),  with  a  characteristic  "bowed  down"  an- 
terior dorsal  profile.  The  body  is  deepest  at  the  pectoral  girdle 
and  the  trunk  elongate.  Anteriorly  the  hypaxial  muscles  do  not 
embrace  the  abdominal  cavity  walls  which  are  therefore  highly 
translucent.  Only  the  muscles  of  the  pelvic  girdle  are  visibly 
evident.  The  abdominal  cavity  is  triangular,  deep  anteriorly. 
Peritoneal  pigment  appears  early  except  in  Benthalhella  which 
lacks  peritoneal  pigment  until  transformation.  The  gut  is  mid- 
ventral.  In  larvae  the  anus  is  anterior  (relative  to  distance  be- 
tween pelvic  fin  insertion  and  anal  fin  origin)  to  position  in 
adults,  far  anterior  in  some  (Benthalhella).  The  head  is  very 


Table  65.    Com 

PARISON 

OF  Selected  Meristic 

Characters  among 

Scopelarchid 

Species. 

Lateral 

Dorsal 

Anal 

Pectoral 

line  scales 

Vertebrae 

alalus 

8-9 

20-22 

23-26 

47-49 

46-47 

hubbsi 

8-9 

23-25 

21-23 

53 

49 

votucris 

9-10 

21-24 

23-26 

48-51 

49-51 

stephensi 

8 

20-22 

18-20 

41-44 

42-43 

michaelsarsi 

7-9 

18-21 

18-21 

40-44 

40-44 

anatis 

7-9 

21-26 

18-22 

45-50 

44-49 

guentheri 

7-8 

24-29 

18-21 

47-52 

47-51 

danae 

6-9 

24-27 

20-22 

50-52 

48-50 

nicholsi 

6-7 

20-23 

20-23 

46-50 

45-48 

kreffti 

9 

25-27 

23-25 

58-59 

55-57 

climax 

7-8 

25-27 

25 

53 

49 

signifer 

9-10 

26-29 

22-25 

49-52 

48-49 

macropmna 

5-6 

35-39 

25-27 

62-65 

60-62 

dentata 

6-8 

17-20 

21-24 

54-58 

54-55 

elongata 

9-10 

24-28 

19-23 

61-65 

62-65 

infans 

8-9 

20-26 

25-28 

55-59 

55-58 

linguidens 

8-9 

28-30 

24-25 

66 

64 

245 


246 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


CAV 


Fig.  127.  Larvae,  juveniles  and  adult  of  Scopelarchidae.  (A.  B)  Rosenblattichthys  volucris.  A  =  14.5  mm  SL,  B  =  26.0  mm  SL,  letters  refer  to 
pigment  spots;  (C,  D)  Scopelarchoides  nicholsi,  C  =  1.5  mm  SL,  D  =  23.0  mm  SL,  letters  refer  to  larval  pigment  spots;  (E,  F)  Benthalhella  denlata. 
E  =  larva,  42.8  mm  SL,  F  =  transforming  specimen,  53.0  mm  SL,  arrows  indicate  position  of  anus;  (G)  Scopelarchus  guenlhen.  juvenile,  48.5 
mm  SL,  DS  =  dermal  pigment  stripes;  (H)  Scopelarchus  analis.  adult,  1 12.5  mm  SL. 


large  and  massive,  exceeding  30%  of  the  SL  in  Rosenblattichthys, 
and  large  but  not  as  large  in  other  genera.  The  eye  is  elliptically 
narrowed,  and  initially  small  in  comparison  with  the  size  of  the 
bony  orbit.  The  interorbital  is  initially  broad  and  narrows  during 
transformation.  Development  of  the  eyes  is  described  for  Ben- 
thalhella infans  in  Merrett  et  al.  (1973).  The  snout  is  pointed. 
The  mouth  is  large  and  low,  with  teeth  appearing  in  very  small 
larvae.  The  most  striking  changes  take  place  during  a  period  of 
transformation,  which,  as  described  below,  can  either  be  within 
a  very  short  interval  (ca.  10  mm  in  Benthalhella  dentata)  of 
growth  (any  statements  implying  time  sequence  are  based  solely 
on  increments  of  length)  as  in  Benthalhella,  or  over  a  long  (20 
mm)  to  very  long  (50  mm)  interval. 

Meristic  characters.— Counts  of  fin  rays  (Table  65)  do  not  differ 
between  larval  and  adult  specimens.  Most  scopelarchid  species 
can  be  uniquely  distinguished  from  all  other  species  on  the  basis 
ofmeristic  characters  alone  (Johnson,  1974b:  14).  Rosenhlattich- 
ihys  is  unique  in  precocious  ossification  of  the  pectoral  fin  rays, 
well  in  advance  of  the  pelvic  or  median  fins  (except  caudal).  In 
all  other  scopelarchids  the  lowermost  5  or  6  pectoral  fin  rays 
are  the  last  to  be  formed  and  the  order  of  fin  ray  ossification  is 
caudal  >  dorsal,  anal,  dorsal  pectoral  >  pelvic  >  ventral  pec- 
toral. As  in  all  inioms  the  caudal  is  formed  of  10  +  9  principle 
rays.  In  Scopelarchoides  and  Rosenhlattichthys  the  pelvic  fins 
appear  as  buds  on  the  midlateral  abdominal  cavity  wall,  well 
above  the  level  of  the  intestine.  In  Benthalhella  and  Scopelar- 


chus the  pelvic  fin  buds  appear  ventrolaterally,  at  or  beneath 
the  level  of  the  intestine.  In  Benthalhella  (except  B.  macropinna) 
the  pelvic  fin  insertion  in  larvae  is  distinctly  in  advance  of  the 
dorsal  fin  origin.  In  other  scopelarchid  larvae  the  pelvic  fin 
insertion  is  beneath  or  behind  the  dorsal  fin  base  (but  comes  to 
be  slightly  in  advance  of  dorsal  fin  origin  in  adult  Rosenhlattich- 
thys and  distinctly  in  advance  of  dorsal  fin  origin  in  all  adult 
Benthalhella).  The  adipose  fin  develops  within  the  dorsal  finfold 
which  extends  between  the  dorsal  and  caudal  fin  in  small  larvae. 
In  adults  the  adipose  fin  is  inserted  over  the  posterior  one-third 
of  the  anal  fin  base  (except  B.  dentata  where  inserted  posterior 
to  a  vertical  through  base  of  last  anal-fin  ray).  Ventral  finfold 
extending  from  vent  to  anal-fin  origin  in  smaller  larvae,  and  is 
completely  reabsorbed  in  early  transformation. 

Peritoneal  pigment  sections.  — \n  all  adult  scopelarchids  (except 
B.  elongata)  the  gut  is  enclosed  by  a  uniform  tube  of  brown  to 
jet-black  pigment.  In  larvae  this  pigment  appears  in  discrete 
sections  (except  in  Benthalhella  where  peritoneal  pigment  is 
lacking  prior  to  transformation)  and  in  a  conformation  char- 
acteristic for  each  genus  or  group  of  apparently  related  species. 
All  larvae  larger  than  20  to  22  mm  possess  peritoneal  pigment 
(except  in  Benthalhella).  One  section  only,  unpaired,  forming  a 
saddle-like  canopy  over  the  gut,  is  present  in  Rosenhlattichthys, 
Scopelarchoides  signifer,  and  S.  clima.x  (larvae  of  S.  kreffii  are 
unknown).  Three  sections,  a  single  anterior  section  as  above  and 
two  paired  posterior  sections  are  found  in  Scopelarchoides  nich- 


JOHNSON:  SCOPELARCHIDAE 


247 


olsi.  S.  danae.  and  Scopelarchus.  However  in  S.  nicholsi  and  5. 
danae  the  posterior  sections  appear  significantly  "later"  and 
appear  above  (S.  danae)  or  anterior  (S.  nicholsi)  to  the  pelvic 
fin  bases.  In  Scopelarchus  a.\\  3  sections  appear  in  near  synchrony 
and  the  posterior  sections  appear  well  to  the  rear  of  the  pelvic 
fin  bases.  In  all  cases  the  pigment  section(s)  expand  during  trans- 
formation and  for  all  genera  except  Benihalhella  the  completion 
of  transformation  can  be  defined  as  acquisition  of  the  adult  state 
of  a  complete  and  unbroken  tube  of  peritoneal  pigmentation. 
In  Benthalbella  the  first  appearance  of  peritoneal  pigment  (not 
in  discrete  section  but  uniformly  in  mesentary  dorsal  to  gut  from 
between  pectoral  fin  bases  to  behind  pelvic  fin  bases)  signals  the 
onset  of  the  period  of  "rapid"  transformation. 

Other  larval  pigment.— Jht  larvae  of  Scopelarchoides  and  Ro- 
senblattichthys  are  characterized  by  the  presence  of  well-defined 
pigment  spots  or  areas  (accessory  pigment  of  Johnson.  1974b; 
complementary  pigment  of  Belyanina,  1982a)  apparent  in  the 
smallest  (6- 1 2  mm  SL)  known  larvae.  The  presence  and  location 
of  spots  is  uniquely  diagnostic  for  each  species  possessing  them. 
Pigment  spots  are  present  in  all  larvae  of  Scopelarchoides  and 
Rosenblattichthys.  absent  in  Benthalbella  and  Scopelarchus.  In 
Scopelarchoides  the  middorsal  spot,  if  present,  and  the  mid- 
ventral  spot  are  entirely  behind  the  adipose  base  and  anal  fin 
base  respectively.  In  Rosenblattichthys  the  middorsal  and  mid- 
ventral  (where  present)  spots  are  entirely  in  advance  of  the  bases 
of  these  fins. 

Transformation  pigmentation.— Johnson  (1974b:20)  distin- 
guishes "dermal"  vs  "epidermal"  pigmentation  in  scopelar- 
chids.  Dermal  pigmentation  refers  to  the  major  pigment  stripes 
present  in  some  genera  and  species.  These  develop  "early"  dur- 
ing transformation  and  persist  in  the  adult.  In  most  cases  the 
dermal  pigment  comes  to  be  partially  or  completely  overlain 
by  the  epidermal  pigmentation  associated  primarily  with  the 
scale  pockets.  Dermal  pigment  is  present  in  all  4  species  of 
Scopelarchus  and  in  certain  Scopelarchoides  and  Rosenblatt- 
ichthys, it  is  absent  in  Benthalbella.  The  subequal  pigment  stripes 
oC  Scopelarchus  (Fig.  127),  situated  above  and  below  the  lateral 
line,  are  diagnostic  for  the  genus. 

Gut  morphology.— \n  all  scopelarchids  the  stomach  is  a  heavily 
muscularized,  greatly  elongate  blind  pouch.  In  small  larvae  the 
stomach  does  not  reach  the  pelvic  fin  base,  but  it  expands  pos- 
teriad  during  transformation,  very  "rapidly"  so  in  Benthalbella. 
and  in  all  adults  extends  to  or  nearly  to  a  vertical  through  the 
anus  (which  in  all  is  closely-adjacent  to  the  anal  fin  origin). 
Johnson  (1974b)  and  Wassersug  and  Johnson  (1976)  note  that 
the  tremendous  expansion  of  the  stomach  allows  ingestion  of 
fairly  large  particles  and  hypothesize  that  the  blind  pouch  ar- 
rangement is  a  device  for  maximal  recovery  of  food  energy. 

Transformation.  —  Larvae  of  Benthalbella  undergo  rapid  trans- 
formation after  a  prolonged  period  of  growth  while  retaining 
larval  form.  The  onset  of  transformation  (size  of  smallest  known 
transforming  specimen  =  49.6  mm  SL  in  B.  dentata;  89. 1  mm 
SL  in  B.  elongata:  55. 1  mm  SL  in  B.  infans;  65. 1  mm  SL  in  B. 
macropinna;  no  transforming  specimens  of  B.  lingutdens  are 
known,  but  the  largest  known  larva  is  85.5  mm  SL)  is  signalized 
by  appearance  of  a  lens  pad,  appearance  of  peritoneal  pigment, 
and  invasion  of  the  abdominal  body  wall  by  musculature.  Other 
changes  occurring  during  transformation  include  rapid  elonga- 


tion of  gut  and  stomach,  "migration"  of  anus  from  just  behind 
pelvic  fin  base  to  just  anterior  to  anal  fin  origin,  appearance  of 
gonad,  appearance  of  scales  (especially  lateral  line  scales),  ap- 
pearance of  head  and  body  pigmentation,  reabsorption  of  ven- 
tral adipose  fin.  great  restriction  of  base  of  dorsal  adipose  fin, 
ossification  of  vertebral  column,  change  (from  dorsally  convex 
to  dorsally  concave)  in  curvature  in  vertical  plane  of  anterior 
portion  of  vertebral  column  (Merrett  et  al.,  1973;  Johnson, 
1974b).  The  result  is  a  miniature  adult  at  the  end  of  a  trans- 
formation period  covering  as  little  as  1 0  mm  of  growth  (Johnson, 
1974b:68).  In  other  scopelarchid  genera  these  and  other  adult 
characters  are  acquired  essentially  one  by  one  over  an  increment 
of  growth  ranging  from  15  to  50  or  more  mm  SL  [in  most 
transformation  occurs  over  an  actual  size  (SL)  range  of  1 5  mm 
to  40  or  50  mm].  Implications  of  changes  in  morphology  during 
transformation  in  terms  of  activity,  buoyancy,  feeding  and  other 
aspects  of  biology  are  discussed  for  B.  infans  in  Merrett  et  al. 
(1973). 

Relationships 

The  scopelarchids  were  poorly  known  until  the  completion 
of  Johnson's  ( 1 974b)  revision.  Currently  recognized  are  1 7  species 
grouped  in  4  genera.  Phylogenetic  analysis  involving  hypothe- 
sized derived  states  of  1 9  characters  or  character  complexes 
(Table  66)  supports  allocation  of  species  among  3  of  the  4  genera. 
As  will  be  shown,  Scopelarchoides  remains  a  problem.  In  the 
listing  that  follows  characters  are  given  a  character  number  (de- 
rived state  number).  Documentation  of  character  state  catego- 
rization and  hypothesized  polarity  are  given  in  references  listed 
in  the  key  to  Table  66.  Of  the  19  characters  for  which  polarity 
is  indicated,  6  involve  larval  features  (Table  65:  18,  19,  20,  22, 
23,  24).  Of  13  adult  characters,  5  represented  noval  autapo- 
morphies  (Table  65:  1,4,  13,  14,  15),  3  occur  in  a  sequence  of 
3  or  more  steps  (Table  65:  5,  11,  16),  and  5  represent  reductive 
characters  (Table  65:  6,  7,  8,  9,  12).  Rosenblattichthys  is  dis- 
tinctive in  having  a  greatly  enlarged  head  in  larvae  19  (19)  and 
precocious  development  of  the  pectoral  fins  20  (20).  A  single 
reductive  character  8  (7)  putati vely  links  the  remaining  1 4  species 
of  scopelarchids.  Scopelarchus  is  specialized  in  having  subequal 
dermal  pigment  stripes  above  and  below  the  lateral  line  4  (2), 
unique  support  of  the  first  epibranchial  16(17);  unique  confor- 
mation of  the  three  peritoneal  pigment  sections  22  (24),  and  in 
three  reductive  characters  9  (8),  12(11),  and  23  (25).  Scopelar- 
chus analis  is  linked  with  5.  michaelsarsi  and  5.  Stephens!  by 
one  reductive  character  11  (10).  Scopelarchus  stephensi  and  S. 
michaelsarsi  are  linked  by  a  reduced  number  of  vertebrae  5  (3) 
and  by  early  onset  and  completion  of  metamorphosis  24  (26). 
Benthalbella  is  specialized  in  having  delayed  but  then  extremely 
"rapid"  metamorphosis  24  (27)  and  in  three  reductive  char- 
acters 6  (5),  22  (21),  and  23  (25).  Linking  Benthalbella  dentata. 
B.  infans,  B.  lingutdens  and  B.  elongata  is  the  unique  presence 
of  a  hooklike  process  on  the  urohyal  15  (14)  and  two  reductive 
characters  9  (8)  and  1 1  (9). 

In  dealing  with  the  5  species  included  by  Johnson  (1974b)  in 
the  genus  Scopelarchoides  the  evidence  available  (Table  66,  Fig. 
1 28)  suggests  that  this  group  is  both  unnatural  and  paraphyetic. 
Linking  5.  nicholsi,  S.  danae  and  Scopelarchus  are  unique  se- 
quential and  fully  correlated  novel  autapomorphies:  support  of 
the  first  epibranchial  character  16  (states  15  -  16  -  17),  and 
number  and  position  of  peritoneal  pigment  sections,  character 
22  (states  22  -  23  ^  24).  Further  linking  5.  nicholsi  with  S. 
danae  and  Scopelarchus  are  relative  size  of  the  opercle  and 


248 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  66.    Characteristics  of  the  Scopelarchidae.  Characters  and  character  states  are  defined  and  listed  below.  Positive  integers  indicate 
derived  states,  zeroes  indicate  primitive  states,  letters  denote  states  of  characters  where  polarity  could  not  be  determined. 


Rosenhlalltcblh 

■< 

Scopelo 

rchus 

■s 

copeiarchotdes 

Benlhalhella 

Char- 

volu- 

michael- 

guenih- 

nich- 

sigii:- 

mac  ra- 

Imgui- 

acters 

alatus 

hubbsi 

cris 

stephensi 

sarsi 

a  nail  s 

en 

danae 

flist 

kreffn 

■Imm.x 

fer 

pt  una 

denlala 

ftongata 

inlans 

dens 

Gross  morphology 

1 

0 

0 

0 

0 

0 

0 

0 

0 

0 

?1 

0 

0 

0 

0 

0 

1 

0 

2 

a 

a 

a 

b 

b 

b 

b 

b 

b 

a 

b 

b 

a 

a 

a 

a 

a 

3 

c 

a 

a 

b 

b 

b 

b 

c 

c 

c 

c 

b 

c 

c 

c 

c 

?c 

4 

0 

0 

0 

2 

2 

2 

2 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

Meristic  characters 

5 

0 

0 

0 

3 

3 

0 

0 

0 

0 

4 

0 

0 

4 

4 

4 

4 

4 

Osteological  characters 

6 

?0 

?0 

0 

?0 

0 

0 

0 

0 

0 

?0 

?0 

0 

5 

5 

5 

5 

?5 

7 

?0 

?0 

0 

?6 

6 

6 

6 

6 

6 

?0 

■'0 

0 

0 

0 

0 

0 

?0 

8 

?0 

0 

0 

7 

7 

7 

7 

7 

7 

7 

?7 

7 

7 

7 

7 

7 

7 

9 

0 

0 

0 

8 

8 

8 

8 

0 

0 

0 

0 

0 

0 

8 

8 

8 

8 

10 

b 

b 

b 

a 

a 

a 

a 

a 

a 

b 

b 

a 

b 

b 

b 

b 

b 

11 

0 

0 

0 

10 

10 

10 

9 

9 

9 

0 

0 

0 

0 

9 

9 

9 

9 

12 

0 

?0 

0 

?11 

11 

11 

11 

0 

0 

?0 

?0 

0 

0 

0 

0 

0 

0 

13 

0 

0 

0 

12 

12 

12 

12 

12 

12 

0 

0 

0 

0 

0 

0 

0 

0 

14 

?0 

0 

0 

0 

0 

0 

0 

0 

0 

13 

13 

13 

13 

13 

13 

13 

13 

15 

?0 

?0 

0 

0 

0 

0 

0 

0 

0 

?0 

?0 

0 

0 

14 

14 

14 

14 

16 

?0 

?0 

0 

?17 

17 

17 

17 

16 

15 

?0 

?0 

0 

0 

0 

0 

0 

0 

Developmental  characters 

17 

b 

a 

a 

a 

a 

a 

a 

a 

a 

a 

b 

b 

b 

b 

b 

b 

b 

18 

0 

0 

0 

18 

18 

18 

18 

18 

0 

?0 

0 

0 

0 

0 

0 

0 

0 

19 

19 

19 

19 

0 

0 

0 

0 

0 

0 

?0 

0 

0 

0 

0 

0 

0 

0 

20 

20 

20 

20 

0 

0 

0 

0 

0 

0 

?0 

0 

0 

0 

0 

0 

0 

0 

21 

a 

a 

a 

b 

b 

b 

b 

a 

a 

?a 

a 

a 

b 

b 

b 

b 

b 

22 

0 

0 

0 

24 

24 

24 

24 

23 

22 

?0 

0 

0 

21 

21 

21 

21 

21 

23 

0 

0 

0 

25 

25 

25 

25 

0 

0 

?0 

0 

0 

25 

25 

25 

25 

25 

24 

0 

0 

0 

26 

26 

0 

0 

0 

0 

?0 

0 

0 

27 

27 

27 

27 

27 

KEY: 

Character  state  classification  and  hypothesized  polanty  based  on  detailed  information  presented  in  Johnson  (1974b.  1982)  and  Iwami  and  Abe  (1980).  Both  characters  (boldface,  in  brackets) 
and  character  states  (in  parentheses)  are  numbered  sequentially. 

Gross  morphohgy  —{\\  Luminous  tissue  is  (0)  absent.  (1)  present;  |2|  pelvic-fin  insertion  is  (a)  antenor  to  dorsal-fin  ongin.  (b)  prasienor  to  dorsal-fin  origin;  |31  length  of  pectoral  fin  is  (a) 
subequal  to,  (b)  distinctly  longer  than,  (c)  distinctly  shorter  than  length  of  pelvic  fin;  |4|  dermal  pigment  stnpes  as  equal  or  subequal  stripes  above  and  below  lateral  line  are  (0)  absent.  (2)  present. 

Merislic  characiers  —\5\  Modal  number  of  vertebrae.  Occurs  within  span  of  (3)  40  to  44.  (0)  45  to  51,  (4)  54  lo  65,  hypothesized  character  slate  sequence:  3-0-4. 

Oiieologica/  characlers  —\(t\  An  anterovcntrally  directed  prong  from  opisthotic  reaching  or  nearly  reaching  border  of  proolic  is  (0)  present.  (5)  absent;  |7|  panental  bones  are  (0)  present,  (6) 
absent;  |8]  supraorbital  bones  are  (0)  present.  (7)  absent;  |9|  antorbital  bones  are  (0)  present.  (8)  absent;  |10|  Ethmoid  process  on  first  infraorbital  bone  is  (a)  present,  (b)  absent.  |ll|  Supramaxillary 
bones  are  (0)  large,  one-third  to  one-fourth  the  maxillary  length;  (9)  splinllike.  less  than  one-ninth  of  maxillary  length,  (10)  absent:  hypothesized  character  state  sequence  0-9-10,  [12]  Discrete 
postenor  arm  of  hyomandibular  bone  which  articulates  with  opercle  is  (0)  present,  (11)  absent,  represented  only  by  a  rounded  ridge.  113)  Opercle— (0)  subequal  to  or  less  than,  (12)  distinctly 
great  than— subopercle  in  size.  1I4|  (0)  basibranchial  teeth  present,  basihyal  short,  (13)  basibranchial  teeth  absent,  basihyal  long.  |I5|  Hook-like  process  on  anterodorsal  surface  of  urohyal  is  (0) 
absent.  (14)  present,  |16|  (0)  suspensory  phar>ngobranchial  (PBl)  present,  uncinate  process  (UP)  of  first  epibranchial  (EBl)  and  second  pharyngobrancial  (PB2)  connected  by  a  ligament;  (15) 
FBI  lacking,  support  of  EBl  near  proximal  end  of  PB2  — UP  of  EBl  and  PB2  connected  by  a  ligament;  (16)  PBl  lacking,  support  of  EBI  near  middle  of  PB2.  no  UP.  (17)  PBl  lacking,  support 
of  EBl  at  point  of  articulation  between  PB2  and  EB2.  no  UP.  Hypothesized  character  state  sequence:  0-15-16-17 

Developmental  characiers  — |I7|  Dermal  pigmentation  as  defined  in  text  is  (a)  present,  (b)  absent,  [181  Adipose  fin  (0)  remains  elongate  (extending  antenad  to  over  antenor  anal-fin  base) 
throughout  transformation  penod,  (18)  is  reabsorbed  early  in  transformation,  exhibiting  adult  proportions  in  specimens  20  to  22  mm  SL  and  larger,  II9|  Head  length  in  larvae  (=28  mm  SL)  (0) 
not  exceeding  30%  SL.  (19)  exceeding  30%  SL.  |20|  Pectoral  fin  (0)  not  precocious,  all  other  fins  with  completely  differentiated  rays  pnor  to  ossification  of  ventralmost  rays  {at  least)  of  pectoral 
fin,  (20)  precocious,  all  rays  completely  differentiated  pnor  to  formation  of  complete  complement  of  rays  of  all  other  fins  (except  caudal  fin),  (211  Pelvic  fin  buds  (a)  form  midlaterally,  well  above 
level  of  intestine,  (b)  form  ventrolaterally,  at  or  below  level  of  intestine,  |221  Number  of  pentoncal  pigment  sections  in  larvae  (2 1 )  =  0.  (0)  =  I ,  (22)  =  3,  the  postenor  paired  sections  appeanng 
much  later  in  development  than  the  single  antenor  section,  and  appeanng  entirely  antenor  to  the  pelvic-fin  bases.  (23)  =  3.  the  postenor  paired  sections  appeanng  much  later  in  development 
than  the  single  antenor  section,  and  appeanng  over  the  pelvic-fin  bases.  (24)  =  3.  the  postenor  paired  sections  app)eanng  in  near  synchrony  with  the  single  antenor  section  and  appeanng  entirely 
posterior  to  the  pelvic-fin  bases.  Hypothesized  character  state  sequence:  21  -  0  -  22  -  23  -  24,  |23|  Other  pigment  spots  or  areas  (as  defined  in  text)  are  (0)  present.  (25)  absent,  |24| 
Transformation  is  (26)  gradual,  onset  at  12-14  mm  SL  or  smaller,  completion  at  30-35  mm  SL  or  smaller.  (0)  gradual,  onset  at  16-22  mm  SL  or  larger,  completion  at  40-60  mm  SL  (most 
species,  R  alaius  is  extreme  with  onset  at  9-10  mm  SL  and  not  yet  complete  in  6  (39,9-80. 1  mm  SL)  juveniles  examined  by  Johnson  { 1 974b)],  (27)  abrupt;  onset  at  49  6-89  I  mm  SL  or  larger, 
completion  at  68.3-98.6  mm  SL  or  larger  (size  for  both  onset  and  completion  of  metamorphosis  vanes  among  the  5  species  of  Benlhalhella).  Hypothesized  character  states  sequence:  26  -  0  - 
27. 


subopercle  13  (12)  and  two  reductive  characters  7  (6)  and  1 1 
(9).  Further  linking  S.  danae 'wiih  Scopelarchus  is  a  unique  early 
restriction  of  the  base  of  the  dorsal  adipose  fin  18  (18).  I  am 
convinced  that  the  characters  previously  detailed  warrant  ge- 
neric level  recognition  for  the  group  of  4  species  assigned  to 
Scopelarchus.  Thus  Scopelarchoides  (type-species  S.  nicholsi) 
should  be  restricted  to  S.  nicholsi  and  S.  danae. 

This  leaves  the  three  species  currently  assigned  to  Scopelar- 
choides, viz.  S.  signifer,  S.  climax,  and  5".  kreffti.  These  three 
share  no  known  derived  character  unique  to  just  this  group. 


They  share  a  single  presumably  derivative  character— loss  of 
basibranchial  teeth,  extension  of  length  of  basihyal  tooth  row 
14  (13)— with  Benlhalhella  but  as  noted  by  Johnson  (1974b: 
204)  this  may  represent  adult  retention  of  a  larval  character 
state  common  to  all  scopelarchids.  Scopelarchoides  kreffti.  a 
subtropical  convergence  species,  shares  with  Benlhalhella  an 
increase  in  the  number  of  vertebrae  5  (4)  and  probably  shares 
with  B.  infans  the  presence  of  luminous  tissue  1(1).  Most  os- 
teological characters  arc  unknown  for  5.  climax  and  S.  kreffti 
(as  a  resuU  of  paucity  of  available  material)  and  the  larvae  of 


JOHNSON:  SCOPELARCHIDAE 


249 


-26 

-    3 


—25 

—  24 

—  17 

—  11 
8 

—  2 


—  14 
9 

-  8 


23 

—18 
—16 


-22 
-15 
-12 

-  9 

-  6 


-20 
-19 


-27 
-25 
-21 

-    5 


Fig.  128.     Proposed  relationships  among  scopelarchid  species  based  on  adult  and  larval  characters.  Integers  indicate  derived  character  states, 
listed  in  Table  66,  possessed  by  taxa  above  indicated  point  in  dendrogram. 


S.  kreffti  are  unknown.  I  would  argue  that  the  specializations 
oi Benthalhella,  especially  in  larval  characters  relating  to  a  unique, 
rapid  pattern  of  transformation  preclude  addition  of  5.  signifer, 
S.  climax,  and  presumably  5.  kreffti  to  Benthalhella.  But  with 
S.  climax  and  S.  kreffti  very  poorly  known  and  with  the  only 
"character"  uniting  this  "group"  of  three  being  that  they  are 
"left  over,"  I  remain  with  my  1974b  (p.  217)  compromise. 
Uniting  all  5  species  of  "Scopelarchoides"  and  diagnostically 
separating  them  from  Scopelarchus  and  Benthalhella  are  de- 
velopment and  conformation  of  accessory  pigment  spots  char- 
acter 23,  and  lateral  appearance  of  the  pelvic  fin  bud.  character 
21.  It  is  possible  that  the  state  exhibited  by  Scopelarchoides 
larvae  is  primitive  in  both  cases  (I  doubt  that  lateral  appearance 
of  the  pelvic  fin  buds  is  primitive)  but  until  this  can  be  shown 
through  adequate  outgroup  comparison  and  until  S.  climax  and 
S.  kreffti  are  better  known,  I  refram  from  attempting  the  de- 
scription of  an  additional  genus.  Thus,  for  now,  the  possibly 
paraphyletic  genus  "Scopelarchoides"  is  retained. 

A  summary  of  the  contribution  of  6  ontogenetic  characters 
to  this  analysis  is  presented  below. 

Dermal  pigmentation  (character  #/7j.  — Dermal  pigmentation 
and/or  dermal  pigment  stripes  are  found  in  all  scopelarchid 
genera  except  Benthalhella.  however,  the  fixation  of  such  pig- 
ment into  subequal  stripes  above  and  below  the  lateral  line  is 
diagnostic  of  and  unique  to  the  four  species  of  Scopelarchus. 
This  fixation  is  regarded  as  autapomorphous  for  this  genus. 


.Adipose fin  (character  #18).— Scopelarchoides  danae  shares  with 
Scopelarchus  an  early  reabsorption  of  most  of  the  adipose  (fin, 
resulting  in  restriction  to  essentially  adult  proportions  of  the 
base  of  this  fin  in  specimens  20-22  mm  SL.  In  other  Scopelar- 
choides as  in  Benthalhella  and  Rosenhlattichthys  the  adipose  fin 
remains  elongate,  to  over  the  anterior  anal  fin  rays,  throughout 
transformation,  assuming  adult  proportions  in  specimens  >30 
mm  SL.  In  combination  with  other  characters  uniting  .S.  danae 
with  Scopelarchus  (Fig.  128)  fixation  of  early  restriction  of  the 
dorsal  adipose  base  is  regarded  as  apomorphous  for  this  group. 

Head  length  (character  #19}.  — The  head  in  larval  Rosenhlatt- 
ichthys is  unusually  large,  deep  and  massive,  the  head  length 
exceeding  30%  of  the  SL.  The  head  length  in  other  scopelarchid 
larvae  does  not  exceed  30%  of  the  SL  and  this  is  apparently  the 
caseinchlorophthalmoids(Taning,  1918;Okiyama,  1972,  1974b, 
1981)  and  most  alepisauroids  (Rofen,  1966a:  Johnson,  1982). 
Larvae  of  Omosudis  and  .Alepisaiirus  do  exhibit  exceptionally 
large  heads  (Rofen,  1966b).  The  fixation  of  this  character  in 
Rosenhlattichthys  alone  among  scopelarchids  is  presumed  to  be 
apomorphous. 

Pectoral  fin  development  (character  #20).  — The  order  of  fin  ray 
differentiation  varies  within  and  between  iniomous  families. 
Precocious  pectoral  fin  development  is  unique  to  Rosenhlattich- 
thys among  scopelarchids.  It  is  also  found  in  ipnopids  (Okiyama, 
1972,  1981)  and  myctophids  (Moser  and  Ahlstrom,  1970)  but 


250 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


not  evermannellids,  Omosudis.  or  chlorophthalmids  (Tuning, 
1918;  Rofen,  1966b;  Johnson,  1982).  It  is  presumed  that  pre- 
cocious pectoral  fin  development  in  Rosenblattichthys  is  the 
derived  state. 

Peritoneal  pigment  sections  (character  #22).  — For  an  overview 
of  the  distribution  of  peritoneal  pigment  sections  in  inioms  see 
Johnson  (1982)  and  the  account  of  the  Evermannellidae  in  the 
present  work.  The  single,  transverse  section  seen  in  Rosenblatt- 
ichthys. Scopelarchoides  climax.  S.  signifer  and  presumably  5. 
kreffti  is  here  considered  the  primitive  state.  Loss  of  peritoneal 
pigment  in  the  larvae  of  Benthalbella  is  clearly  apomorphous. 
The  single  and  paired  conformation  of  the  3  sections  in  Sco- 
pelarchoides nicholsi.  S.  danae  and  Scopelarchus  is  unique  to 
this  lineage  among  inioms.  The  seemingly  sequential  progres- 
sion of  states  22  -  23  -  24  (Table  66:  character  22)  and  the 
correlation  of  these  states  with  states  15-16-17  of  character 
16  strongly  reinforce  the  concept  of  monophyly  for  this  lineage. 

Larval  pigment  spots  (character  #23).  —  Deep-lying  pigment  spots 
or  areas  occur  widely  among  iniomous  fishes  (TSning,  1918; 
Gibbs,  1959;  Anderson  et  al.,  1966;  Rofen,  1966a;  Moser  and 
Ahlstrom,  1970;  Johnson,  1982)  and  their  presence  is  here  pre- 
sumed to  be  primitive.  As  noted  above,  the  position  and  relative 
size  of  the  spots  differs  between  and  is  diagnostic  of  Scopelar- 
choides (all  5  species)  vs  Rosenblattichthys. 

Transformation  (character  #24).— Larvae  of  Benthalbella  are 
unique  among  scopelarchids  and  possibly  among  inioms  in 
achieving  very  large  size— 50  to  100  mm  or  more  (varying  by 
species)  while  retaining  a  purely  larval  form  and  then  exhibiting 
a  very  "rapid"  (based  on  size  increment  relative  to  total  size) 
transformation.  This  pattern  is  regarded  as  autapomorphous  for 
this  genus.  Larvae  of  two  central- water  species  of  Scopelarchus. 
S.  stephensi  and  S.  michaelsarsi.  exhibit  a  gradual  transfor- 
mation typical  for  most  inioms,  but,  relative  to  other  scopelar- 
chids, exhibit  onset  and  completion  of  transformation  at  sub- 
stantially smaller  sizes.  This  is  regarded  as  an  apomorphous 


feature  linking  these  two  species  (as  does  the  possibly  redundant 
character  5,  reduction  in  number  of  vertebrae). 

Johnson  ( 1 982:62-10 1 )  reviews  some  49  characters  seemingly 
related  to  the  question  of  sister-group  relationship  of  the  sco- 
pelarchids and  evermannellids.  Found  were  derived  states  in 
eight  characters— multiple  peritoneal  pigment  sections,  lateral 
attachment  of  dermosphenotic,  restricted  insertion  of  RAB  (Ro- 
sen, 1973)  muscle,  reduction  in  number  of  supraneurals,  and 
loss  of  the  following:  sclerotic  bones,  antorbital  bones,  tooth- 
plate  of  second  pharyngobranchial  and  basibranchial  denti- 
tion—characteristic of  all  alepisauroids  (Alepisauridae,  Ano- 
topteridae,  Evermannellidae,  Omosudidae,  Paralepididae)  but 
not  the  Scopelarchidae  (at  least  primitively).  Also  found  were 
5  derived  states  characteristic  of  the  Evermannellidae  +  Alep- 
isauridae +  Omosudidae  but  not  the  Scopelarchidae.  viz.  pos- 
session of  eight  infraorbital  bones,  reduction  in  number  of  ep- 
urals  and  loss  of  the  following:  body  scales,  lateral  line  scales, 
suspensory  pharyngobranchial.  Admittedly  many  of  the  features 
listed  are  "loss"  characters  and  thus  potentially  worrisome,  but 
why  should  they  uniformly  be  absent  in  the  groups  indicated 
and  not  in  the  Scopelarchidae  if  their  correlated  loss  is  not 
indicative  of  relationship?  On  the  basis  of  the  large  number  of 
derived  states  shared  among  alepisauroids  but  not  shared  by 
scopelarchids  Johnson  (1982)  excludes  the  scopelarchids  from 
the  alepisauroids  and  links  them  (tentatively)  with  chloroph- 
thalmoids.  Only  a  single  derived  state— gap  in  ossification 
between  first  centrum  and  the  skull— links  the  scopelarchids 
with  chlorophthalmoids,  but  this  feature  is  found  in  no  alepi- 
sauroid.  It  should  be  reemphasized  that  the  characters  discussed 
in  Johnson  (1982)  were  specifically  chosen  to  explore  the  hy- 
pothesis of  sister-group  relationship  of  evermannellids  and  sco- 
pelarchids—a  notion  rejected.  Many  additional  characters  need 
to  be  studied  for  any  rigorous  analysis  of  iniom  relationships. 
It  is  clear  that  the  contribution  of  larval  characters  to  this  anal- 
ysis will  be  great. 

Field  Museum  of  Natural  History,  Roosevelt  Road  at 
Lake  Shore  Drive,  Chicago,  Illinois  60605. 


Evermannellidae:  Development  and  Relationships 
R.  K.  Johnson 


THE  Evermannellidae  is  one  of  five  families  included  by 
Johnson  (1982,  the  most  recent  revision)  in  the  primarily 
oceanic  Alepisauroidei.  Excluded  from  this  group  are  the  Sco- 
pelarchidae, long  the  supposed  sister  group  of  the  evermannel- 
lids, but  tentatively  allied  by  Johnson  with  the  chlorophthal- 
moids. All  evermannellids  are  oceanic  and  mesopelagic, 
occupying  (as  juveniles  and  adults)  a  wide  vertical  range  in  the 
upper  1,000  m,  and  are  not  known  to  exhibit  diel  vertical  mi- 
gration. Evermannellids  are  relatively  large-bodied  (to  184.5 
mm  SL)  predators,  capable  of  engorging  large  food  particles, 
and  concentrating  most  frequently  on  fish  although  Coccorella 
may  more  frequently  prey  on  squid.  The  family  contains  7  species 


arranged  in  3  genera.  Evermannellids  are  distinguished  among 
other  alepisauroids  by  the  following  combination  of  characters: 
( 1)  an  externally  visible  tripartite  division  of  the  tail  musculature 
with  the  epaxial  and  hypaxial  muscles  separated  by  a  midlateral 
band  of  muscle  tissue,  the  lateralis  superficialis;  (2)  lack  of  scales; 
(3)  greatly  reduced,  edentate  basihyal;  (4)  restriction  of  gill  teeth 
to  ceratobranchial  of  second  arch;  (5)  presence  of  tubular  or 
semitubular  eyes  in  6  of  7  species;  (6)  lack  of  external  keels  on 
body.  The  genera  and  species  are  distinguished  by  gross  mor- 
phological (eye,  laterosensory  pores,  gut  morphology,  luminous 
tissue),  meristic,  morphometric,  osteological,  pigment  and  lar- 
val characters  (Table  67). 


JOHNSON:  EVERMANNELLIDAE 


251 


Fig.  129.  Larvae  and  juveniles  and  Evermannellidae.  (A)  E.  balbo.  showing  larval  phase  pigmentation,  D  3553  II,  8-10  mm  SL;  (B)  E.  indica. 
showing  juvenile  phase  pigmentation,  ORSTOM  CY  III-5,  28.0  mm  SL;  (C)  O.  normalops.  illustrating  larval  phase  pigmentation  and  multiple 
peritoneal  pigment  sections  (shown  in  solid  black),  UH  73/8/38,  10.5  mm  SL;  (D)  C.  allantica.  showing  juvenile  phase  pigmentation,  RHB  2960, 
6.3  mm  SL;  (E)  C.  allantica,  arrow  shows  location  of  cephalic  extension  of  pyloric  caecum,  ACRE  I2-18A,  25.2  mm  SL  (peritoneal  pigment 
sections  not  shown). 


Development 

Eggs  of  evermannellids  are  unknown.  Larvae  are  known  for 
all  species  and  developmental  series  have  been  partly  illustrated 
and  described  (Schmidt,  1918;  Rofen,  I966d;  Wassersug  and 
Johnson,  1976;  Johnson,  1982).  Osteological  examination  has 
been  confined  to  adults.  Development  is  direct,  transformation 
gradual,  adult  characteristics  are  acquired  essentially  one  by  one 
but  for  the  most  part  such  acquisition  is  complete  in  specimens 
exceeding  30  mm  SL. 

For  all  species  the  great  majority  of  larval  specimens  has  been 
taken  in  the  upper  100  m  but  only  the  larvae  of  three  species 
(Evermannella  balbo,  E.  indica,  Odontostomops  normalops)  have 
been  commonly  taken  in  hauls  to  50  m  or  less.  The  distributional 
ranges  of  larvae  and  adults  are  coextensive  and  there  is  no 
evidence  (the  data  are  very  incomplete)  for  seasonality  in  re- 
productive effort.  Evermannellids  are  synchronous  hermaph- 
rodites. 

The  following  paragraphs  describe  those  characters  most  ev- 
ident in  the  early  life  history  of  evermannellids  including  those 
of  value  in  distinguishing  genera  and  species. 

Gross  aspect  (Fig.  /29A  — Larvae  and  smaller  juveniles  of  all 
three  genera  are  similar  in  general  proportions  and  in  having  a 
relatively  smaller  eye,  smaller  lens,  broader  interorbital,  and 
larger  snout  than  larger  juveniles  and  adults.  The  body  is  deepest 
just  behind  the  pectoral  fin  base.  The  anterior  dorsal  profile 
descends  gradually  and  is  not  bowed  down.  The  eye  in  larvae 
of  Evermannella  and  Coccorella  but  not  Odontostomops  is  el- 
liptically  narrowed,  broader  dorsoventrally  than  antero-poste- 
riorly.  The  gut  cavity  is  essentially  triangular  and  quite  deep 
anteriorly.  The  snout  is  pointed,  the  mouth  large,  and  teeth 
appear  in  very  small  larvae.  The  most  striking  changes  in  body 


proportions,  in  all  evermannellid  larvae,  are  correlated  with  the 
transition  from  individuals  with  a  "larval  phase"  pigment  pat- 
tern to  those  with  a  "juvenile  phase"  pigment  pattern  (see  pig- 
mentation, below),  with  the  result  that  individuals  exceeding 
ca.  25  mm  in  the  latter  category  are  essentially  miniature  adults. 

Meristic  characters.— Counts  of  fin  rays  (Table  67)  do  not  differ 
between  larval  and  adult  specimens.  The  caudal  is  the  first  fin 
to  form,  it  develops  10  +  9  principal  rays,  as  in  all  Aulopiformes 
and  Myctophiformes  (sensu  Rosen,  1973).  Next  to  form,  in 
order,  are  the  dorsal,  pelvic,  anal  and  pectoral  fins.  The  pelvic 
fins  do  not  greatly  change  position  during  ontogeny,  they  appear 
ventrolaterally  beneath  the  posterior  half  of  the  dorsal  fin  and 
are  inserted  beneath  the  anterior  half  of  the  dorsal  fin  in  adults. 
An  adipose  fin  connects  the  incipient  dorsal  fin  with  the  caudal 
fin  in  small  larvae  but  loses  this  connection  and  shrinks  in  extent 
with  growth  of  the  individual,  inserted  over  posterior  one-third 
of  anal  fin  base  in  adults.  There  is  apparently  no  variation  in 
the  above-described  features  among  evermannellid  larvae. 

Peritoneal  pigment  sections  (Fig.  129).  — In  all  adult  everman- 
nellids the  gut  is  completely  enclosed  by  a  uniform  lube  of  dark 
brown  to  black  peritoneal  pigment.  In  larvae,  this  peritoneal 
pigment  appears  in  discrete  sections.  In  Odontostomops  there 
are  12  or  more  peritoneal  pigment  sections,  typically  13  to  15. 
In  Evermannella  and  Coccorella  there  are  invariably  3  sections, 
one  centered  over  and  medial  to  the  pectoral  fin  insertion,  one 
centered  (or  nearly  so)  under  the  dorsal  fin  insertion,  and  one 
(roughly)  centered  between  the  posteriormost  pelvic  fin  ray  base 
and  the  anal  fin  origin.  In  all  cases  the  sections  are  unpaired 
and  are  connected  broadly  over  the  dorsal  surface  of  the  stom- 
ach. In  small  larvae  the  sections  form  canopy-like  continuous 


252 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  67.    Characteristics  of  the  Evermannellidae.  In  the  list  that  follows  only  characters  useful  in  distinguishing  evermannellid  taxa  are 
included.  Those  characters  also  included  in  phylogenetic  analysis  are  numbered;  presumed  primitive  states  denoted  by  0.  presumed  derivative 

states  by  integers. 


Coccorella 
allantica 


Coccorella        Everntannella     Evermannelta     Evermanneita     Evermannella    Odonloscomops 
atrata  ahblromi  halbo  tndtca  megatops  normatops 


Gross  morphology 

(1)  Eye  (each  state  includes  a  suite  of  presumably  cor- 
related features  listed  in  Johnson,  1982,  p.  68):  (0) 
nontubular,  (I)  semitubular,  (2)  tubular 

(2)  Pyloric  caecum  with  cephalic  extension:  (0)  absent, 
(3)  present 

(3)  Luminous  tissue,  associated  with  ventral  wall  of 
intestine  and  pyloric  caecum;  (0)  absent,  (4)  pres- 
ent 

(4)  Medial  snout-pad  pore  (Johnson,  1982,  p.  8)  is:  (0) 
present,  (5)  absent 

Meristic  characters 

(5)  Dorsal  fin,  modal  number  of  rays;  (0)  12  or  13,  (6) 
10  or  II 

(6)  Number  of  lateral  line  segments: 
(0)  S43,  (7)  £34,(8)  SI8 
—Anal  fin  rays 

—  Vertebrae 

Morphometric  characters  (as  thousandths  of  SL) 
—Body  depth  at  anal-fin  origin 
—Horizontal  diameter  of  eye 

—  Vertical  diameter  of  eye 

—  Interorbital  width 

—  Length  of  longest  palatine  tooth 

Osteological  characters 

(7)  Basisphenoid;  (0)  present,  (9)  absent 

(8)  Ethmoid  cartilage:  (0)  not  forming  orbital  septum, 
(10)  considerably  expanded  posteriorly  forming  an 
orbital  septum 

(9)  Supraorbitals;  (0)  present,  (II)  absent 

(10)  Vertically  elongate  fossa  centered  at  dentary  sym- 
physis; (0)  absent,  (12)  present 

(11)  Jaw  and  palatine  teeth;  (0)  dentary  teeth  in  two  se- 
ries, at  least  some  dentary  and  palatine  fangs 
barbed,  (13)  dentary  teeth  uniserial,  all  fangs  un- 
barbed 

(12)  Basihyal  toothplate;  (0)  covers  dorsal  and  dorsolat- 
eral surface  of  basihyal,  (14)  covers  only  posterior 
2/3  of  dorsum  of  basihyal,  (15)  absent 

(13)  Toothplate  of  fourth  pharyngobranchial:  (0)  bears 
teeth,  ( 1 6)  edentate 

(14)  Toothplate  of  fifth  ceratobranchial:  (0)  bears  teeth, 

(17)  edentate 

Developmental  characters 

( 1  5)  Number  of  peritoneal  pigment  sections;  (0)  three, 

(18)  twelve  or  more 

(16)  Juvenile  phase  pigmentation;  (19)  characterized  by 
development  of  three  distinct  rows  of  very  large 
melanophores,  each  row  associated  with  one  of 
three  main  divisions  of  tail  musculature,  (0)  juve- 
nile phase  pigmentation  not  as  above,  with  many 
more  melanophores  and  no  distinct  trilateral  pat- 
tern 


1 

1 

2 

2 

2 

2 

0 

3 

3 

0 

0 

0 

0 

0 

4 

4 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

5 

0 

7 

7 

S 

8 

8 

8 

0 

26-30 

27-29 

29-32 

33-37 

27-31 

29-31 

30-35 

48-50 

45-47 

47-49 

52-54 

48-52 

48-50 

48-52 

144-191 

171-210 

173-200 

145-181 

136-173 

148-162 

135-170 

40-65 

47-65 

67-81 

52-72 

49-93 

74-85 

27-42 

42-76 

54-70 

69-87 

59-81 

60-97 

86-110 

28-40 

32-47 

47-61 

17-26 

9-19 

8-20 

4-17 

36-52 

71-96 

80-100 

46-69 

54-69 

48-73 

61-69 

53-69 

10 

10 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

11 

13 


13 


12 


12 


12 


12 


14 

15 

0 

0 

0 

0 

0 

0 

16 

0 

0 

0 

0 

0 

17 

17 

17 

0 

17 

17 

0 

18 


19 


19 


19 


19 


JOHNSON:  EVERMANNELLIDAE 


253 


sheets  over  the  dorsal  and  dorsolateral  margins  of  the  gut  and 
these  sections  expand  ventrad  as  well  as  longitudinally  with 
growth.  In  specimens  larger  than  35  to  45  mm  SL  the  peritoneal 
pigment  sections  coalesce  to  form  the  complete  gut-enclosing 
pigment  tube  characteristic  of  adults. 

Other  pigmentation  (Fig.  129).— The  major  pattern  of  body  pig- 
mentation in  evermannellid  larvae  occurs  in  two  phases,  a  larval 
phase  and  a  juvenile  phase,  with  a  gradual  transition  between 
the  phases.  In  smaller  larvae  (less  than  12-15  mm  SL)  the  most 
prominent  body  pigmentation  consists  of  a  pattern  of  pigment 
bands  arranged  along  the  myosepta.  Typically  these  bands  are 
arranged  in  groups  (symmetrically  distributed  in  epaxial  and 
hypaxial  myotomal  bands  in  the  tail  region,  nonsymmetrical 
and  predominantly  epaxial  in  the  trunk  region),  resulting  in  a 
characteristic  barred  appearance.  In  larvae  larger  than  1 2  to  15 
mm  SL  the  body  pigmentation  characteristic  of  adults  begins 
to  appear.  In  Odontostomops  juvenile  phase  pigmentation  is 
characterized  by  the  development  of  numerous  highly  punctate 
melanophores  generally  distributed  over  the  head  and  body.  In 
Evermannella  the  juvenile  phase  is  typically  characterized  by 
the  development  of  three  rows  of  very  large  melanophores,  each 
row  associated  with  one  of  the  3  main  divisions  of  the  trunk/ 
tail  musculature.  The  median  row,  that  associated  with  the  lat- 
eralis superficialis,  is  limited  to  the  tail.  Both  of  the  other  rows, 
epaxial  and  hypaxial,  extend  the  length  of  the  body,  from  the 
posterior  border  of  the  head  (or  nearly  so)  to  the  caudal  peduncle. 
In  Coccorella  the  juvenile  phase  pigmentation  tends  to  be  in- 
termediate in  state  between  that  of  Odontostomops  and  Ever- 
mannella. the  developing  melanophores  tend  to  be  larger  and 
more  prominent  than  in  Odontostomops.  but  much  more  nu- 
merous and  not  arranged  in  rows  as  in  Evermannella.  Body 
pigmentation  in  juveniles  larger  than  25  to  30  mm  SL  is  similar 
to  that  in  adults.  Development  of  adult  pigmentation  in  ever- 
mannclid  larvae  is  associated  with  gradual  (all  statements  im- 
plying time  course  are  based  solely  on  size  increments)  disap- 
pearance of  the  larval  myoseptal  pigment  bands.  Four  of  the 
seven  evermannellid  species  (Coccorella  atlantica.  C.  atrata. 
Evermannella  megalops.  Odontostomops  normalops)  are  highly 
melanistic  as  adults.  In  Evermannella  balho.  E.  indica.  and  es- 
pecially E.  ahlstromi  the  pigmentation  in  adults  tends  to  be 
much  more  mottled,  with  numerous,  variably-sized  melano- 
phores (some  very  large)  on  a  light  brown  (in  alcohol)  ground 
color.  Obscured  in  adults  is  the  longitudinal  tnlateral  melano- 
phore  pattern  characteristic  of  juveniles. 

Gut  morphology  (Fig.  129).  — \n  all  evermannellids  the  stomach 
is  a  heavily  muscularized  blind  sac.  The  stomach  expands  pos- 
teriad  with  larval  growth  reaching  its  full  extension  (to  a  vertical 
just  behind  the  pelvic  fin  base)  in  specimens  exceeding  20-25 
mm  SL.  Larvae  of  Coccorella  are  distinguished  by  the  unique 
possession  of  a  pyloric  caecum  that  expands  anteriad  with  growth 
and  enters  the  head  in  larger  larvae,  juveniles  and  adults  (Fig. 
129E).  The  caecum  is  visible  as  a  short,  blind,  bud-like  sac  on 
the  ventro-anterior  margin  of  the  intestine  in  the  smallest  known 
larvae  of  Coccorella.  Wassersug  and  Johnson  (1976)  describe 
in  detail  the  structure  and  development  of  this  remarkable  or- 
gan. Neither  Evermannella  nor  Odontostomops  nor  (as  far  as  is 
known)  any  other  alepisauroid  possess  a  pyloric  caecum. 

rra«s/orwa?/o/i.  — Development  of  juvenile  phase  pigmentation 
signals  the  onset  of  transformation  in  all  evermannellid  larvae. 


2,8,12,19 


Fig.  1 30.  Proposed  relationships  among  evermannellid  species  based 
on  adult  and  larval  characters.  Integers  indicate  derived  character  states, 
listed  in  Table  67,  possessed  by  taxa  above  indicated  point  in  dendro- 
gram. 


Transformation  in  Evermannellidae  is  gradual,  adult  characters 
are  essentially  acquired  one  by  one,  and  there  are  no  abrupt  and 
radical  changes  in  morphology.  In  all  evermannellid  species, 
individuals  larger  than  25  to  30  mm  SL  are  (except  for  final 
fusion  of  pentoneal  pigment)  essentially  miniature  adults  and 
can  be  distinguished  readily  on  the  basis  of  adult  characters 
(e.g.,  eye  morphology,  presence  or  absence  of  dentary  fossa, 
posterior  extent  of  lateral  line,  arrangement  of  cephalic  latero- 
sensory  pores,  dentition,  pigmentation,  meristic  and  morpho- 
metric  characters).  Final  fusion  of  the  peritoneal  pigment  sec- 
tions occurs  by  about  35  mm  SL  (Coccorella.  Evermannella)  or 
by  about  45  mm  SL  (Odontostomops). 

Relationships 

The  evermannellids  were  poorly  known  until  the  completion 
of  Johnson's  (1982)  revision.  Currently  recognized  are  7  species 
in  3  genera  (Fig.  130).  Phylogenetic  analysis  involving  presum- 
ably derived  states  of  1 6  characters  or  character  complexes  sup- 
ported previous  allocation  of  species  among  the  3  genera.  In  the 
listing  that  follows  characters  are  given  as  character  number 
(derived  state  number).  Of  the  1 6  characters,  2  involved  larval 
features  (Table  67:  15,  16).  Of  the  14  adult  characters,  5  rep- 
resented novel  autapomorphies  (Table  67:  2.  3,  8,  10,  II),  3 
exhibited  a  sequence  of  3  steps  (Table  67:  1,6,  12)  and  6  rep- 
resent reductive  characters  (Table  67:  4,  5,  7,  9,  13,  14).  Odon- 
tostomops is  specialized  in  having  12  or  more  serially  arranged 
pentoneal  pigment  sections  15  (18)  and  in  two  reductive  char- 
acters 7  (9)  and  9  (11).  Coccorella  exhibits  autapomorphies  in 
four  characters:  cephalic  extension  of  pyloric  caecum  2(3),  pres- 
ence of  luminous  tissue  3  (4),  posterior  expansion  of  ethmoid 
cartilage  8  (10),  arrangement  and  morphology  of  dentary  and 
palatine  teeth  II  (13)  and  is  apomorphous  in  two  additional 
reductive  characters  12  (14)  and  14  (17).  Coccorella  atrata  is 
apomorphous  in  two  reductive  characters,  12(15)  and  13  (16). 
Linking  Coccorella  and  Evermannella  are  intermediate  states 
in  the  two  3-step  characters  1(1)  and  6  (7).  Evermannella  shows 


254 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


autapomorphies  in  three  characters:  unique  pattern  of  juvenile 
phase  pigmentation  16  (19)  and  presence  of  vertically  elongate 
fossa  at  dentary  symphysis  10  (12),  presence  of  fully  tubular 
eye,  1  (2),  unique  to  them  among  evermannellids,  and  show 
further  reduction  in  the  number  of  lateral  line  segments  6  (8). 
A  single  reductive  character  14(17)  also  shared  with  Coccorella 
links  E.  indica  and  E.  ahlstromi  and  E.  megalops.  A  final,  ques- 
tionable character  5  (6)  links  the  latter  two.  In  each  case  well- 
defined  autapomorphous  features  support  the  hypothesis  of 
monophyly  of  each  genus  and  the  information  available  appears 
to  adequately  support  most  of  the  proposed  scheme. 

Details  concerning  the  contribution  of  two  larval  characters 
to  this  analysis  are  discussed  below. 

Peritoneal  pigment  ^frt/0/15.  —  Discrete  peritoneal  pigment  sec- 
tions are  striking  features  of  most  aulopiform  but  not  mycto- 
phiform  fishes  (Johnson,  1974b,  1982;  Okiyama,  1974b,  this 
volume).  A  single  dorsomedial  section  characterizes  the  larvae 
of  all  Aulopus  (Okiyama,  this  volume),  chlorophthalmoids  and 
(primitively)  scopelarchids.  Multiple  (3  or  more,  serially  ar- 
ranged, paired  or  unpaired)  sections  occur  in  ipnopids  (Bathy- 
pterois),  bathysaurids,  synodontids,  harpadontids,  paralepidids, 
Oinosudis  and  evermannellids.  Peritoneal  pigment  sections  are 
paired,  left  and  right,  in  synodontoids  (sensu  Johnson,  1982) 
but  single  and  connected  dorsomedially  over  the  gut  in  alepi- 
sauroids.  Peritoneal  pigment  sections  are  apparently  lacking  in 
notosudids,  some  ipnopids,  Alepisaurus,  neoscopelids  (perito- 
neal pigment  present  but  not  in  a  discrete  section,  see  Okiyama, 


this  volume)  and  myctophids.  Johnson  (1982)  concludes  that  a 
single  dorsomedial  section  is  primitive  for  aulopiform  fishes. 
Three  unpaired  sections  are  found  in  larvae  of  Coccorella,  Ev- 
ermannella,  Omosiidis  and  the  paralepidine  barracudina  Par- 
alepis  atlantica  (said  by  Rofen,  1966a:238,  to  be  ".  .  .  the  most 
primitive  species  in  the  Paralepididae.").  Larvae  of  Odontosto- 
niops  norinalops  exhibit  1 2  or  more  unpaired  peritoneal  pig- 
ment sections,  unique  in  the  order,  and  a  feature  regarded  as 
autapomorphous. 

Juvenile  phase  pigmentation— ]o\\nson  (1982)  regarded  fixa- 
tion of  the  trilateral  longitudinal  pattern  of  juvenile  phase  pig- 
mentation, as  described  above,  as  autapomorphous  for  the  genus 
Evermannella. 

It  has  long  been  supposed  (Gregory  and  Conrad.  1936;  Mar- 
shall, 1955;  Gosline  et  al.,  1966)  that  the  Scopelarchidae  and 
Everrriannellidae  are  closely  related.  This  supposition  was  based 
mainly  on  the  occurrence  of  tubular  eyes  in  both  groups.  John- 
son (1982)  argues  against  this  notion,  rejecting  any  close  rela- 
tionship of  the  Evermannellidae  and  Scopelarchidae,  placing 
the  latter  (tentatively)  among  the  chlorophthalmoids,  and  plac- 
ing the  Evermannellidae  as  the  sister  group  of  the  Omosudidae 
plus  Alepisauridae.  The  evidence  for  these  conclusions  is  pre- 
sented in  Johnson  ( 1 982)  and  briefly  summarized  in  the  account 
of  the  Scopelarchidae  in  the  present  work. 

Field  Museum  of  Natural  History,  Roosevelt  Road  at 
Lake  Shore  Drive,  Chicago,  Illinois  60605. 


Myctophiformes:  Relationships 
M.  Okjyama 


IN  the  traditional  concept,  the  order  Myctophiformes  is  con- 
sidered to  be  a  monophyletic  assemblage  with  taxa  having 
much  the  same  levels  of  organization,  even  though  they  have 
undergone  considerable  adaptive  radiation  including  some  ex- 
tremely specialized  forms  for  particular  habitats  (Goody,  1969; 
Marshall  and  Staiger,  1975;  Johnson,  1982). 

Modem  definition  of  the  order  including  16  families  was  first 
established  by  Gosline  et  al.  (1966).  They  recognized  the  fol- 
lowing two  suborders: 

Myctophoidei:  Aulopidae,  Synodontidae,  Bathysauridae, 
Harpadontidae,  Bathypteroidae,  Ipnopidae,  Chlorophthal- 
midae,  Myctophidae  and  Neoscopelidae. 

Alepisauroidei:  Notosudidae  (=Scopelosauridae),  Paralepidi- 
dae, Omosudidae,  Alepisauridae,  Anotopteridae,  Ever- 
mannellidae and  Scopelarchidae. 

This  dichotomous  system  has  been  generally  followed  by  re- 
cent workers  (Rosen  and  Patterson,  1969;  Marshall  and  Staiger, 
1975;  Sulak,  1977),  despite  some  minor  changes  or  disagree- 
ments in  the  definition  of  family  limits.  On  the  other  hand, 
Gosline  (1971)  proposed  the  idea  of  splitting  the  order  into  four 
groups  (!)  without  giving  rigorous  evidence. 

Rosen  (1973)  reevaluated  the  relationships  among  the  Myc- 
tophiformes and  produced  a  very  different  provisional  classi- 


fication based  on  a  cladistic  analysis  of  the  group,  where  all  of 
the  myctophiform  fishes  (except  Myctophidae  and  Neoscopeli- 
dae) form  a  monophyletic  group,  and  likewise  all  the  alepisau- 
roid  families  (except  Giganturidae)  form  a  monophyletic  assem- 
blage. His  phyletic  hypothesis  is  radically  different  from  those 
of  Gosline  et  al.  (1966)  and  Johnson  (1982). 

Notosudidae  was  later  transferred  from  Alepisauroidei  to 
Myctophoidei  (Bertelsen  et  al.,  1976),  and  furthermore,  Sco- 
pelarchidae was  removed  from  Alepisauroidei  (sensu  lato)  in 
the  recent  study  of  Johnson  (1982)  who  further  subdivided  the 
order  into  five  possible  major  groups  in  three  perceived  lineages 
(Fig.  131). 

Among  these  studies,  Johnson  (1982)  is  unique  in  carefully 
evaluating  larval  characters  such  as  the  peritoneal  pigment  sec- 
tions and  the  stomach  pigmentation  in  juveniles,  in  considering 
myctophiform  phylogeny  with  special  references  to  Scopelar- 
chidae and  Evermannellidae. 

As  finely  reviewed  by  Kendall  ( 1 982),  myctophiforms  provide 
an  excellent  example  for  elucidating  systematic  relationships 
among  fishes  using  larval  characters,  because  larvae  are  known 
for  representatives  of  most  of  the  families  and  in  some  cases 
nearly  all  of  the  species  within  the  families.  Potential  usefulness 
of  the  larval  groups  in  this  connection  has  been  well  documented 
for  several  families  such  as  Myctophidae  (Moser  and  Ahlstrom, 


OKIYAMA:  MYCTOPHIFORMES 


255 


1 


AULOPIDAE 


<  w  w    —  O  t^    -• 


TTtH  myctophidae 
ttH  neoscopelidae 


y 


ttttHNOTOSUDIDAE 


I  nil    I  SCOPELARCHIDAE 


"pCHLOROPHTHALMIDAE 
ttH  IPNOPIDAE 


—   3 


TTT 


TTT 


tH  synodontidae     ^ 


harpadontidae 


TT 


T-i  bathysauridae 


^ 


TTTTTTTnr 


paralepididae 


N 


II  iiMiii  II    I  ANOTOPTERIDAE 


Mill 


-tH  EVERMANNELLIDAE 

St 


TTTTTrr 


tHOMOSUDIDAE 
ttH  ALEPISAURIDAE 


J 


Fig.  131.     Possible  interrelationships  among  myctophiform  fishes  (Johnson,  1982). 


1972,  1974),  Scopelarchidae  (Johnson,  1974b,  1982),  Notosu- 
didae  (Bertelsen  et  a!.,  1976)  and  Evermannellidae  (Johnson, 
1982).  At  higher  taxonomic  levels.  Okiyama  (1974b,  1979b, 
1981)  considered  the  relationships  among  families  with  partic- 
ular reference  to  the  peritoneal  pigment  sections  in  association 
with  the  meristic  features  of  the  axial  skeleton,  notably  precau- 
dal  and  caudal  vertebrae.  Larval  characters  of  possible  system- 
atic importance  among  Myctophoidei  in  Okiyama  ( 1 979b)  have 
been  closely  analyzed  by  Kendall  (1982)  in  establishing  familial 
interrelationships  on  the  basis  of  the  cladistic  method,  although 
several  larval  stages  critical  to  this  were  not  available  at  that 
time. 

Since  current  knowledge  reveals  slightly  different  conclusions 
for  larval  characters  of  potential  phylogenetic  importance  from 


those  employed  in  Okiyama  ( 1 979b),  some  comments  are  given 
below  for  a  revised  character  catalogue  with  a  discussion  of 
possible  evolutionary  direction.  The  determination  of  this  di- 
rectional change  is  generally  based  on  the  assumption  that  the 
family  Aulopidae,  as  presently  considered,  represents  the  prim- 
itive character  state. 

In  the  following  discussion,  the  character  states  believed  to 
be  primitive  are  all  identified  with  a  "0,"  and  those  believed  to 
be  derived  are  designated  by  a  positive  integer. 

Peritoneal  pigment  sections  {1).—The  development  of  the  dis- 
crete peritoneal  pigment  sections  is  a  remarkable  feature  of  lar- 
val myctophiform  fishes.  Nothing  is  known  of  their  function, 
but  the  systematic  importance  of  this  unique  structure  has  been 


256 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  68.    Distribution  of  Larval  Character   States  Among 
Myctophiform  Families. 


Characters 

Family 

1 

T 

3 

4 

5 

6 

7 

8 

D* 

Aulopidae  (Au) 

0 

0 

0 

0 

0 

0 

0 

0 

0 

Myctophidae  (My) 

3 

1 

0 

1 

0 

0 

0 

0 

3 

Neoscopelidae  (Ne) 

3 

1 

0 

1 

0 

1 

2 

0 

5 

Chlorophthalmidae  (Ch) 

0 

0 

0 

0 

0 

0 

0 

0 

0 

Ipnopidae  (Ip) 

0 

0 

1 

1 

0 

0 

0 

0 

2 

Notosudidae  (No) 

3 

0 

0 

0 

1 

0 

1 

0 

3 

Scopelarchidae  (Sc) 

0 

1 

0 

0 

1 

0 

0 

0 

-> 

Bathysauridae  (Ba) 

1 

1 

t 

0? 

0 

0 

1 

1 

5 

Harpadontidae  (Ha) 

2 

1 

0 

0 

0 

0 

1 

0 

3 

Synodonlidae  (Sy) 

2 

1 

0 

0 

0 

0 

1 

0 

3 

Alepisauridae  (Al) 

3 

0 

1 

1 

0 

1 

2 

6 

Anotoptendae  (An) 

3 

0 

0 

0 

0 

0 

1 

3 

Evermannellidae  (Ev) 

1 

1 

0 

0 

1 

0 

-> 

5 

Omosudidae  (Om) 

1 

0 

0 

0 

0 

1 

2 

4 

Paralepididae  (Pa) 

1 

0 

0 

0 

0 

0 

1 

3 

*  Number  of  denved  character  slates. 


repeatedly  emphasized  (i.e.,  Okiyama,  1974b,  1979b,  1981. 
Johnson,  1974b,  1982).  Contrary  to  earlier  understanding  (Oki- 
yama, 1974b),  much  diversity  of  this  character  has  been  re- 
vealed. Based  on  the  number  and  shape  of  the  sections,  a  pro- 
visional classification  is  as  follows:  (A)  Dorsomedial  pigment 
sections;  (A-1)  Single  patch— Aulopidae,  Chlorophthalnms. 
Bathytyphlops.  Rosenblattichthys.  Scopelarchoides  (in  part);  (A- 
2)  Many  (three  or  more)— Bathysauridae,  Bathypterois  (in  part), 
Sudis.  Omosudidae,  Evermannellidae;  (A-3)  Single  to  many 
patches  with  growth— Paralepididae  (except  Sudis).  (B)  Paired 
pigment  sections— Harpadontidae,  Synodontidae.  (C)  Dorso- 
medial and  paired  pigment  sections— Scopelarchoides  (in  part), 
Scopelarchus.  (D)  No  pigment  sections— Neoscopelidae  (except 
Solivomer),  Myctophidae,  Ipiiops.  Bathymicrops.  Bathypterois 
(in  part),  Benthalhella.  Notosudidae,  Alepisauridae,  Anotop- 
teridae. 

Rare  exceptions  are  also  known  for  several  of  these  types. 
The  only  known  exception  to  the  presence  of  the  A-3  type  in 
paralepidids  is  in  Notolepis  coatsi  with  a  single  pigment  section 
throughout  all  stages  (Efremenko,  1978,  1983a).  However,  the 
ontogenetic  development  of  this  section  into  the  extensive  per- 
itoneal pigment  tube  around  the  gut  as  in  other  paralepidids 
reveals  little  phylogenetic  difference  for  this  exception.  Among 
those  having  B-type,  some  Synodus  reportedly  lack  the  peri- 
toneal pigment  sections  and  may  represent  an  extremely  spe- 
cialized character  state  (Cressey,  1981).  On  the  contrary,  a  my- 
tophid,  Protomyctophuin  anderssoni.  is  known  to  develop  the 
serially  arranged  paired  pigment  patches  similar  to  those  of  B- 
type(MoserandAhlstrom,  1974;  Efremenko,  1976).  Their  over- 
all resemblance  including  this  pigmentation  may  be  a  result  of 
a  simple  convergence. 

As  is  clear  from  the  above  classification,  character  states  are 
remarkably  diverse  in  the  Scopelarchidae  and  Ipnopidae.  C- 
type,  peculiar  to  the  former,  is  of  particular  significance  in  sug- 
gesting the  possible  direction  differentiating  the  paired  and  un- 
paired character  states  (Johnson,  1974b).  Unclear  limits  of  the 
family  are  partly  responsible  for  the  confusion  in  Ipnopidae. 

It  is  generally  agreed  that  the  presence  of  a  single,  dorsomedial 
peritoneal  pigment  section  (A-1  type)  represents  the  primitive 
state.  Since  A-3  and  C  types  are  referable  to  the  ranges  of  either 
A-1  or  2,  four  states  are  recognized  as  in  Johnson  ( 1 982);  (0)  = 


A  single,  dorsomedial  peritoneal  pigment  section.  (1)  =  Multi- 
ple (3  or  more),  serially  arranged,  unpaired  peritoneal  pigment 
sections.  (2)  =  Multiple  (3  or  more),  serially  arranged,  paired 
peritoneal  pigment  sections.  (3)  =  Peritoneal  pigment  section 
absent. 

Position  of  anus  f2A— Contrary  to  the  usual  pattern  of  the  anus 
location  immediately  anterior  to  the  anal  fin  origin  in  much 
teleosts,  a  more  or  less  wide  preanal  interspace  is  commonly 
shared  by  many  taxa  of  this  order.  This  character  can  be  of  much 
use  in  distinguishing  the  groups  of  Myctophiformes  (Rosen, 
1 97 1 ;  Okiyama,  1 979b).  The  character  states  of  the  diverse  anus 
location  relative  to  the  pelvic  fins  are  not  recognized  herewith 
due  to  the  unclear  patterns  of  occurrence. 

The  ontogenetic  rearward  shift  of  the  anus  is  restricted  to 
some  speciose  families  such  as  Scopelarchidae,  Paralepididae 
(except  Sudis),  Notosudidae,  and  Myctophidae  (in  part).  There 
is,  however,  a  sharp  contrast  in  the  final  condition  among  them: 
no  preanal  interspace  in  Myctophidae  and  Scopelarchidae  vs  a 
distinct  space  in  the  remaining  two.  As  in  Kendall  ( 1982),  who 
employed  this  character  in  the  first  step  of  branching,  two  char- 
acter states  are  recognized.  (0)  =  Anus  with  interspace  from  the 
origin  of  the  anal  fin.  (1)  =  Anus  just  in  front  of  the  ongin  of 
the  anal  fin. 

Fin  features  (3).  — Except  Bathysauridae  with  magnificently  en- 
larged fins,  the  elongated  pectoral  fins  are  the  pronounced  larval 
character  found  in  many  representatives  of  this  order.  The  Ip- 
nopidae displays  the  most  diverse  pattern  of  specialized  pectoral 
form  in  being  bifid,  large  and  fan-like,  or  extremely  elongated. 
Parallel  features  are  known  to  occur  sporadically  in  some  spe- 
cialized myctophidae  (Moser  and  Ahlstrom,  1970,  1974).  Sco- 
pelarchidae is  another  member  of  less  cohesiveness  in  this  char- 
acter; prominent  pectoral  fins  are  peculiar  to  Rosenblattichthys, 
the  most  specialized  genus  in  this  family  (Johnson,  1974b). 
Likewise,  only  the  aberrant  genus  Sudis  has  elongated  pectoral 
fins  in  Paralepididae.  The  character  states  recognized  are:  (0)  = 
All  fins  short.  (1)  =  Only  pectoral  fins  elongated.  (2)  =  All  fins 
elongated. 

Sequence  of  fin  formation  i^-^A- Although  current  knowledge  is 
far  from  complete,  dichotomous  patterns  can  be  recognized  in 
the  sequence  of  fin  formation,  especially  in  the  pectoral  fins. 
Johnson  (1982)  defined  the  derived  character  state  of  Rosen- 
blattichthys by  the  development  of  pectorals  prior  to  all  fins 
except  caudal.  The  precocious  nature  of  this  fin  apparently  rep- 
resents the  derived  state.  (0)  =  Pectoral  fins  not  precocious.  (1)  = 
Pectoral  fins  precocious. 

Eye  shape  (5).  — Moser  and  Ahlstrom  (1974)  showed  that  two 
types  of  eyes,  i.e.,  round  and  narrow,  reflect  the  major  two 
lineages  of  Myctophidae  with  several  exceptions.  These  patterns 
are  commonly  duplicated  at  familial  levels  in  this  order.  In  view 
of  the  specialized  morphology  of  the  narrow  eyes  including  the 
peculiar  choroid  tissue  and  following  the  suggested  phylogeny 
of  Myctophidae  (Moser  and  Ahlstrom,  1974),  round  or  nearly 
round  eyes  are  regarded  primitive.  The  states  recognized  are: 
(0)  =  Eyes  rounded  or  nearly  rounded.  (1)  =  Eyes  narrowed. 

Head  armature  (6).— The  development  of  head  spines  is  rare 
or  rather  exceptional  particularly  in  the  adult  myctophiform 
fishes.  However,  larvae  of  at  least  five  families  have  head  ar- 
mature. These  include  preopercular  spines  and  supraorbital  and/ 


OKIYAMA:  MYCTOPHIFORMES 


257 


Table  69.    Similarity  Matrix  of  1 5  Families  of  Mvctophiformes.  Based  on  the  total  number  of  characters  shared  in  the  same  state  regardless 

of  the  primitive  or  denved  (below  the  diagonal)  and  that  of  the  shared  derived  states  (above  the  diagonal,  with  similarity  index  in  parentheses). 

Subordinal  groups  are  indicated  by  enclosure.  Similarity  index  is  calculated  by  the  following  formula:  P„  =  (C,/\/S,S,)  x  100,  where  S,  and  S, 

are  number  of  derived  characters  in  families  i  and  j,  and  C„  is  number  of  the  shared  derived  states  between  the  same  set  of  families. 


My 


Ip 


Sy 


Au 

- 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

My 

5 

- 

|3(77)| 

0 

1  (40)  1  (33) 

1  (40)        1  (26) 

1(33) 

1  (33)         2 

(47) 

1(33) 

1(26) 

0 

0 

Ne 

3 
g 

[6] 
5 

3 

0 

1(22 

)  1  (26) 

1  (32) 

1(20) 

1(26) 
0 

1  (26)          1  (55) 
0                 0 

1(26) 
0 

2(40) 
0 

2(45) 
0 

0 

Ch 

0 

0 

0 

0 

0 

Ip 

6 

5 

3 

6^^ 

\^"-^ 

0 

0 

1(32) 

0 

0                2 

(58) 

0 

0 

0 

0 

No 

5 
6 
3 

4 

5 
3 

3 
3 

2 

5 
6 

3\ 

4 

^ 

1(45) 

1(20) 

1(33) 
1(41) 

1 (33)          1 

24) 

2(67) 

0 

1(26) 

1(26) 
2(64) 
1(20) 

0 
0 
0 

1  (33) 

So 

- 

1(32) 

1(41) 

0 

0 

Ba 

3 

2 

2 

2 

[2^2) 

2(52) 

0 

1  (26) 

Ha 

5 
5 
2 

5 

5 
4 

4 

4 
5 

5 
5 
2 

3 

3 
4 

5 
5 

2 

5 
5 

5 

"^ 

3(100) 

0 

1(33) 
1(33) 

1(26) 
1(26) 

0 
0 

1(33) 

Sy 

- 

0 

1(33) 

A! 

0 

2 

1 

1 

12^7) 

2(37) 

3(61) 

1  (24) 

An 

5 

4 

3 

5 

3 

6 

3              5 

5 

5 

4 

~~"\ 

,^^^^ 

^26) 

1(29) 

2(67) 

Ev 

3 

3 

3 

3 

2 

4 

5               5 

5 

5 

2 

4"^ 

,^^^^^. 

3(67) 

2(58) 

Om 

4 
5 

2 
3 

4 

2 

4 
5 

2 
3 

3 
4 

2               4 

3 
5 

3 

5 
3_ 

5 

7 

5 

"^ 

2(52) 

Pa 

3 

6 

5 

- 

or  frontal  ridges.  Development  of  head  armature  generally  oc- 
curs in  the  forms  with  a  massive  head  more  than  30%  of  body 
length,  thus  suggesting  the  specialized  condition  of  this  char- 
acter. Several  myctophid  species  {Lampanyctus)  having  pre- 
opercular  spines  provide  a  fine  example  of  this  trend,  while  this 
is  not  the  case  in  Scopelarchidae.  According  to  Nafpaktitis(1977), 
the  character  state  of  Neoscopelus  is  assigned  to  the  Neosco- 
pelidae.  The  states  recognized  are:  (0)  =  Head  armature  absent. 
(1)  =  Head  armature  present. 

Body  shape  f/A  — The  general  body  shape  can  range  from  ex- 
tremely slender  and  elongate  to  stubby  and  deep.  These  are 
tentatively  grouped  into  three  character  states  with  possible  evo- 
lutionary trends  towards  the  opposing  directions  from  the  mod- 
erately slender  body  shape  shared  by  primitive  groups  such  as 
Aulopidae  and  Chlorophthalmidae.  The  character  states  rec- 
ognized are:  (0)  =  Body  moderately  elongate.  (1)  =  Body  ex- 
tremely slender  and  elongate.  (2)  =  Body  stubby  and  deep. 

Pigment  spots  or  area  ("(SA— Johnson  (1982)  suggested  the  po- 
tential importance  of  pigmentation  other  than  the  peritoneal 
sections  in  the  systematics  of  the  Myctophiformes,  even  at  high 
taxonomic  levels.  A  difficulty  in  this  regard  is  how  to  recognize 
the  meaningful  character  states.  Based  on  the  various  pigmen- 
tation patterns  in  the  tails  of  larvae  (posterior  to  the  anus  except 
for  the  caudal  fin)  such  as  (a)  absent,  (b)  present  along  only  the 
ventral  midline,  and  (c)  present  along  lateral  or  dorsal  surfaces 
of  body  sometimes  forming  clear  bands,  formal  recognition  of 
this  character  is  undertaken.  Since  patterns  (a)  and  (b)  are  shared 
commonly  during  the  ontogeny  of  the  same  species,  two  char- 
acter states  are  recognized  with  the  assumption  that  (c)  repre- 
sents the  derived  state.  (0)  =  Pigment  spots  or  areas  in  tail  absent 
or  present  along  only  the  ventral  midline.  (1)  =  Pigment  spots 
or  areas  in  tail  present  along  lateral  or  dorsal  surface. 

The  primitive  or  derived  states  for  these  eight  characters  are 
summarized  in  Table  68.  Family  level  designation  of  character 


states  is  mostly  based  on  the  assumption  of  Johnson  (1982)  that 
"possession  by  one  or  more  representatives  of  a  particular  OTU 
of  a  state  considered  primitive  indicates  (except  where  contrary 
evidence  can  be  cited)  the  primitiveness  of  that  state  for  that 
OTU." 

A  similarity  matrix  based  on  the  total  numbers  of  characters 
shared  in  the  same  state,  regardless  of  whether  the  states  are 
primitive  or  derived,  is  given  below  the  diagonal  in  Table  69. 
Above  the  diagonal  are  shown  the  numbers  of  derived  characters 
shared  in  the  same  state  and  the  similarity  index  calculated  on 
the  same  data.  These  two  sets  of  figures  are  expected  to  reveal 
certain  clues  to  clarify  the  interfamilial  associations  of  this  order 
from  the  larval  standpoint. 

AiiLOPOiDEi:  Aulopidae 

So  far  as  the  selected  larval  characters  are  concerned,  the 
Aulopidae  can  not  be  separated  from  the  Chlorophthalmidae. 
This  unclear  distinction  is  due  to  the  limited  numbers  of  char- 
acters selected,  because  other  larval  and  adult  features  shown 
in  Table  70  reveal  the  trenchant  differences  between  them.  Of 
these,  the  possession  of  maxillary  teeth  and  fulcral  scales,  and 
the  earlier  differentiation  of  the  peritoneal  pigment  spots  well 
justify  the  distinct  and  less  specialized  systematic  status  of  the 
Aulopidae.  Other  aspects  of  sharp  contrast  such  as  in  the  den- 
tition, particularly  of  the  basihyal,  and  gut  morphology  sub- 
stantiate the  above  conclusion. 

Although  the  diversity  within  the  Aulopidae  once  suggested 
on  the  basis  of  larval  characters  (Okiyama,  1974b)  has  proved 
to  be  unacceptable,  there  still  remain  problems  concerning  the 
monotypic  nature  of  this  family.  As  mentioned  elsewhere  (Oki- 
yama, 1979b),  it  is  likely  that  the  Myctophiformes  evolved  along 
several  lines,  one  of  the  major  trends  being  the  elongation  of 
the  body  shape  accompanying  an  increase  in  vertebral  number. 
Obviously,  aulopids  lie  near  the  base  of  this  trend  with  clear 
orientation  toward  an  increase  in  the  number  of  abdominal 
components.  The  uniquely  elongated  larval  oesophagus  in  A. 


258 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table   70. 


Anatomical  Differences  of  Early  Stages  Betwfen 
AuLOPvs  and  Chlorophthalmus. 


Table  71. 


Distribution  of  Larval  Character  States  among  Four 
Genera  of  the  Ipnopidae. 


Autopus 


Chtorophlhatmits 
(afler  Rosen  (1971) 
and  Sulak(1977)] 


D« 


Maxillary  teeth 
Vomerine  teeth 


Basihyal 


Fulcral  scale 
Gut  morphology 


Peritoneal  pig- 
ment sections 


Present 

Only  two  widely  sepa- 
rated at  opposing 
anterolateral  comer 

Ovoidal  with  slightly 
indented  tip;  teeth 
absent 


Present 

Moderately  elongated, 

straight;  intestine 

slightly  fat 
Single;  distinct  at  less 

than  3.5  mm  SL 


Absent 

Transverse  row  of  six 
teeth  divided  into 
two  rows  of  three 
each 

Triangular  with  similar 
anterior  indentation; 
a  transverse  row  of 
six  teeth  divided  into 
two  series 

Present(?) 

Short,  compact  with 
slender  stomach;  in- 
testine fat 

Single;  distinct  at  more 
than  5  mm  SL 


japonicus  is  a  probable  indication  of  this  evolutionary  trend 
(Okiyama,  1974b).  Among  recent  congeners,  A.  damasi  may  be 
the  most  generalized  species  in  view  of  its  smallest  number  of 
vertebrae  (20  +  16)  similar  to  the  known  counts  in  the  fossil 
aulopids  (Goody,  1969;  Rosen  and  Patterson,  1969).  Further- 
more, this  species  is  clearly  separable  from  congeners  by  the 
mode  of  direct  association  between  the  first  haemal  spine  and 
anal  pterygiophores  (Okiyama,  1979b).  A  look  at  the  larvae  of 
A.  damasi  would  be  enlightening  in  clarifying  the  problem  in 
question. 

Myctophoidei:  Neoscopelidae, 
Myctophidae 

The  two  families  of  this  suborder  are  readily  discriminated 
from  the  others  by  the  greatest  similarity  index  value  based  on 
a  suite  of  derived  characters  (1  and  2)  not  shared  by  any  other 
families.  The  smaller  sizes  at  metamorphosis  are  also  peculiar 
to  these  families.  These  larval  evidences  offer  strong  support 
for  the  views  of  Moser  and  Ahlstrom  (1974)  and  Johnson  (1982), 
warranting  a  distinct  subordinal  ranking.  My  observation  of  the 
vertebrae  of  Solivomer  (see  Table  57  in  my  Myctophiformes: 
Development,  this  volume)  also  disclosed  their  closer  linkage 
than  assumed  by  Johnson  (1982). 

The  similarity  matrix  in  Table  69  would  offer  little  support 
for  Rosen's  scheme  to  transfer  these  families  to  a  different  order. 

Chlorophthalmoidei:  Notosudidae, 

Scopelarchidae,  Chlorophthalmidae, 

Ipnopidae 

The  larval  character  states  indexed  in  Table  7 1  are  less  prom- 
ising in  support  of  this  familial  assemblage,  because  only  the 
Notosudidae  and  Scopelarchidae  share  a  single  derived  char- 
acter state  (narrow  eye).  It  seems  that  this  ambiguity  is  also 
associated  with  the  inadequate  numbers  of  characters  in  ques- 
tion. 

Although  the  admitted  cohesiveness  of  larval  characters  of 
Chlorophthalmidae  may  be  altered  by  the  discovery  of  larval 
Bathysauropsis  or  Parasudis.  larval  characters  support  the  tra- 
ditional view  that  it  is  one  of  the  basal  stocks  of  this  order,  lying 


Bathytyphlops 
Ipnops 

Bathymicrops 
Bathypterois 


10  0  0        0        2 

10  0  10        4 

10  0  10        4 

10  0  0         15 


'  Number  of  denved  character  slates. 


at  a  somewhat  advanced  place  along  a  line  different  from  the 
Aulopidae.  Trenchant  characters  in  this  connection  such  as  the 
dentition  and  the  mode  of  anal  fin  support  are  shared  with  the 
Ipnopidae. 

Members  of  the  Notosudidae,  the  most  cohesive  family  in 
this  suborder,  have  the  greatest  numbers  of  derived  characters 
of  the  group.  Marshall  ( 1 966a)  and  Bertelsen  et  al.  ( 1976)  stated 
that  it  seems  most  closely  related  to  Chlorophthalmidae.  The 
superficial  resemblance  of  larval  stages  between  this  and  the 
Paralepididae  was  also  suggested  (Ahlstrom,  1972a).  On  the 
other  hand,  the  similarity  matrix  indicates  its  affinity  with  An- 
otopteridae,  along  with  Scopelarchidae.  Of  these  associations, 
the  last  grouping  based  on  a  single  derived  state  in  character  5 
(narrow  eye)  appears  less  arguable.  Other  features  such  as  the 
maxillary  teeth  and  the  uncommon  morphology  of  the  corpus 
cerebelli  suggest  the  aberrant  systematic  status  of  this  family. 

Since  Table  68  provides  few  clues  to  discuss  the  confused 
family  limits  of  the  Ipnopidae,  the  same  coding  of  the  character 
states  is  applied  to  the  four  genera  of  this  family  (Table  71). 
Except  for  the  distinct  larval  status  of  Bathypterois.  derived 
characters  shared  among  the  remaining  three  genera  do  not 
reveal  the  generic  linkages  suggested  by  Sulak  (1977).  By  the 
same  reasoning  as  discussed  before  concerning  the  relationships 
between  Aulopidae  and  Chlorophthalmidae,  the  derived  state 
in  character  1  (peritoneal  pigment  sections)  shared  by  Ipnops 
and  Bathymicrops  includes  the  different  states  of  gut  morphol- 
ogy. It  seems  these  genera  form  a  loose  but  distinct  assemblage 
warranting  family  rank.  Besides  the  shared  dentition  mentioned 
before,  the  close  fit  of  general  larval  morphology  between  Bathy- 
typhlops and  Chlorophthalmus  may  suggest  their  relationship. 

The  diverse  larval  characters  of  Scopelarchidae  were  elabo- 
rately enalyzed  in  the  light  of  adult  systematics  (Johnson,  1 974b). 
It  is  remarkable  that  this  family  has  no  phenetic  similarity  with 
Alepisauridae  in  terms  of  catalogued  characters.  On  the  other 
hand,  two  derived  states  in  character  2  (anus  location)  and  5 
(eye  shape)  shared  with  Evermannellidae  give  the  greatest  sim- 
ilarity index  value.  Johnson  (1982)  suggested  the  independent 
occurrence  of  the  tubular  eyes  in  adults  of  both  families,  but 
traditional  concepts  of  their  close  association  should  be  reevalu- 
ated using  larval  evidence. 

Synodontoidei:  Bathysauridae, 
Harpadonti[5ae,  Synodontidae 

Accepted  linkage  between  Synodontidae  and  Harpadontidae 
is  clearly  substantiated  by  the  larval  characters,  while  familial 
allocation  of  Saurida  remains  to  be  solved.  Synodus  lucioceps, 
having  the  intermediate  state  of  larval  characteristics  between 
these  families,  may  be  important  here.  The  relationships  among 
four  genera  are  thus  indistinct  from  the  standpoint  of  the  larvae, 
but  Saurida  appears  to  be  the  most  generalized.  Possible  phy- 


OKJYAMA:  MYCTOPHIFORMES 


259 


logenetic  association  between  Aulopidae  and  these  families  has 
been  suggested  on  the  basis  of  larval  characters  and  the  similar 
mode  of  anal  fin  support  (Okiyama,  1974b,  1979b).  To  these 
can  be  added  the  peculiar  structures  on  the  chorion  surface  of 
the  extremely  transparent  eggs,  the  pigmentation  patterns  in  the 
newly  hatched  larvae,  and  the  mode  of  reproduction  shared  by 
these  families,  characters  which  favor  their  close  association. 

Bathysauridae  is  distinguished  from  other  families  of  this 
suborder  by  some  trenchant  differences  in  the  peritoneal  pig- 
ment sections  and  the  mode  of  reproduction,  while  two  derived 
states  are  shared  by  all  families.  The  phylogenetic  relationship 
of  these  families  depends  on  whether  the  above  mentioned  dif- 
ferences are  due  to  divergence.  Larval  stages  of  Bathysauridae 
are  surely  highly  specialized,  adapting  to  a  prolonged  pelagic 
life,  but  larval  dentition  described  in  detail  by  Rosen  (1971) 
and  Johnson  (1974)  and  the  character  state  of  the  axial  skeleton, 
including  the  mode  of  anal  fin  support  (Okiyama,  1976b)  are 
of  particular  interest  in  showing  the  pattern  common  to  Ipno- 
pidae. 

Alepisauroidei:  Paralepididae, 

Anotopteridae,  Evermannellidae, 

Omosudidae,  Alepisauridae 

The  similarity  matrix  provides  certain  indication  of  the  co- 
hesiveness  of  this  suborder.  Most  remarkable  is  their  common 
sharing  of  the  derived  state  of  character  8.  Regarding  the  per- 
itoneal pigment  sections  dividing  five  families  into  two  groups, 
some  comments  are  warranted  for  Alepisauridae.  As  discussed 
by  Johnson  ( 1 982),  this  character  state  is  very  tentatively  defined 
due  to  the  inadequate  state  of  available  material.  Even  so,  a 
distinct  family  pair  of  Alepisauridae  and  Omosudidae  can  be 
readily  separated  from  the  remaining  families  by  the  many  de- 
rived character  states  shared  by  them.  Although  the  possibility 
of  their  convergence  cannot  be  fully  rejected  in  view  of  the  clear 
contrast  in  the  ontogenetic  aspects  of  the  pectoral  fins,  the  close 
similarity  between  Alepisaurus  ferox  and  Omosudis  lowei  (trop- 
ical western  Pacific  specimen)  (see  my  Myctophiformes:  De- 
velopment, Fig.  1 1 2B,  E,  F,  this  volume),  in  head  armature  and 
pigment  pattern  is  extremely  striking. 


An  association  between  the  Anotopteridae  and  Paralepididae, 
particularly  the  more  elongated  paralepidids  such  as  Stemo- 
nosudis  and  Macroparalepis  (Rofen,  1 966a,  c),  can  be  seen  from 
the  larval  standpoint.  In  addition  to  their  shared  derived  char- 
acter states  (character  7  and  8),  a  fleshy  projection  on  the  lower 
jaw  tip  peculiar  to  Anotopteridae  and  Stemonosudis  macrura. 
and  the  similar  larval  dentition  (huge  canines)  may  substantiate 
the  above  association.  Their  disagreement  in  the  character  of 
the  peritoneal  pigment  sections  is  probably  associated  with  the 
odd  systematic  position  of  Anotopteridae  lying  at  "an  extreme 
specialized  end-point  of  the  paralepidid  line"  (Rofen,  1966a,  c). 

On  the  basis  of  the  larval  characters,  two  subfamilies  of  Par- 
alepididae are  well  separated.  As  compared  with  the  relative 
constancy  of  conservative  characters  in  larval  Paralepidiinae, 
the  many  derived  character  states  of  larval  Sudinae  are  too 
specialized  to  be  consistent  with  the  accepted  subfamilial  level. 
The  latter  may  be  an  earlier  offshoot  preceding  the  remarkable 
paralepidine  radiation.  The  complete  lack  of  intermediate  forms 
between  them  offer  strong  support  for  this  suggestion. 

As  in  Scopelarchidae  (Johnson,  1974b),  the  systematics  of 
Evermannelidae  were  studied  in  detail  using  a  large  character 
suite,  including  larval  aspects  (Johnson,  1982).  So  far  as  the 
present  analysis  is  concerned,  this  family  seems  variously  as- 
sociated with  families  of  Alepisauroidei  such  as  Paralepididae, 
Alepisauridae  and  Omosudidae,  besides  Scopelarchidae.  It  is  of 
interest  that  limited  character  states  shared  by  Evermannellidae 
and  Alepisauridae  are  restricted  to  derived  ones,  probably  sug- 
gesting their  close  association.  Perhaps,  an  Evermannellidae  and 
Scopelarchidae  linkage  is  much  more  loose,  if  valid. 

Concerning  the  possible  three  main  lineages  in  this  order,  the 
larval  evidence  is  less  promising.  However,  additional  larval 
evidence  regarding  developmental  sequences,  including  osteol- 
ogy as  well  as  internal  morphology,  would  provide  much  more 
fruitful  information  for  elucidating  the  phylogeny  of  this  inter- 
esting group. 

Ocean  Research  Institlite,  University  of  Tokyo,  1-15-1, 
MiNAMiDAi,  Nakano-ku,  Tokvo  164,  Japan. 


Gadiformes:  Overview 
D.  M.  Cohen 


GADIFORMES  is  a  particularly  interesting  order  with  which 
to  work  because  it  encompasses  a  high  degree  of  diversity 
that  suggests  the  existence  of  several  lineages,  apparent  conver- 
gence and  reductive  trends  to  trap  the  unwary,  a  useful  fossil 
record  that  allows  a  consideration  of  the  distribution  in  time  of 
some  important  taxa  and  character  states,  and  new  suites  of 
characters  based  on  the  study  of  ELH  stages. 

Although  study  of  the  classification  of  gadiforms  dates  from 
pre-Linnean  times,  there  is  still  insufficient  properly  evaluated 
data  available  to  derive  a  phyletic  classification.  In  fact,  there 
is  not  even  agreement  as  to  what  should  be  included.  Berg  (1947) 


restricted  the  order  to  the  muraenolepids,  bregmacerotids,  mor- 
ids,  and  gadids  (including  merlucciids)  and  excluded  the  mac- 
rourids.  He  noted  primitive  and  advanced  characters  in  his 
gadiforms  and  suggested  derivation  from  primitive  fishes.  Rosen 
and  Patterson  (1969)  revived  an  expanded  Gadiformes  dating 
at  least  from  the  time  of  Gill,  which  included  not  only  gadoids 
and  macrouroids  but  also  ophidioids  and  zoarcoids,  and  which 
they  placed  in  a  supraorder  Paracanthopterygii,  postulated  as 
being,  "in  many  ways  more  primitive  than  the  acanthoptery- 
gians"  and  representing  "a  spiny-finned  radiation  more  or  less 
comparable  morphologically  with  that  of  the  Acanthopterygii" 


260 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


MERLUCCIUS 


BRECMACEROS 


EUCLICHTHYS 


CADINAE  (2) 
LOTINAE  (2) 
EUCLICHTHYS   ( u) 
MURAENOLEPIS  (2) 
PHYCINAE   12) 
MORIDAE  («-5) 
MELANONUS   ( if- 5) 
MERLUCCIUS   (21 
BRECMACEROS  (2) 


MELANONUS 


HYPURAL  RAYS 


Fig.  133.  Numbers  of  hypural  bones  (in  parentheses)  and  fin  rays 
supported  by  hypural  bones  in  nine  groups  of  gadiform  fishes.  Data 
from  Fahay  and  Markle  (this  volume)  and  original. 


TOTAL  CAUDAL  RAYS 

Fig.  132.  Total  caudal  rays  in  eight  groups  of  gadiform  fishes.  Data 
from  Fahay  and  Markle  (this  volume)  and  original. 

and  including  in  addition  to  their  gadiforms  the  polymixoids, 
percopsiforms.  batrachoids.  iophiiforms,  and  gobiesocoids. 
Gosline  (1968)  analyzed  the  characters  used  in  defining  the  ex- 
panded Gadiformes  and  concluded  that  ophidioids  and  zoar- 
coids  are  perciform  derivatives,  while  gadoids  are  widely  sep- 
arate and  probably  close  to  the  percopsiforms  (Gosline,  1963a). 
Marshall  and  Cohen  (1973),  whom  I  follow  for  present  purposes, 
restricted  the  Gadiformes  to  the  gadoids  and  macruroids  but 
did  not  consider  the  question  of  relationships.  In  the  following 
brief  preliminary  consideration  of  the  order,  I  discuss  several 
characters,  mention  the  groups  that  I  think  must  be  considered, 
and  outline  some  of  my  ideas  about  the  course  of  evolution  in 
the  gadiforms. 

Characters 

Several  character  complexes  that  require  consideration  are 
discussed  below.  Others  are  noted  later  under  groups  in  which 
they  occur.  Additional  relevant  information  is  presented  by  Fa- 
hay and  Markle  and  Dunn  and  Matarese  in  subsequent  sections 
of  this  volume. 

Caudal  fin.— Considering  the  fact  that  well  over  half  the  known 
species  of  gadiform  fishes  lack  the  slightest  vestige  of  a  caudal 
fin,  it  is  a  little  astonishing  how  much  importance  has  been 
attached  to  the  origin  and  homologies  of  the  various  skeletal 
supports  and  of  the  fin  rays  themselves.  There  is  no  denying, 
however,  that  when  present  the  gadiform  caudal  complex  is 
unique  in  several  respects.  Most  fish  groups  may  be  character- 
ized by  a  set  number  of  branched  caudal  rays.  Furthermore,  the 
branched  rays  are  generally  supported  by  only  hypurals.  In  gad- 
iforms with  tail  fins,  the  number  of  branched  caudal  rays  is 


highly  variable,  as  is  their  skeletal  support.  Bregmaceros  may 
have  as  few  as  1 2  branched  caudal  rays,  most  of  which  are 
supported  by  hypurals,  while  at  the  upper  end  of  the  range,  the 
lotine  Brosmc  may  have  as  many  as  43  branched  rays,  which 
are  supported  by  hypurals.  epurals,  and  haemal  and  neural  spines. 
This  high  degree  of  variation  in  an  otherwise  conservative  an- 
atomical complex  lends  credence  to  the  idea  of  Boulenger(1902) 
and  Regan  ( 1 903b)  that  the  caudal  fin  of  gadiforms  is  a  structure 
newly  evolved  from  an  essentially  tailless  condition  such  as  that 
of  the  macrourids  or  of  some  merlucciids.  It  was  partly  to  test 
Regan's  hypothesis  that  Barrington  (1937)  compared  the  de- 
velopment of  the  caudal  fin  of  Gadus  with  that  of  Pleuwnectes 
and  concluded  that,  although  the  tail  of  Gadus  was  unique  in 
several  respects,  it  could  have  been  derived  from  an  ordinary 
homocercal  tail  that  was  less  specialized  than  that  of  Pleuw- 
nectes. I  agree  with  Barrington.  Barrington  commented  also  on 
the  presence  in  gadids  of  a  high  number  of  procurrent  caudal 
rays,  which  he  interpreted  as  being  far  posterior  dorsal  and  anal 
rays,  so  that  the  functional  caudal  of  a  cod  is  composed  of 
elements  of  three  fins,  dorsal,  anal,  and  caudal  proper.  This 
interpretation  has  been  neither  falsified  nor  verified  by  the  study 
of  early  life  history  stages.  Barrington  coined  the  term  pseu- 
docaudal  for  what  he  took  to  be  this  kind  of  fin.  In  his  lectures 
and  during  conversations  with  me.  Ahlstrom  disagreed  with 
Barrington's  explanation  and  its  acceptance  by  Marshall  and 
Cohen  (1973)  because  procurrent  rays  lack  pterygiophores.  It  is 
instructive  to  note  in  this  respect  the  caudal  fin  structure  of 
Muraenolepis  (see  Fig.  1 43  of  Fahay  and  Markle  in  this  volume), 
which  has  confluent  vertical  fins  and  in  which  the  distinctive, 
elongate  pterygiophores  grade  into  hypurals.  It  is,  in  fact,  im- 
possible to  distinguish  between  the  last  anal  pterygiophore  and 
the  first  hypural  or  parhypural.  But  see  Fahay  and  Markle  later 
in  this  volume. 

A  variety  of  controversial  interpretations  (Gosline,  1963a; 
Monod,  1968;  Rosen  and  Patterson,  1969)  have  been  advanced 
concerning  supposed  sequences  effusions  and  deletions  of  bony 
elements  in  gadiform  tails.  This  particular  use  of  caudal  fin 
structure  in  phylogeny  has  yet  to  be  proven,  as  few  hypotheses 
have  been  verified  or  falsified. 

For  purposes  of  classification  within  the  order,  at  least  four 


COHEN:  GADIFORMES 


261 


BREGMACEROTIDAE 


EUCLICHTHYS 


MERLUCCIUS 


10  20  30  40  50 

BRANCHED  CAUDAL  RAYS 

Fig.  134.     Branched  caudal  rays  in  seven  groups  of  gadiform  fishes.  Data  from  Fahay  and  Markle  (this  volume)  and  onginal. 


caudal  fin  characters  require  comment.  They  are:  1)  presence 
or  absence  of  a  caudal  fin;  2)  number  of  hypurals;  3)  relationship 
between  branched  caudal  rays,  hypurals,  and  procurrent  caudal 
rays;  4)  presence  or  absence  of  X- Y  bones. 

Although  vestiges  of  a  caudal  fin  are  sometimes  found  in  a 
few  macrourid  species,  it  is  essentially  absent  from  all  of  them. 
The  same  is  true  of  the  merlucciid  genus  Lyconus  and  also 
Steindachneria.  Loss  of  a  caudal  fin  has  certainly  occurred  two 
times  and  perhaps  more. 

The  number  of  hypurals  is  a  useful  systematic  character.  There 
are  almost  always  4  or  5  in  morids  and  Melanonus,  and  almost 
always  2  in  gadids,  Merliiccius.  Bregmaceros,  and  Muraenolepis: 
Euclichthys  has  4,  nearly  fused  to  2. 1  consider  the  lower  number 
to  be  an  advanced  character;  the  study  of  developmental  series 
has  verified  this  interpretation  for  Raniceps  at  least  (Dunn  and 
Matarese,  this  volume).  Certainly  the  loss  of  hypurals,  whether 
through  deletion  or  fusion  has  occurred  several  times  in  gadi- 
forms. 

The  evolutionary  complexity  of  the  caudal  fin  in  gadiforms 
is  particularly  apparent  when  considering  the  numbers  of  dif- 
ferent kinds  of  caudal  fin  rays  (Figs.  132-134  and  Fahay  and 
Markle,  this  volume.  Table  76).  Morids  in  general  have  caudal 
fins  that  are  small  and  probably  of  reduced  importance  in  pro- 
pulsion, and  which  1  interpret  as  a  derived  state;  they  also  have 
generally  fewer  total  rays,  which  Fahay  and  Markle  (this  volume) 
consider  an  ancestral  state,  and  unbranched  rays  that  tend  to 
be  short  and  contribute  little  to  overall  caudal  fin  size;  yet, 
morids  have  4-5  hypurals.  Melanonus  also  has  a  weakly  de- 
veloped caudal  fin  but  has  4-5  hypurals  and  many  rays.  Gadine 
fishes  on  the  other  hand,  have  well-developed  caudal  fins  with 
many  rays,  both  branched  and  unbranched,  but  have  only  2 
hypurals.  Gadines  are  in  general  good  swimmers,  and  one  of 
the  most  active  of  all,  Pollachius  vtrens,  has  the  most  total  caudal 
fin  rays  (70  in  one  specimen)  of  any  gadiform  fish.  (Sluggish 
fishes  like  the  lotines,  Brosme  and  Lota,  also  have  numerous 
caudal  fin  rays  but  have  rounded  caudal  fins  and  must  swim  in 
a  very  different  way,  probably  using  the  caudal  fin  as  an  exten- 
sion of  the  body  rather  than  as  an  oar.)  Although  numbers  of 
different  kinds  of  fin  rays  may  prove  useful  in  taxonomy,  the 
relationship  of  branched  to  unbranched  or  total  caudal  fin  rays 
is  variable  and  has  limited  apparent  value  in  the  present  context. 

Many  gadiform  fishes  have  in  their  caudal  fin  skeletons  a  pair 
of  bone  splints  resembling  neural  and  haemal  spines.  These 
structures  have  been  mentioned  in  the  literature  as  accessory 


bones  or  X  and  Y  bones  and  have  been  interpreted  as  modified 
relict  pterygiophores  or  detached  neural  and  haemal  spines  whose 
centra  have  been  lost  (Rosen  and  Patterson,  1969).  1  agree  with 
Markle  (1982)  that  the  absence  in  any  gadiform  of  X  and  Y 
bones  is  a  derived  character. 

Dorsal  and  anal  fins.— Gadiform  fishes  have  1,  2,  or  3  external 
dorsal  fins  and  1  or  2  external  anal  fins.  The  number,  size,  and 
location  of  these  fins  have  been  used  for  hundreds  of  years  to 
characterize  groups  of  species.  Prior  to  the  recognition  of  Mor- 
idaeasa  distinct  family  (Svetovidov,  1937),  convergence  in  this 
character  was  not  recognized;  most  ichthyologists  lumped 
gadids  and  morids  with  similar  fin  patterns. 

Svetovidov  ( 1948)  assumed  on  functional  grounds  that  a  sin- 
gle dorsal  and  single  anal  is  the  primitive  condition  and  arranged 
the  gadid  genera  in  a  transition  series  based  on  increasing  num- 
ber of  fins  and  the  distance  of  their  separation  from  each  other. 
His  hypothesis  is  supported  by  the  presence  in  all  gadiforms  of 
a  single,  continuous,  postanal  series  of  pterygiophores,  present 
even  over  areas  that  lack  fin  rays.  Complete  or  partial  division 
of  the  exterior  fin  has  occurred  several  times,  for  example  in 
the  gadines,  Euclichthys.  Merluccius.  and  in  the  morid  genera 
Mora,  Halargyreus.  Lepidion,  Laemonema.  and  Tripterophycis. 

Although  only  a  few  gadiforms  have  a  single  dorsal  fin,  the 
condition  has  a  broad  taxonomic  distribution;  examples  are  the 
gadid  Brosme.  the  merlucciid  Lyconus.  Melanonus,  and  the  ma- 
crouroidine  rattails.  Nearly  all  gadiforms  have  2  or  3  dorsals, 
but  even  in  those  with  3,  there  are  only  two  series  of  pteryg- 
iophores. From  fewer  to  more  dorsals  would  seem  to  be  a  rea- 
sonable transition  series.  But  it  certainly  has  occurred  more  than 
once,  even  within  Gadidae,  as  Markle  (1982)  has  demonstrated. 

Pectoral  radials. —  Mosx  gadiforms  have  five  pectoral  radials. 
Muraenolepis  has  more;  Bregmaceros  has  fewer;  both  are  in- 
terpreted as  derived  conditions. 

First  neural  spine.  —  Many  gadiforms  have  the  first  neural  spine 
closely  adpressed  to  the  occipital  crest.  I  take  this  as  a  derived 
character.  Muraenolepis  has  a  free  spine,  but  it  is  modified  by 
the  presence  of  a  prominent  wing-shaped  enlargement  extending 
on  either  side  of  the  occipital  crest. 

Olfactory  lobes.  — In  his  classical  monograph  on  the  Gadidae, 
Svetovidov  (1948)  discussed  the  position  of  the  olfactory  lobes 


262 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  135.     Dorsal  view  of  cranium  in  three  genera  of  gadiform  fishes;  left.  Rhinocephalus  planiceps:  center,  Palaeogadus  intergerinus;  right, 
Merluccius  merluccius.  From  Fedotov  ( 1 976). 


of  the  brain  and  used  their  advanced  position,  adjacent  to  the 
nasal  capsule,  as  his  primary  character  for  defining  the  Gadi- 
formes.  This  is  a  derived  character,  which  has  been  found  also 
in  cyprinids,  galaxiids,  and  mormyrids.  Svetovidov  noted  that 
the  olfactory  lobe  is  located  in  an  intermediate  position  in  the 
gadid  Raniceps.  A  posterior  location  of  the  lobe  was  subse- 
quently recorded  in  Melanonus  and  several  macrourids  and  an 
intermediate  location  in  merlucciids,  Steindachneria,  the  gadid 
Raniceps,  and  two  macrourids  (Marshall  and  Cohen,  1973). 
Svetovidov  ( 1 969)  pointed  out  the  size  dependent  nature  of  this 
character,  especially  in  Merluccius  (which  I  have  verified  in  M. 
bilinearis  and  M.  productus).  Further  investigation  is  required, 
especially  in  species  that  mature  at  small  sizes. 

V-shaped  crest  on  skull.— As  long  ago  as  1903b  Regan  noted 
the  shared  presence  in  Merluccius  and  Macruronus  of  prominent 
V-shaped  ridges  on  the  frontals,  which  converge  on  the  supra- 
occipital  crest.  These  structures  have  subsequently  been  found 
in  the  extinct  genera  Rhinocephalus  and  Palaeogadus  (Fig.  135) 
as  well  as  in  some  fossil  percopsiforms  (Rosen  and  Patterson, 
1969)  and  are  present  in  varying  degrees  in  Lyconus  and  Stein- 
dachneria. 

Groups  and  Their  Relationships 

In  this  section  I  briefly  discuss  those  taxonomic  units  that  I 
think  require  consideration  and  explain  as  best  possible  the 
reasons  for  their  placement  on  Fig.  1 36. 

"Protocodus"  is  an  unnamed  species'  from  the  Paleocene  of 
Greenland  (discussed  by  Rosen  and  Patterson,  1969  and  Fe- 


'  The  name  "Protocodus"  is  used  as  a  designation  of  convenience 
and  does  not  have  formal,  nomenclatural  significance. 


dotov,  1 976;  I  too  have  examined  it),  which  is  the  oldest  known 
non-otolith  gadiform.  It  has  a  number  of  characters  that  may  be 
interpreted  as  primitive  for  the  group,  including  five,  slender, 
well-separated  hypurals,  X-Y  bones,  numerous  procurrent  rays, 
and  a  V-shaped  ridge  on  the  frontals.  It  has  a  dorsal  and  anal 
fin  configuration  much  like  that  of  Merluccius  (Rosen  and  Pat- 
terson, 1969). 

Muraenolepis  is  a  highly  distinctive  genus  with  four  or  more 
species.  It  has  such  primitive  characters  as  a  single  anal  and 
long-based  second  dorsal  fin,  a  dermal  basibranchial  plate  (Ro- 
sen and  Patterson,  1969),  the  similarity  of  the  lower  hypurals 
to  pterygiophores  and  to  caudal  fin  elements,  and  a  free  first 
neural  spine.  Derived  characters  include  12-14  pectoral  radials, 
a  single  epural,  first  dorsal  fin  a  single-rayed  anteriorly  placed 
filament,  vertical  fins  confluent  around  the  tail,  an  oblique  pat- 
tern of  squamation,  and  modifications  of  the  first  neural  spine. 
Muraenolepis  is  not  obviously  related  to  any  other  gadiform 
and  appears  to  represent  an  ancient  lineage. 

Bregmaceros  is  another  distinctive  genus  with  no  obvious 
close  relatives.  Like  Muraenolepis  it  retains  a  dermal  basi- 
branchial plate,  but  this  is  a  primitive  character,  as  is  possession 
of  a  uroneural  and  a  set  of  X-Y  bones  in  the  tail.  Derived 
characters  include  the  conjunction  of  the  first  neural  spine  with 
the  occipital  crest,  a  large  consolidated  hypural  plate  supporting 
many  branched  rays,  a  unique  lateral  line  system,  only  two 
pectoral  radials,  and  a  long  dorsal  ray  on  top  of  the  head.  The 
tropical  pelagic  habitat  of  these  fishes  is  also  different  from  that 
of  any  other  gadiform.  If  fusion  of  the  first  neural  spine  with 
the  occipital  crest  has  occurred  only  a  single  time,  then  Breg- 
maceros must  have  originated  after  Rhinocephalus. 

Rhinocephalus  is  an  Eocene  fossil,  the  skull  of  which  has  been 
described  in  some  detail  and  compared  with  other  gadiforms 
by  Rosen  and  Patterson  ( 1 969).  They  mention  and  illustrate  a 


COHEN:  GADIFORMES 


263 


RECENT. 


PLEISTOCENE. 


PLIOCENE. 


MIOCENE. 


OLIGOCENE. 


EOCENE. 


PALEOCENE. 


"PROTOCODUS" 


Fig.  136.     Phylogenetic  bush  showing  hypothetical  inter-relationships  among  gadirorm  fishes.  Beginning  of  soHd  Unes  based  on  fossils,  not 
including  otoliths  or  scales. 


V-shaped  indge  on  the  frontals  and  also  lateral  flanges  on  the 
rear  of  the  skull  that  characterize  gadines  and  at  least  some 
morids.  They  write,  "The  skull  roof  of  Rhinocephalns  shows 
many  features  common  to  morids,  merlucciids,  gadids.  and 
macrourids  .  .  .  ."  In  addition,  the  first  neural  spine  is  free  from 
the  supraoccipital  crest. 

Eucltchthys  (Fig.  137),  represented  by  a  single  South  Austra- 
lian and  New  Zealand  species,  was  incorrectly  placed  in  Moridae 
but  removed  by  Svetovidov  (1969),  who  pointed  out  some  sim- 
ilarities to  Macrouridae.  Enclichlhys  can  not  be  placed  in  any 
currently  recognized  family.  It  has  a  free  first  neural  spine,  which 
may  indicate  an  origin  prior  to  Palaeogadus.  lacks  an  otophysic 
connection,  has  four  hypurals  nearly  fused  to  two,  and  in  two 
specimens  has  only  one  of  the  X-Y  bones.  As  in  morids,  which 
are  more  specialized  than  macrourids  and  could  not  have  given 
rise  to  them,  Eitclichthys  has  an  asymmetrical,  rather  reduced 
caudal  fin.  Perhaps  this  curious  fish  is  a  modem  representative 
of  a  macrourid  progenitor. 

Macrouroidinae  is  represented  by  two  small  genera  and  has 
been  treated  both  as  a  subfamily  of  Macrouridae  (Marshall, 
1973)  and  a  separate  family  (Okamura,  1970a).  It  has  single 
dorsal  and  anal  fins  and  a  number  of  distinctive  features  in  the 
head  skeleton  and  may  represent  the  most  primitive  tail-less 
macruroid. 

Macrourinae-Trachyrincinae,  which  may  well  constitute  two 


quite  separate  groups,  has  20-25  genera  and  contains  more  than 
half  of  all  gadiform  species  (Okamura,  1970a;  Marshall,  1973). 
The  caudal  fin  is  absent  in  most,  vestigial  in  a  few;  the  first 
neural  spine  is  free,  and  there  is  no  V-shaped  ridge.  Eggs  of  the 
few  species  for  which  information  is  available  have  a  distinctive 
hexagonal  pattern;  many  species  have  light  organs. 

Bathygadinae,  with  two  genera,  differs  from  other  macrourids 
in  having  a  large,  terminal  mouth,  dorsal  rays  longer  than  anal 
ones,  and  in  a  variety  of  other  ways  summarized  by  Okamura 
(1970a),  who  interprets  most  of  the  bathygadine  characters  as 
primitive  ones.  Differences  in  functional  morphology  between 
bathygadines  as  pelagic  feeders  and  macrourines  as  benthic  to 
benthopelagic  feeders  have  been  described  by  McLellan  (1977). 

Melanonus  has  two  meso-to-bathypelagic  species  formerly 
placed  in  Moridae,  where  they  do  not  belong  as  they  lack  an 
otophysic  connection,  have  a  single  dorsal  fin,  and  have  lost  the 
X-Y  bones.  Otherwise,  they  seem  similar  to  Moridae.  The  first 
neural  spine  is  joined  to  the  occipital  crest,  suggesting  an  origin 
after  Rhinocephalus.  A  separate  family  was  proposed  by  Mar- 
shall (1965). 

Moridae  consists  of  12-15  genera,  some  highly  diverse,  and 
all  characterized  by  possession  of  an  otophysic  connection,  4  or 
5  hypurals,  X-Y  bones,  a  joined  first  neural  spine,  and  distinctive 
otoliths;  many  species  have  light  organs.  Morids  probably  di- 
verged from  the  main  Rhinocephalus-Palaeogadus-Merluccius 


264 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES- AHLSTROM  SYMPOSIUM 


Fig.  137.    Euclichlhys  polynemus,  holotype.  From  McCulloch  (1926). 


line  after  fusion  of  the  neural  spine  and  at  least  some  of  their 
evolution  is  in  parallel  with  the  gadids. 

Palaeogadits  is  a  well-known  Eocene  fossil  genus  in  which  the 
V-shaped  crest  has  been  retained,  but  specializations  include  a 
joined  first  neural  spine  and  only  two  hypurals.  It  is,  in  fact, 
very  similar  to  modem  Merluccius.  DaniPchenko  (1950),  who 
reviewed  Palaeogadus.  believed  that  it  gave  rise  independently 
to  Lotinae  and  Gadinae  as  well  as  to  Merluccius. 

Phycinae,  as  recently  modified  by  Markle  (1982),  is  presently 
included  in  the  family  Gadidae.  Fahay  and  Markle  (this  volume) 
would  like  to  escort  it  out.  An  early  Oligocene  fossil  genus, 
Eophycis  (Jerzmanska,  1968)  has  been  suggested  as  a  precursor 
of  Phycts  and  Urophycis,  and  probably  arose  independently  of 
other  gadid  subfamilies,  which  supports  Fahay  and  Markle's 
position. 

Lotinae  is  a  gadid  subfamily  that  I  mainly  leave  to  Fahay  and 
Markle  and  Dunn  and  Matarese.  I  note,  however,  Mujib's  ( 1 967) 
conclusion  based  on  cranial  osteology  that  Lotinae  could  have 
arisen  from  Merlucciinae.  Lotines  have  no  V-shaped  crest  but 
retain  X-Y  bones.  Hypurals  are  two,  the  first  neural  spine  is 
joined  to  the  occipital  crest,  and  there  are  more  branched  rays 
than  in  any  other  gadid. 

Gadinae  has  about  a  dozen  genera,  all  of  which  have  three 
external  dorsal  and  two  external  anal  fins  and  a  large  caudal, 
even  though  there  are  only  two  hypurals.  Derived  characters 
include  fused  frontals,  absence  of  X-Y  bones,  and  a  joined  neu- 
ral spine;  Fahay  and  Markle  and  Dunn  and  Matarese  (this  vol- 
ume) give  more. 

Merluccius,  with  about  a  dozen  closely  related  species  (Inada, 
1981b),  has  been  treated  as  the  type  of  a  separate  family  or  as 
a  subfamily  of  Gadidae.  Primitive  characters  include  a  V-shaped 
ridge  and  X-Y  bones.  Advanced  ones  are  the  joined  first  neural 
spine  and  the  reduced  number  of  hypurals.  Merluccius  appears 
to  be  the  modem  representative  of  a  lineage  commencing  with 
"Protocodus"  and  extending  through  Rhinocephalus  and  Pa- 
laeogadus. which  it  closely  resembles  (Rosen  and  Patterson, 
1969). 

Macruronus,  which  has  three  nominal  species  found  in  tem- 
perate waters  of  the  southern  hemisphere,  is  basically  a  Mer- 
luccius wiih  a  much  reduced  caudal  fin.  I  mention  it  here  because 
it  has  been  referred  incorrectly  to  Macrouridae  and  considered 
by  some  to  be  a  link  between  that  family  and  Merluccius. 

Lyconus.  with  several  pelagic  oceanic  species,  is  probably  re- 


lated to  Merluccius.  It  lacks  a  caudal  fin  and  has  a  single  dorsal 
fin. 

Steindachneria,  is  a  monotypic  tropical  western  Atlantic  ge- 
nus with  luminescent  organs,  a  wide  separation  between  the 
anus  and  urogenital  openings,  and  no  caudal  fin.  It  has  been 
placed  in  Macrouridae  and  also  considered  a  separate  family 
(Marshall  and  Cohen,  1973).  It  may  be  closer  to  Merluccius 
than  to  any  other  known  gadiform. 

Classification 

How  best  to  classify  gadiforms  for  working  purposes  in  a  way 
that  approximates  their  possible  phylogenetic  relationships  is 
diflicult  because  the  existence  of  fossils,  which  appears  to  help 
indicate  lineages,  creates  logical  traps  for  the  classifier.  The  fol- 
lowing arrangement,  unfortunately  based  on  gaps  for  some  groups 
and  on  a  continuum  for  others,  is  an  interim  suggestion  for 
further  testing. 

Euclichthys  is  accorded  family  status  for  the  first  time  because 
it  can  not  be  placed  in  any  gadiform  family.  Gadidae  is  restricted 
to  the  gadines,  and  Lotidae  and  Phycidae  are  recognized  at  the 
full  family  level  (family  group  names  for  the  latter  two  date  at 
least  from  Goode  and  Bean,  1 896),  because  available  evidence 
indicates  an  independent  origin  from  Palaeogadus  for  each  of 
the  three  groups.  If  merlucciids  were  reduced  to  subfamily  rank 
and  placed  with  gadines,  lotines,  and  phycines  in  a  more  inclu- 
sive family  Gadidae,  then  consistency  would  require  the  inclu- 
sion of  at  least  two  other  well-defined  apparent  derivatives  of 
the  Rhinocephalus- Palaeogadus- Merluccius  stem,  Moridae  and 
Melanonidae.  In  the  present  instance  I  believe  that  splitting  is 
more  useful  than  lumping. 

Suborder  Muraenolepoidei 
Family  Muraenolepidae 
Suborder  Bregmacerotoidei 
Family  Bregmacerotidae 
Suborder  Macrouroidei 
Family  Euclichthyidae 
Family  Macrouridae 
Subfamily  Macrouroididae 
Subfamily  Trachyrincinae 
Subfamily  Macrourinae 
Subfamily  Bathygadinae 
Suborder  Gadoidei 


COHEN:  GADIFORMES 


265 


Family  Merlucciidae 
Subfamily  Merlucciinae  (including  "Protocodus,"  Rhi- 
nocephalus.  Palaeogadus,  Merluccius.  Macruronus, 
and  Lyconus) 
Subfamily  Steindachneriinae 
Family  Gadidae 
Family  Lotidae 


Family  Phycidae 
Family  Moridae 
Family  Melanonidae 

Life  Sciences  Division.  Los  Angeles  County  Museum  of 
Natural  History,  900  Exposition  Boulevard,  Los 
Angeles,  California  90007. 


Gadiformes:  Development  and  Relationships 
M.  P.  Fahay  and  D.  F.  Markle 


AS  treated  herein,  the  Gadiformes  includes  about  63  genera 
and  400+  species  (Nelson,  1976)  divided  into  eight  fam- 
ilies (Gosline,  1968;  Marshall  and  Cohen,  1973);  (but  see  Cohen, 
this  volume).  They  are  primarily  marine  with  familial  distri- 
bution "centers"  as  follows:  Muraenolepididae— high  latitudes, 
southern  hemisphere;  Bregmacerotidae  — tropical  and  sub- 
tropical, world-wide;  Melanonidae— tropical  and  sub-tropical, 
world-wide;  Moridae— world-wide;  Macrouridae— deep  sea, 
world-wide;  Steindachneriidae— tropical  W.  Atlantic;  Merlucci- 
idae—mid-latitudes,  both  hemispheres;  and  Gadidae— high  lat- 
itudes, northern  hemisphere  with  minor  freshwater  and  south- 
em  hemisphere  components. 

Meristic  characters  of  genera  within  each  family  are  presented 
in  Table  72  (except  that  macrourid  characters  will  be  found  in 
Table  75).  Gadiforms  characteristically  have  relatively  high  ver- 
tebral counts,  with  caudal  centra  outnumbering  precaudal  cen- 
tra, usually  by  a  wide  margin.  The  first  two  centra  lack  ribs  and 
parapophyses.  Vertical  fins  have  numerous  rays  and  long  bases, 
with  posterior  dorsal  and  anal  rays  separate  from  caudal  fin  rays 
except  in  Miiraenolepis  and  macruronines.  Pectoral  fins  are  typ- 
ically high  on  the  body  and  pelvic  fins  typically  thoracic  or 
jugular  in  position.  Mental  barbels  are  found  in  many  genera 
and  mouth  position  ranges  from  terminal  to  inferior. 

Present  State  of  KnowLedge  and 

Characters  of  Early  Life 

History  Stages 

Literature  on  gadiform  eggs  and  larvae  is  heavily  weighted 
towards  gadids  and  merlucciids.  within  which  the  commercially 
important  gadines  and  Merluccius  have  received  most  attention. 
Gadine  larvae  were  among  the  first  marine  fish  larvae  to  be 
described.  In  fact.  G.  O.  Sar's  discovery,  early  in  the  1860's, 
that  cod  eggs  and  larvae  were  pelagic,  helped  initiate  fisheries- 
oriented  ichthyoplankton  surveys.  In  addition  to  their  com- 
mercial importance,  gadines  and  Merluccius  are  found  in  shelf 
waters  where  their  early  stages  are  more  accessible  than  those 
of  other  gadiforms  which  are  largely  residents  of  slope  and  oceanic 
waters. 

Published  descriptions  of  gadiform  early  life  history  stages 
are  listed  in  Table  73.  We  especially  note  the  seminal  work  on 
young  gadids  done  by  Johannes  Schmidt  in  the  early  1900's. 
Although  he  stressed  pigment  patterns  over  other  develop- 


mental features.  Schmidt  was  one  of  the  first  to  look  at  several 
species  in  a  systematic  fashion. 

In  the  following  review,  we  summarize  gadiform  characters 
in  brief  family  synopses  as  well  as  through  a  limited  survey  of 
the  ontogeny  of  selected  characters.  Our  purposes  are,  respec- 
tively, to  point  out  what  appear  to  be  easily  observed  diagnostic 
early  life  history  characters  and  to  contribute  to  discussions  of 
gadiform  phylogeny. 

Gadiformes.— The  gut  of  gadiform  larvae  coils  early  in  ontogeny 
and  combined  with  a  tapering  postanal  region  and  rounded 
head,  contributes  to  an  overall  tadpole-like  appearance.  These 
features  are,  in  part,  a  reflection  of  vertebral  and  vertical  fin  ray 
elements  (Table  72)  and  are  not  diagnostic.  Although  it  has  not 
been  documented  in  all  families  and  is  not  always  easily  ob- 
served, yolk-sac  and  first-feeding  gadiform  larvae  have  an  anus 
that  exits  laterally  through  the  finfold  rather  than  medially  as 
is  usual  in  teleost  larvae.  Some  secondary  caudal  rays  develop 
before  some  primary  in  forms  with  a  caudal  fin. 

In  Table  74  we  summarize  some  developmental  features  of 
each  family.  A  rather  widespread  trend  is  for  the  pelvic  fin  to 
be  the  earliest  forming  fin.  There  does  not  seem  to  be  any  char- 
acter unique  or  diagnostic  for  young  gadiforms.  The  features  of 
body  shape,  anus  morphology  and  pelvic  fin  development  in 
combination  with  specific  familial  characters  appear  to  be  the 
most  useful  for  initial  identification.  Transformation  is  gradual 
and  direct  with  no  striking  changes  in  ontogeny. 

Muraenolepididae.  —  \  single  planktonic  juvenile  (see  discus- 
sion of  planktonic  juveniles  below)  of  Muraenolepis  sp.  is  shown 
in  Fig.  138 A.  The  distinctive  first  dorsal  fin,  composed  of  one 
or  two  rays,  the  confluent  vertical  fins,  meristic  characters  (Ta- 
bles 72  and  76),  chin  barbel,  restricted  gill  opening  and  capture 
locality  (53°48.7'S,  38°18.7'W)  preclude  all  other  teleosts  and 
agree  with  characters  described  for  Muraenolepis  (Svetovidov, 
1948).  The  lateral  premaxillary  spines  (Fig.  138A)  were  not 
shown  in  a  schematic  illustration  of  an  early  Muraenolepis  (North 
and  White.  1982)  or  in  larvae  described  by  Efremenko  (1983b) 
and  are  not  reported  for  adults.  It  is  possible  that  they  are  not 
found  in  larvae  of  all  species  of  Muraenolepis,  but  for  present 
purposes  we  consider  them  a  unique  and  diagnostic  larval  spe- 
cialization of  the  family. 


266 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  72.  Meristic  Characters  in  Gadiformes.  (See  Table  75  for  characters  of  the  Macrouridae.)  Characters  of  the  caudal  fin  are  contained 
in  Table  76.  "Number  of  species"  includes  number  of  nominal  species  followed  by  number  surveyed  for  meristic  characters.  Primary  sources  of 
data:  Gunther,  1887;  Goode  and  Bean,  1896;  Ehrenbaum,  1905-1909;  Thompson,  1916;  Norman,  1930;  D'Ancona,  1933a;  Parr,  1946;  Jensen, 
1948;  Svetovidov,  1948;  Koefoed,  1953;  Andriyashev,  1954;  Rass,  1954;  Smith,  1961;  Scott,  1962;  Lmdberg  and  Legeza,  1969;  Leim  and  Scott! 
1966;  Templeman,  1968;  Fitch  and  Barker,  1972;  Miller  and  Lea,  1972;  Hart,  1973;  Inada  and  Nakamura,  1975;  Brownell,  1979;  Cohen,  1979; 
Cohen  and  Russo,  1979;  Inada,  1981a;  Inada,  1981b;  Matarese  et  al.,  1981;  Yabe  et  al.,  1981;  Demir,  1982;  Markle,  1982;  Fahay,  1983;  Paulin, 
1983. 


Family 
genus 

Number 

of 
species 

Prc- 

caudal 

Vertebrae 
Caudal 

Total 

Fin  rays 

D, 

Dj 

D, 

D  total 

A, 

A. 

A  total 

Pelvic 

Pectoral 

Muraenolepididae 

Muraenolepis  4  +  /1         20-21     46-49      67-69 

Bregmacerotidae 

Bregmaceros  9  +  /7         13-14     32-42      43-59 


127-141 


34-65 


129-142   98-112        98-112    4    37-38 

35-66   42-69         42-69    5-7'   16-21 
or 
15-21   27-31 


Melanonidae 

Melanonus 

111 

13 

47 

58-62 

72-78 
or 
5-8 

59-70 

72-78 

50-58 

50-58 

5-7 

10-16 

Fam.  Incertae  sedis 

Euclichthys 

?/l 

15 

55 

70 

15 

74 

89 

15 

77 

92 

5 

20 

Moridae 

Antimora 

2/2 

24-25 

33-35 

57-61 

4-7 

48-56 

54-60 

36-49 

36-49 

5-7 

17-25 

Auchenoceros 

1/1 

11 

37 

46-49 

1-2 

13 

51 

65  (+10) 

62-82 

62-82 

2' 

23 

(+10) 

(holo- 
type) 

Brosmiculus 

1/1 

17 

33-34 

50-51 

58  or 
10 

53-56 

58 
63-66 

56-62 

56-62 

5 

— 

Eretmophorus 

1/1 

— 

— 

— 

4-5 

66-77 

70-82 

64-73 

64-73 

5 

22 

Gadella 

1/1 

— 

— 

50 

9-12 

55-64 

64-74 

56-67 

56-67 

6-7 

20-25 

Halargyreus 

1/1 

— 

30-35 

51-58 

6-8 

47-60 

53-67 

17-26 

21 

-29 

39-53 

5 

17-20 

Laemonema 

147/14 

15-17 

42-45 

50-63 

5-6 

48-75 

53-80 

45-72 

45-72 

(1)2(3) 

15-26 

Lepidion 

l->/l 

17-18 

42-45 

54-63 

4-7 

49-62 

54-68 

40-54 

40-54 

(5)  6-8 

17-23 

Loiella 

6?/6 

13-15 

27-34 

41-50 

5-8 

46-69 

51-73 

42-61 

42-61 

7-9 

19-26 

Mkroteptdium 

1/1 

11-12 

33-34 

44-46 

7-9 

39-42 

46-51 

39-42 

39-42 

2 

19-22 

Mora 

1/1 

15 

35 

50-54 

7-11 

42-53 

49-60 

16-22 

15 

-22 

30-44 

5-6 

18-25 

Physiculus 

127/12 

12-16 

34-42 

48-59 

7-11 

44-71 

53-79 

43-79 

43-79 

(3)  5-7 

20-28 

Pseudophyas 

3/3 

— 

— 

42-51 

8-14 

40-63 

51-67 

39-68 

39-68 

5-6 

19-27 

Rhynchogadus 

1/1 

_ 

_ 

_ 

5 

50-56 

55-61 

44-49 

44-49 

7 

20-22 

Salilota 

1/1 

— 

— 

— 

9-11 

56 

65-67 

57 

57 

— 

_ 

Svelovidovia 

2/2 

14-15 

43 

57-58 

5-7 

57-63 

62-70 

56-62 

56-62 

10-11 

17-18 

Tripterophycis 

7/2 

- 

- 

67-72 

4-7 

12-17 

29- 

-39 

48-58 

95-112 

95-112 

5 

15-20 

Gadidae  (Lotinae) 

Brosme 

1/1 

19-21 

44-46 

63-66 

85-108 

85-108 

62-77 

62-77 

5 

22-24 

Lola 

1/1 

23-26 

37-39 

59-66 

9-16 

65-93 

75-108 

63-85 

63-85 

6-8 

18-21 

Molva 

3/3 

25-36 

37-48 

63-84 

10-16 

61-85 

74-98 

57-82 

57-82 

6-7 

18-21 

Gadidae  (Phycinae) 

Ciliata 

2/2 

(12) 
13-15 

30-34 

44-48 

1  + 

45-55 

46-56  + 

40-46 

40-46 

7' 

15-17 

Enchelyopus 

1/1 

15-17 

38-39 

49-55 

1  + 

45-52 

46-53  + 

34-49 

34-49 

5-6  (7)' 

15-19 

Gaidropsarus 

14/14 

13-17 

32-36 

44-53 

1  + 

45-70 

46-71  + 

38-60 

38-60 

5-9' 

14-24 

Phycis 

3/3 

15-16 

or 
18-19 

28-29 

or 
32-35 

8-11 

54-65 

63-70 

47-65 

47-65 

(2)3' 

15-19 

Raniceps 

1/1 

11 

33-34 

44-45 

3 

61-67 

64-70 

55-61 

55-61 

6 

21-22 

Vrophycis 

7/7 

13-17 

30-37 

44-52 
(56,  57) 

8-13 

43-68 

53-78 

40-58 

40-58 

2- 

15-18 

Gadidae  (Gadinae) 

Arclogadus 

2/2 

19-22 

36-40 

54-62 

10-16 

15-24 

19- 

-25 

17-24 

18- 

-25 

6-7 

17-23 

Boreogadus 

27/2 

18-20 

35-39 

49-58 

9-16 

12-19 

16- 

-25 

49-55 

14-23 

18- 

-24 

39-44 

6 

18-21 

Eleginus 

2/2 

21-24 

37-41 

55-64 

11-16 

14-24 

18- 

-24 

19-27 

18- 

-26 

6 

18-22 

Gadiculus 

1/1 

— 

— 

39-43 

9-12 

11-17 

15- 

-18 

12-19 

15- 

-19 

6 

14-15 

Gadus 

77/3 

18-22 

31-37 

49-58 

10-17 

11-24 

10- 

-22 

45-62 

16-27 

12- 

-25 

35-48 

6-7 

19-22 

FAHAY  AND  MARKLE:  GADIFORMES 


267 


Table  72.    Continued. 


Mela  nogra  mm  us 

Merlangius 

Microgadus 

Micromesislius 
Pollachius 
Theragra 
Trisoplerus 

Merlucciidae 

Merluccius 

Macruronus 

Lyconus 

Steindachneriidae 
Steindachneria 


Number 

of 
species 

Venebrae 

Fin  rays 

Family 
fienus 

Prc- 
caudal 

Caudal 

Tolal 

D, 

Dj 

D, 

D  lolal 

A, 

A; 

A  total 

Pelvic 

Pectoral 

1/1 

2/1 
2/2 

2/2 
2/2 
2/1 
3/3 


1/1 


19-21  33-36  52-57  14-18 

23(?)        -  53-57  12-17 

17-22  34-38  53-60  9-15 

24-26  30-33  54-60  11-14 

20-23        32  52-56  11-15 

18-20  31-34  48-52  10-14 

-            -  44-55  11-16 


12/12   21-29  24-31 
3/3   20-21'  58-60' 
2/1 


48-58 
78-81' 


13 


1,7-12 
8-11 
90  + 

8-12 


19-26 

18-25 
15-21 

10-15 
16-24 
12-19 
16-28 


34-45 
105-120 


123  + 


19-24 
19-22 
16-24 

21-27 
15-24 
14-23 
16-27 


56-67 


21-28 
28-38 
(12) 
18-29 
33-41 
23-34 
15-24 
25-36 


35-46 

86-105 
90+  65  + 


20-25 
20-25 
16-28 

22-30 
16-24 
15-23 
17-30 


45-53 


10-11  +  113 


86-105 
65  + 


6-7 

6 
6-7 

6 

6 

6-7 

6 


(6)  7  (8) 
9-10 
10 


19-21 
19-20 

(16) 
18-19 
18-23 
17-20 
18-21 

(13) 
17-19 


12-18 

14-18 

13 


14-15 


'  Four  rays  in  larvae. 
-  Three  rays  in  larvae. 
^  n  =  2  (A/,  novaezelandiae). 


Bregmacerotidae.  — Larval  and  juvenile  bregmacerotids  appear 
distinctive  in  the  early  acquisition  of  a  cephalic  dorsal  fin  ray. 
Larvae  have  been  described  (Table  7  3 ),  but  eggs  are  undescribed. 
Characters  are  reviewed  in  this  volume  by  Houde. 

Melanonidae. —Eggs,  larvae  and  young  stages  have  not  been 
previously  described  for  melanonids  (Cohen,  1973).  Early  stages, 
however,  are  moderately  abundant  in  some  oceanic  collections. 
In  the  smallest  specimens  seen  by  us  (ca.  15  mm  SL)  the  fins 
are  all  formed  and  they  have  the  general  body  shape  of  adults 
(Fig.  1 38B).  Notable  features  of  this  stage  are  the  small  eye,  dark 
peritoneum  and  distinctive  caudal  fin. 

Monrfae.— Considering  the  diversity  of  the  family,  very  little  is 
known  of  the  early  life  history  stages  of  morids.  Eggs  with  oil 
globules  have  been  described  for  Physiculus  dalwigkii  (De 
Gaetani,  1928),  Mora  mora  (D'Ancona,  1933a),  Physiculus  ca- 
pensis  (Brownell,  1979),  Salilota  australis  (de  Ciechomski  and 
Booman,  1981)  and  Laemonema  longipes  (Kuroda  et  al.,  1 982). 

Pelagic  juveniles  of  some  morids  have  not  yet  been  related 
to  adult  forms  and  have  been  placed  in  three  genera,  Rhyncho- 
gadus  Tortonese,  1948;  Svetovidovia  Cohen,  1973  and  Eret- 
mophorus  Giglioli,  1889.  One  of  these  forms,  S.  vitellius  (Koe- 
foed,  1953),  is  shown  in  Fig.  I38C,  D.  This  form  appears  to  be 
the  juvenile  stage  of  Laemonema.  In  our  largest  specimen,  55 
mm  SL  (MCZ  59773),  the  pelvic  fin  has  two  rays  plus  two  or 
three  remnants.  This  is  a  reduction  from  a  count  of  9-1 1  in 
smaller  specimens.  To  date  it  has  not  been  possible  to  assign 
this  form  to  a  known  adult.  A  second  type  of  Svetovidovia  is 
shown  in  Fig.  I38E.  D'Ancona  (1933a)  suspected  that  Eret- 
mophorus  kleinenbergi  was  the  young  of  Lepidion  lepidion  but 
Cohen  (1973)  apparently  was  not  convinced  of  the  relationship. 
Finally,  Rhynchogadus  Tortonese,  1948  (= Hypsirhynchus)  is  a 
pelagic  form  referrable  to  no  known  adults  and  may  also  rep- 
resent an  early  stage  of  a  species  whose  adult  form  is  known 
under  another  name. 

The  early  stages  of  morids  appear  stocky  anteriorly,  with  well 
developed  to  voluminous  pelvic  fins,  frequently  with  more  in- 
ferior than  superior  procurrent  caudal  fin  rays,  and  relatively 


voluminous  posterior  sections  of  dorsal  and  anal  fins  (see  Figs. 
138C-E,  139A-F).  Earliest  stages  may  be  difficult  to  separate 
from  some  merlucciids  and  gadines. 


Macrouhdae.— There  is  a  moderate  amount  of  early  life  history 
information  available  on  macrourids  but  considering  that  the 
family  contains  over  a  third  of  all  extant  gadiform  species  (Nel- 
son, 1976),  a  great  deal  remains  unknown.  Eggs  have  been  de- 
scribed by  Gilchrist  (1904),  Sanzo  (1933a),  de  Ciechomski  and 
Booman  (1981)  and  Grigorev  and  Serebryakov  (1981).  All  de- 
scribed macrourid  eggs  range  from  about  I  to  4  mm  (Marshall 
and  Iwamoto,  1973).  Most  are  less  than  2  mm,  have  a  single 
oil  globule  and  characteristic  honey-comb  ornamentation  on 
the  chorion  (see  Boehlert,  this  volume). 

Larvae  and  pelagic  juveniles  have  been  infrequently  described 
and  only  for  macrourines  (Table  73).  Early  ontogenetic  stages 
of  trachyrhynchine  and  macrouroidine  macrourids  are  still  not 
known  though  Johnsen  (1927)  illustrated  and  discussed  meta- 
morphosed Trachyrhyncfius  juveniles. 

Only  one  pelagic  juvenile  bathygadine  is  known  (Fig.  1408). 
The  specimen,  tentatively  referred  to  Gadomus.  can  be  recog- 
nized by  its  long  second  dorsal  (relative  to  anal)  fin  rays,  short 
interspace  between  dorsal  fins,  moderate-sized  barbel  and  fine 
jaw  teeth.  Other  important  characters  of  this  specimen  are  its 
laterally  placed  thoracic  pelvic  fins  and  paired  preopercular  skin 
flaps.  The  general  appearance  of  young  bathygadines  approaches 
that  of  morids,  with  the  lack  of  caudal  fin  and  presence  of 
pedunculate  pectoral  fins  the  obvious  differences. 

Numerous  macrourine  larvae  have  been  described.  The  spec- 
imen illustrated  as  Fig.  140C  appears  identical  to  Johnsen's 
(1927)  "AH  1"  macrourid  larva  while  Fig.  MOD  is  similar  in 
appearance  to  Merrett's  (1978)  Coryphaenoides  rupestris.  In  both 
cases  meristic  characters  agree  with  Coryphaenoides  (sensu  lato), 
but  we  are  unable  to  provide  further  identification  at  this  time. 

A  number  of  more  elongate  types  are  also  known.  A  specimen 
belonging  to  either  Cetonurus  or  Nezumia  is  shown  in  Fig.  1 4 1  A. 
A  similar  specimen,  also  with  seven  branchiostegals,  is  shown 
in  Fig.  14 IB;  its  meristic  characters,  however,  do  not  permit 


268 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  73.    Published  Descriptions  of  Early  Life  History  Stages  in  Genera  within  Eight  Gadiform  Families. 


Family 
genus 


Family 
genus 


Muraenolepididae 
Muraenolepis 


Moridae 
Auchenoceros 

Erelmophorus' 
Gadella 

Mora 


Gaidropsarus 


Physiculus 


Rhynchogadus' 

Salilota 
Svetovidovia 


Melanonidae 

Melanonus 

Bregmacerotidae 
Bregmaceros 


Gadidae  (Lotinae) 
Brosme 


Lota 


Molva 


Gadidae  (Phycinae) 

Ciliata 


Enchelyopus 


North  and  White,  1982 
Efremenko,  1983b 


Robertson,  1975a 

Crossland.  198P 

Mazzarelli,  1917 

Lo  Bianco,  1911 

Sparta,  1928 

Lo  Bianco,  1911 

DeGaetani,  1926 

D'Ancona,  1933a 

Lo  Bianco,  1911 

Mancuso,  1926 

DeGaetani,  1928 

D'Ancona,  1933a 

Pertseva-Ostroumova  and  Rass,  1973- 

Brownell,  1979 

Lo  Bianco,  1911 

Cipna,  1927 

D'Ancona,  1933a 

Weiss,  1975 

de  Ciechomski  and  Booman,  1 98 1 

Koefoed,  1953 

Fahay,  1983^ 


None 


Munro,  1950 

Clancey,  1956 

D'Ancona  and  Cavinato,  1965 

Aboussouan,  1968c 

Pertseva-Ostroumova  and  Rass,  1973 

Belyanina,  1974 

Houde,  1981 


Mcintosh,  1893 

Schmidt,  1905b 

Ehrenbaum,  1905-1909 

Rass,  1949 

Ehrenbaum,  1905-1909 

Meshkov,  1967 

Jude,  1982b 

Mcintosh,  1893 

Mcintosh  and  Masterman,  1897 

Heincke  and  Ehrenbaum,  1900 

Schmidt,  1906b,  1907b 

Ehrenbaum,  1905-1909 

D'Ancona,  1933a 

Russell,  1976 


Ehrenbaum,  1905-1909 

Dando,  1975 

Russell,  1976 

Agassiz,  1882 

Agassiz  and  Whitman,  1885 

Brook,  1890 

Ehrenbaum  and  Strodtman,  1904 

Ehrenbaum,  1905-1909 

Dannevig,  1919 

Colton  and  Marak,  1969 

Russell,  1976 


Phycis 


Raniceps 


Urophycis 


Gadidae  (Gadinae) 
Arctogadus 
Boreogadus 

Eleginus 


Gadiculus 
Gadus 


Melanogrammus 


Merlangius 


Microgadus 

Micromesistius 


Roule  and  Angel,  1930 

D'Ancona,  1933a 

Vodyanitsky  and  Kazanova,  1954 

Fives.  1970b 

Schmidt,  1905a,  1906a 

Ehrenbaum,  1905-1909 

Russell,  1976 

Bini,  1971 

Dekhnik,  1973 

Brownell,  1979 

Demir,  1982 

Markle,  1982 

Facciola,  1882 

Emery,  1886 

Manon,  1894b 

D'Ancona,  1933a 

Russell,  1976 

Heincke  and  Ehrenbaum,  1900 

Schmidt,  1907b 

Ehrenbaum,  1905-1909 

Kennedy  and  Fitzmaurice,  1969 

Russell,  1976 

Aggassiz,  1882 

Aggassiz  and  Whitman,  1885 

Hildebrand  and  Cable,  1938 

Bigelow  and  Schroeder,  1953 

Miller  and  Marak,  1959 

Barans  and  Barans,  1972 

Serebryakov,  1978 


Zvyagina,  1961 
Schmidt,  1905a,  1906a 
Rass,  1949 

Kuz-min-Karovaev,  1930 
Khaldinova,  1936 
Ponomareva,  1949 
Rass,  1949 
Mukhacheva,  1957 
Aronovich  et  al.,  1975 
Dunn  and  Vinter,  1984 
Schmidt,  1906a 
Roule  and  Angel,  1930 
Heincke  and  Ehrenbaum,  1900 
Masterman,  1901 
Schmidt,  1905a,  1906a 
Dannevig,  1919 
Uchidaet  al.,  1958 
Mukhacheva  and  Zvyagina,  I960 
Russell,  1976 
Matarese  et  al.,  1981 
Mcintosh  and  Pnnce,  1890 
Heincke  and  Ehrenbaum,  1900 
Schmidt,  1905a,  1906a 
Dannevig,  1919 
Russell,  1976 

Heincke  and  Ehrenbaum,  1900 
Schmidt,  1905a,  1906a 
Ehrenbaum,  1905-1909 
D'Ancona,  1933a 
Dekhnik,  1973 
Russell,  1976 
Booth,  1967 
Matarese  et  al.,  1981 
Schmidt,  1905a,  1906a 
D'Ancona,  1933a 
Seaton  and  Bailey,  1971 


FAHAY  AND  MARKLE:  GADIFORMES 


269 


Table  73.    Continued. 


Family 
genu'. 


Family 

senui 

Source 

Steindachneriidae 

Steindachneria 

None'' 

Macrouridae 

Ateleobrachium^ 

Gilbert  and  Burke.  1912 

Coetorhynchus 

Sanzo,  1933a 

de  Ciechomski  and  Booman,  1981 

Gilchnst,  1905 

Coryphaenoides 

Johnsen,  1921 

Merrett,  1978 

Stein,  1980b 

Grigorev  and  Serebryakov,  1 98 1 

Hymenocephalus 

Sanzo,  1933a 

Krohnius^ 

Costa,  1869 

Smitt,  1895 

Rouleand  Angel,  1930? 

Sanzo,  1933a 

"Macrouridae" 

Ehrcnbaum,  1905-1909^ 

Murray  and  Hjort,  1912 

Johnsen,  1927 

Evseenko.  1982b 

Macrourus 

Yanulov,  1962 

de  Ciechomski  and  Booman,  1981 

Efremenko,  1983a 

Malacocephalus 

Marshall,  1964 

Mesobtus 

Hubbs  and  Iwamoto,  1977 

Odontoinacrurus 

Maul  and  Koefoed,  1950 

Maul,  1951 

Koefoed,  1953 

Marshall,  1964 

ISphagehra  nchusl^ 

Backus  etal.,  1965 

Trachyrhynch  us 

Johnsen,  1927 

Pollachius 
Theragra 

Tnsoplerus 


Merlucciidae 

Lyconus 

Macruronus 

Merluccius 


Weiss,  1974 

Russell,  1976 

Coombs  and  Hiby,  1979 

de  Ciechomski  and  Booman,  1981 

Lisovenko  et  al.,  1982 

Mcintosh,  1893 

Gorbunova,  1954 

Matarese  et  al.,  1981 

Mcintosh,  1893 

Schmidt,  1905a,  1906a 

Ehrenbaum.  1905-1909 

D'Ancona,  1933a 

Rass,  1949 

Russell,  1976 


None 

None 

Aggassiz  and  Whitman,  1885 

Raffaelle,  1888 

Schmidt,  1907a 

Ehrenbaum,  1905-1909 

Kuntzand  Radcliffe,  1917 

D'Ancona,  1933a 

Ahlstrom  and  Counts,  1955 

Miller,  1958 

Fischer,  1959 

Marak,  1967 

Sauskan  and  Serebryakov,  1968 

Santander  and  de  Castillo,  1969 

Colton  and  Marak,  1969 

de  Ciechomski  and  Weiss,  1974 

Russell,  1976 

Brownell,  1979 

Markleetal.,  1980 

Fahay,  1983 


'  No  adull  specimens. 

'  May  refer  to  Svetovidovia.  a  larval  stage  name, 
'  Illustration  only 

*  No  published  descnptions  Mead  ( 1 963)  reports  collection  of  1 5  larvae,  9.0  to  66.0  mm  SL  (MCZ  43083). 
'  Name  applied  to  lar\'al  stage   Referred  to  Coryphaenotdes  acrolepts  by  Johnsen  (1927). 
^  Name  applied  to  larval  stage   Probably  Nezunita  (Marshall  and  Iwamoto.  1973). 

'  Ehrenbaum's  "Macrtindae"  plate  (hg.  108)  illustrates  a  Mauroltcus  egg,  a  Lophius  larva,  a  percifonn  larva  possibly  referrabte  to  Carangidae  or  Serranidae  and  a  92-mm  macround  larva 
resembling  Krohntti^ 
"  Possibly  referrable  to  Atlantic  specimen  of  Mesobnts  (see  Hubbs  and  Iwamoto,  1977). 


identification  by  process  of  elimination  anci  its  identity  must 
await  further  study.  One  of  the  most  elongate  of  the  known 
macrouind  larvae  is  that  of  Odontomacrurus  murrayi  (Maul  and 
Koefoed,  1 950).  Among  these  elongate  types  there  is  a  tendency 
for  caudal  spotting,  either  as  supranal  melanophores(Fig.  141  A) 
or  as  midlateral  spots  or  bars  (Fig.  14 IB).  With  development, 
there  is  a  marked  change  in  mouth  orientation  from  oblique  to 
almost  horizontal  in  the  elongate  Mesohius  herryi  (Hubbs  and 
Iwamoto,  1977)  (Fig.  142C)  and  in  Coryphaenoides  (Stein, 
1980b). 

Known  macround  larvae  can  be  characterized  by  their  mod- 
erate (Gadomus)  to  very  elongate  tail,  lack  of  caudal  fin  and 
moderate  to  very  elongate  pectoral  fin  peduncle.  Some  adult 
diagnostic  characters  (Table  75),  such  as  numbers  of  branchio- 
stegal  rays  and  retia  mirabilia  are  present  early  and  are  crucial 
to  identification  (Merrett,  1978;  Stein,  1980b),  while  others, 
such  as  dorsal  spine  serrations,  develop  late  and  cannot  be  used 
(Merrett,  pers.  comm.,  and  unpublished  observations).  Addi- 


tional characters  (the  interspace  between  dorsal  fins,  anterior 
extent  of  anal  fin  origin,  the  position  of  fin  origins  relative  to 
centra  or  myomeres,  the  relative  size  and  shape  of  pectoral  fin 
peduncles  and  larval  pigmentation)  are  not  known  or  not  re- 
ported, but  appear  to  offer  promise  in  characterizing  groups  of 
larval  macrourids  (see  Table  75  and  Figs.  140,  141A-B  and 
I42C). 

Steindachneriidae.— Steindachneria  was  aligned  with  the  ma- 
crourids in  early  works  (Jordan  and  Evermann,  1896-1900), 
with  merlucciids  by  Norman  ( 1 966),  Marshall  ( 1 966b)  and  Nel- 
son ( 1976),  and  as  a  separate  family  (Marshall  and  Cohen,  1 973). 
Eggs  are  not  known  and  larvae  have  not  previously  been  de- 
scribed although  Mead  (1963)  mentions  specimens,  9.0  to  66.0 
mm  SL  (MCZ  43083).  An  early  planktonic  juvenile  was  avail- 
able and  is  illustrated  in  Fig.  142B-C.  Noteworthy  features  are 
the  distinctive  striated  photogenic  organs  on  the  ventral  surface 
of  the  gut  (Cohen,  1964a),  genital  papilla  and  orifice  separated 


270 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  74.    Developmental  Characters  in  Gadiform  Families  and  Gadid  and  Merlucciid  Subfamilies. 


Muraenolepididae 

Bregmacerotidae 

Mclanonidae 

Mondae 

Gadidae 

Lolinae 

Phycinae 

#  dorsal  fins  (exter- 

2 

1  single  ray  on  head 

1  (deeply 

2(3) 

1-2 

2  (first  some- 

nally) 

plus  1  fin  (divided 
by  low  midsec- 
tion) 

divided) 

times  modi- 
fied) 

#  anal  fins  (external- 

1 

1  (divided  by  low 

1 

1(2) 

1 

1 

ly) 

midsection) 

First  fin  to  form 

Vertical  fins  (pelvic 

Pelvic  (or  anterior 

? 

Pelvic 

Pelvic  (except 

Pelvic 

rays 

late) 

dorsal  ray) 

Lota) 

#  pelvic  rays 

4 

3-4  (larvae) 
5-7  (adults) 

5-7 

2-11 

3-4  (larvae) 
5-8  (adults) 

3-4  (larvae) 
2-9  (adults) 

Pelvic  fin  elongate? 

No 

Yes 

No 

Yes  (most) 

Yes 

Yes 

Pectoral  ray  forma- 

7 

Late 

7 

Midsequence 

Late 

Late 

tion 

to  late 

Body  shape 

Moderately  elongate 

Elongate 

Moderately 
elongate 

Tapers  to  nar- 
row peduncle 

Elongate 

Stocky 

#  vertebrae 

64-69 

43-59 

58-62 

41-63 

59-84 

44-55 

Egg  diameter  (mm) 

1.5-1.6 

? 

7 

0.52-1.16 

1.3-1.5 

0.63-0.98 

Chorion 

Smooth 

? 

7 

Smooth 

Pitted/smooth 

Smooth 

#  oil  globules 

1 

? 

7 

1 

1 

Multiple  to  1 

Miscellaneous 

Premaxillary  spines 

Small  eye; 

Barbel  forms 

Barbel  forms 

Barbels  form  on 

(at  least  1  form); 

dense  pig- 

on lower 

on  lower  jaw 

lower  jaw  (and 

barbel  forms  on 

ment 

jaw;  few  lu- 

in juveniles 

snout  in 

lower  jaw 

minescent 

some);  pterotic 
spines  in  few 

Gadidac 

Merlucciidae 

C«  A.  rt  .-1 »  ^ 

inenidae 

Macroundae 

Gadinae 

Merlucciinae 

Macruroninae 

#  dorsal  fins  (external- 

3 

2  (second  divided           2 

2 

(1)2 

ly) 

by  low  midsec- 

#  anal  fins  (externally)      2 


First  fin  to  form  rays 

#  pelvic  rays 
Pelvic  fin  elongate? 
Pectoral  ray  formation 

Body  shape 

#  vertebrae 

Egg  diameter  (mm) 
Chorion 

#  oil  globules 
Miscellaneous 


Caudal  (pelvic 

last) 
6-7 
No 
Late 

Moderately 
elongate 

39-64 

1.0-1.9 

Smooth 

None 

Barbel  forms  on 
lower  jaw  dur- 
ing/after juve- 
nile stage 


tion 
1  (divided  by  1 
midsection) 

Caudal  (pelvic 

7 

Moderately 

Late 

Elongate 

48-58 
0.8-1.2 
Smooth 
1 


1  (anterior  rays  elon- 

1 (anterior  rays  elon- 

1 

gate  in  Macruron- 

gate) 

us) 

Dorsal  and  anal  (pel- 

9 

Pelvic  (with  dorsal  and 

vic  late) 

anal) 

7-9 

8 

(0)5-17 

No 

No 

Moderately  to  very 

Late 

Late  (pedunculate) 

Late 
(pedunculate) 

Attenuated  (reduced 

Attenuated  (no  caudal 

Attenuated  (no  caudal 

caudal  fin) 

fin) 

fin) 

77-78 

7 

80-116  + 

0.99-1.16 

7 

1.0-2.0 

Smooth 
1 

7 

9 

Hexagonal  pattern 

Luminescent  organ 

Some  luminescent; 

present;  lacks  caudal 

barbel  forms  on  low- 

fin 

er  jaw,  lacks  caudal 
fin 

Fig.  138.  (A)  Muracnolepis  sp.,  32.5  mm  SL,  British  Antarctic  Survey,  53°48.7'S,  38°18.7'W.  (B)  Melanonus  sp.,  30.6  mm  SL,  MCZ  58619, 
35°19'S,  07°30'E.  (C)  Svewvidovia.  13.0  mm  NL,  Fahay,  1983.  (D)  Svetovidovia,  44.4  mm  SL,  HML  H6901,  38°49.5'N,  54"'18.0'W.  (E) 
"Svelovidovia."  44.1  mm  SL,  HML  H9455,  43°21.94'N,  60°32.34'W. 


FAHAY  AND  MARKLE:  GADIFORMES 


271 


'^X/^^^^^//y^y^^^^^^ 


^$$i$s^i^^- 


272 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  139.  (A)  Eretmophorus  kleinenbergi.  105  mm,  Mazzarelli.  1917.  (B)  Rhynchogadus  hepaticus,  21.9  mm.  Cipria,  1927.  (C)  Mora  moro, 
12  mm,  De  Gaetani,  1926.  (D)  Gadella  maraldi.  18.8  mm,  Sparta,  1928.  (E)  Physicutus  nematopus,  9.2  mm,  CALCOFI  5604,  Sta.  103  G  40.  (F) 
Physiculus  nematopus.  14.1  mm,  ventral  view,  CALCOFI  5604,  Sta.  103  G  40. 

from  the  anus,  small  pedunculate  pectoral  fin,  silvery  eye  and     are  unknown.  Merlucciids  have  moderately  pedunculate  pec- 
lack  of  caudal  fin.  torals;  Merluccius  approaches  the  gadines  in  pigmentation  and 

sequence  of  fin  formation  (caudal  first),  while  macruronines 
Merlucciidae.  —  Egs,s,  larvae  and  juveniles  of  Mfr/wcaiw  are  well  approach  the  macrourids  in  pectoral  morphology  and  reduction 
described  (Table  73),  while  those  of  Lyconus  and  Macruronus     of  caudal  fin. 


Fig.  140.  (A)  Macrouridae,  1 1.2  mm  TL,  HML  uncat.,  off  Newfoundland.  (B)  Gadomus  sp.,  30+  mm  TL,  MCZ  58621,  25°48'N,  91°40'W. 
(C)  Corvphaenoides  sp.,  39  mm  TL,  MCZ  58622,  40°04'N,  68°07'W  (pectoral  fin  damaged).  (D)  Coryphaenoides  sp.,  30+  mm  TL,  MCZ  58623, 
34°27'N,  71°19'W. 


FAHAY  AND  MARKLE:  GADIFORMES 


273 


274 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  75.  Meristic  and  Other  Selected  Characters  tn  Genera  of  Macrouridae.  "Number  of  species"  refers  to  minimum  nominal  species 
followed  by  numbers  surveyed  for  characters.  Primary  sources  of  data:  Gunther,  1887;  Gilbert,  1893;  Goode  and  Bean,  1896;  Gilbert,  1905; 
Gilbert  and  Burke,  1912;  Gilbert  and  Hubbs,  1916;  Gilbert  and  Thompson,  1916;  Koefoed,  1927;  Hubbs,  1934;  Parr,  1946;  Koefoed,  1953; 
Smith,  1961;  Scott,  1962;  Iwamoto,  1966;  Makushok,  1966;  Okamura,  1970b;  Hart,  1973;  Marshall,  1973;  Marshall  and  Iwamoto,  1973;  Iwamoto, 
1974;  Iwamoto  and  Stein,  1974;  Hubbs  and  Iwamoto,  1977;  Iwamoto,  1978;  Merrett,  1978;  Iwamoto,  1979;  McCann  and  McKjiight,  1980; 

Trunov,  1981;  Merrett  et  al.,  1983. 


Nature  of  second 

Longest  rays 

Number  of 

Retia 

Branchio- 

Precaudal 

First 

Pectoral 

spinous  ray  of 

(dorsal,  anal, 

Genus 

species 

Mirabilia 

stegal  rays 

vertebrae 

Pelvic  rays 

dorsal  rays 

rays 

first  dorsal  fin 

about  equal) 

Trachyrhynchinae 

Trachyrhynch  us 

4/3 

2 

7 

14 

6-7 

9-12 

18-26 

(ray) 

D 

Macrouroidinae 

Macrouroides- 

1/1 

3 

7 

_ 

0 

_ 

22-25 

(ray) 

Dor  s 

Squalogadus- 

1/1 

3 

7 

12-13 

5 

- 

23-26 

(ray) 

Dor  s 

Bathygadinae 

Gadomus 

11/9 

4 

7 

11-13 

8-9 

11-14 

13-25 

Smooth 

D 

Bathygadus 

14/10 

2 

7 

11-13 

7-11 

8-13 

10-20 

Smooth 

D 

Macrourinae 

Cynomacrurus- 

1/1 

0 

6 

— 

7-8 

— 

15-17 

Smooth 

A 

Ondontomacrurus'^ 

1/1 

2 

6 

— 

7-8 

8-10 

8-11 

Smooth 

A 

Lepidorhynchus 

1/1 

9 

6 

— 

8-9 

11-13 

16-17 
(19) 

Smooth 

— 

Mahia 

1/1 

? 

6 

_ 

7 

11 

17-18 

Smooth 

A 

Coelorhynchus 

58/40 

4 

6 

11-12 

7 

9-14 

(13,  14) 
15-21 
(22) 

Smooth 

A 

Hyomacrums 

2/1 

2 

6 

13 

8 

— 

— 

Serrate 

— 

Coryphaenoides 

46/24 

4 

6 

11-16 

(6) 
7-11 
(12) 

9-14 

15-25 

Serrate 

A 

Macrounis 

3/2 

4 

6 

16 

8-9 

10-13 

17-21 

Serrate 

A 

Nematonums 

4/1 

5 

6 

15 

(7) 
9-11 

10-12 

18-21 

Serrate 

A 

Chalinura 

8/6 

6 

6 

12-13 

8-14 

9-12 

18-22 

Serrate 

A 

Lionurus 

3/2 

6 

6 

11-13 
(15) 

8-11 

10-11 

15-20 

Serrate  or 
smooth 

A 

Mesobius'-^ 

2/2 

2 

7 

11 

6-9 

10-12 

12-16 

Weakly 
serrate 

A 

Echwomacrurus- 

2/2 

0 

7 

— 

9-12 

11-13 

16-19 

Serrate  or 
smooth 

A 

Hymenocephalus 

17/3 

2 

7 

10-11 

7-15 

10-13 
(14) 

11-18 

(18-22) 

Weakly 
serrate 
or  smooth 

A 

Cetonurus 

2/2 

2 

7 

10 

8-10 

9-12 

16-19 

Serrate 

A 

Paracetonurus 

4/3 

2 

7 

11 

(5) 

6-9 

9 

8-11 

17-21 

Serrate 

A 

Kumba 

1/1 

2 

7 

- 

11 

21 

Smooth 

A 

Parakumba 

1/1 

7 

7 

12 

10 

11 

22 

Weakly 
serrate 

A 

Macrosmia 

1/1 

2 

7 

12 

11-12 

11-13 

22 

Weakly 
serrate 

A 

Mataeocephalus 

6/3 

2 

7 

13-14 

7-9 

10-12 

(19) 
22-26 

Serrate 

A 

Trachonurus 

1/1 

2 

7 

12-13 

7 

7-11 

13-18 

Smooth 

A 

Sphagemacrurus ' 

7/2 

2 

7 

11-12 

(8) 
10-13 

12-13 

18-22 

Serrate 

A 

Nezumia 

44/32 

2(4) 

7 

13-14 

6-17 

10-15 

13-27 

Serrate 

A 

Pseudonezumia 

1/1 

n 

8 

_ 

6 

10 

16 

Serrate 

A 

Mataeocephalus' 

4/3 

2 

7 

14 

8-10 

11-16 

16-22 

Smooth 

A 

Venlrifossa 

16/10 

2 

7 

10-14 

8-11 
(13-15) 

11-15 

18-27 

Serrate 

A 

'  Juvenile  phase  known  lo  have  prominently-spoued  pigment  pattern  (but  see  note  on  Sphagemacrurus  in  Table 
■  Includes  bathypelagic  species. 


73). 


FAHAY  AND  MARKLE:  GADIFORMES 

Table  75.    Extended. 


275 


Mouth  posnion 

Number  of  hghl    (terminal,  sub-ter-  Chm  barbel 
organs  (bulbous               minal.  (present 

or  tubular)  mfenor)  or  absent) 


Position  of  anus  and  urogenital 
opening  relative  lo  anal  fin 
ongin  and  pelvic  fin  bases 


A. 
ong- 


Plv. 

bases 


Anal  fin  ongin 

antenor  to 
postenor  end 
of  gut  cavity? 


Anteriormost 
fin  ongin 


Distance  between  dorsal  fins 


<DI 
base 


=  DI 

base 


>DI 

base 


Inf 


Pr 


No 


D2or  s 


0 

Inf 

Ab 

X 

No 

D2 

Single 

fin 

0 

Inf 

Ab 

X 

No 

D2 

Single 

fin 

0 

Term 

Pr/Ab 

X 

No 

D2 

X 

0 

Term 

Pr/Ab 

X 

No 

D2 

X 

0 

Term 

Ab 

X 

IB 

Term 

Ab 

X 

— 

A 

X 

IB 

Term 

Pr 

X 

- 

- 

IB 

Sub-T 

Pr 

X 

Yes 

A 

X 

IB 

Inf 

Pr 

X 

or 

X 

Yes 

A  (usu.) 

X 

or         X 

or 

X 

or 

(usu.) 

IT 

0 

Inf 

Pr 

X 

Yes 

_ 

0 

Sub-T, 
Inf 

Pr 

X 

Yes 

A 

X 

or         X 

or 

X 

0 

Inf 

Pr 

X 

Yes/no 

D2 

X 

0 

Sub-T, 
Inf 

Pr 

X 

Yes/no 

A 

X 

0 

Sub-T 

Pr 

X 

Yes/no 

A 

X 

or 

X 

0 

Inf 

Pr 

X 

Yes/no 

A 

X 

IB 

Sub-T 

Ab 

X 

Yes 

A 

X 

0 

Inf 

Pr 

X       ?       X 

- 

A 

X 

IT 

Sub-T 

Pr/Ab 

X 

No 

A 

X 

Oor  IB 

Sub-T 

Pr 

(tiny) 

X 

Yes 

A 

X 

0 

Sub-T, 
Inf 

Pr 

X 

Yes 

A(?) 

IB 

Sub-T 

Pr 

(tiny) 

X 

Yes 

A 

X 

1B(?) 

Sub-T 

Pr 

(tiny) 

X 

— 

A 

X 

0 

Sub-T 

Pr 

X 

Yes 

A 

X 

0 

Inf 

Pr 

X      or       X 

Yes 

A 

X 

Oor  IB 

Sub-T 

Pr 

X 

Yes 

A 

X 

IB 

Sub-T 

Pr 

X 

Yes 

A 

X 

IB 

Sub-T 

Pr 

X 

(X) 

Yes 

A 

X 

or 

X 

0 

Inf 

Pr 

X 

_ 

A 

X 

IB 

Sub-T 

Pr 

X 

Yes 

A 

X 

IB 

Sub-T, 
Inf 

Pr 

X 

(X) 

Yes 

A 

X 

or 

X 

276 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  141.     (A)  Macrouridae  (Macrourinae).  60  mm  XL,  HML  H6818,  39°52.5'N,  58°54.0'W.  (B)  Macroundae  (Macrounnae),  1 5  mm  TL,  MCZ 
58624,  ag-SS'N,  79°54'W. 


FAHAY  AND  MARKLE:  GADIFORMES 


277 


~>«^ 


it 

n 

J  pi 


^' 


-sjft'S"-' 


^■^...JM^.X 


.p^- 


Fig.  142.    (A)  Steindachnena  argentearlA  mm  TL,  GCRL  01962.  28°45'N,  89°36'W.  (B)  Steindachneria  argentea.  24  mm  TL,  GCRL  01962, 
ventral  view.  (C)  Mesobius  berryi.  23.4  mm  TL.  Hubbs  and  Iwamoto,  1977. 


Gadidae— The  early  life  history  stages  of  gadids  are  well  known 
(Table  73)  and  are  reviewed  and  characterized  in  this  volume 
by  Dunn  and  Matarese. 

Selected  Characters 

•Eg?-?.— Eggs  are  undescribed  for  three  gadiform  families:  Breg- 
macerotidae,  Melanonidae  and  Steindachneriidae.  Efremenko 
(1983b)  recently  described  muraenolepidid  eggs  and  Markle 
(1982)  summarized  information  for  the  remaining  families  and 
noted  that  a  relatively  small  egg  (<  1  mm)  with  an  oil  globule 
was  a  widespread  and  probably  primitive  character.  The  oil 
globule  has  apparently  been  lost  only  in  the  gadines,  a  group 
showing  numerous  derived  stales,  including  relatively  large  eggs 
(Markle,  1982). 

Except  in  the  gadid,  Brosme,  and  macrourids,  chorion  or- 


namentation appears  to  be  restricted  to  ubiquitous  pores  seen 
with  scanning  electron  microscopy  (Lannig  and  Hagstrom,  1975). 
In  B.  brosme  the  chorion  pores  are  many  times  larger  than  in 
other  gadiforms  and  give  the  egg  a  pitted  appearance  (Markle 
and  Frost,  MS).  In  macrourids  an  elaborate  "honey-comb"  or- 
naments the  chorion.  This  ornamentation,  like  the  pores,  has 
an  unknown  function.  The  uniqueness  of  the  "honey-comb" 
(Boehlert,  this  volume)  and  its  presence  in  all  known  macrourine 
eggs  suggests  an  autapomorphy,  at  least  for  the  subfamily.  Ad- 
ditional information  on  egg  morphology  of  merlucciids,  ma- 
crourids, morids  and  Steindachneria  could  contribute  to  a  dis- 
cussion of  the  unsettled  status  and  relationships  of  the  latter. 

Transient  early  life  history  characters— After  hatching  there  are 
at  least  six  characters  that  can  be  considered  ontogenetically 


Fig.  143.     (A)  Gadiis  morhua.  1  1.0  mm,  Fahay.  1983.  (B)  Brosme  bwsme.  14.0  mm,  Fahay,  1983.  (C)  Urophycis  chuss.  9.5  mm,  Fahay,  1983. 
(D)  Merluccius  productus.  10.1  mm,  Ahlstrom  and  Counts,  1955. 


FAHAY  AND  MARKLE:  GADIFORMES 


279 


transient;  larval  pigmentation,  lateral  maxillary  fangs,  pterotic 
spines,  pedunculate  pectoral  fin  bases,  sequence  of  develop- 
mental events  and  the  presence  of  a  pelagic  juvenile  stage.  Many 
of  these  characters  are  incompletely  known  for  the  order  and 
only  tentative  phylogenetic  statements  can  be  made. 

Embryonic  and  larval  pigmentation  patterns  are  quite  vari- 
able. In  gadoids  there  appears  to  be  widespread  occurrence  of 
postanal  bands,  usually  one  or  two,  and  melanophores  at  the 
notochord  tip.  Similar  patterns  occur  in  Merlnccius  bilinearis 
(Merlucciidae),  Physiculus  capensis  (Moridae)  and  Coelorhyn- 
chus  sp.  (Macrouridae).  However,  even  within  one  subfamily 
such  as  the  gadines,  there  are  genera  without  any  banding  (e.g., 
Melanogranuniis)  as  well  as  much  variation  in  number  of  bands 
(Pollachius.  Gadus).  Eye  pigmentation  at  hatching  varies  de- 
pending on  development  stage  at  hatching,  for  example,  unpig- 
mented  in  Pollachius  and  pigmented  in  Gadus.  Embryonic  and 
larval  pigmentation  seems  variable  in  the  well  studied  gadids 
as  well  as  in  other  gadiforms  so  that  an  evaluation  of  phylo- 
genetic significance  seems  premature  at  this  time. 

Lateral  premaxillary  spines  are  only  known  in  Muraenolepis 
(Fig.  1 38A)  and  larval  pterotic  spines  are  only  known  in  Phycis, 
Gaidropsarus  (Demir,  1982;  Markle,  1982)  and  Ciliata  (Dunn 
and  Matarese,  this  volume).  Both  characters  appear  to  be  apo- 
morphies,  providing  phylogenetic  information  at  the  generic 
level  at  least.  The  western  Atlantic  Phycis.  P.  chesteri,  lacks 
larval  pterotic  spines  and  may,  in  fact,  belong  in  Urophycis 
(David  Methven,  pers.  comm.). 

The  lack  of  developmental  series  outside  the  gadoids  pre- 
cludes discussion  of  many  developmental  sequence  characters. 
However,  it  does  seem  possible  to  make  some  tentative  state- 
ments about  the  first  fin  to  form  rays.  On  the  basis  of  our 
examination  of  a  larval  series  provided  by  A.  W.  North  of  the 
British  Antarctic  Survey,  Muraenolepis  does  not  form  pelvic 
rays  first.  This  contrasts  with  most  gadiforms  where  the  pelvic 
is  the  first  fin  to  form  (Table  74).  Other  exceptions  seem  to  be 
in  gadines  and  merlucciids  where  the  caudal  or  dorsal  and  anal 
fins  form  before  the  pelvic.  The  latter  condition  may  represent 
a  derived  character  state.  However,  the  tail-less  macrourids  are 
precluded  from  showing  this  character. 

The  pectoral  fin  base  is  strongly  pedunculate  (stylopterous) 
during  the  larval  period  in  macrourids  and  steindachneriids, 
moderately  pedunculate  in  morids  and  narrow-based  (but  less 
pedunculate)  in  bregmacerotids,  Meriuccius  and  gadids  (Fig. 
143A-D).  Strong  expression  of  this  character  is  associated  with 
loss  of  the  caudal  fin  (macrourids.  steindachneriids)  or  delayed 
caudal  fin  formation  (morids)  and  may  reflect  a  compensatory 
response  of  larvae  associated  with  larval  locomotion. 

In  the  life  history  of  most  gadiforms  there  is  a  benthic  or 
engybenthic  adult  phase.  In  all  of  these  groups  (muraenolepi- 
dids,  morids,  most  gadids,  merlucciids,  most  macrourids)  as 
well  as  in  pelagic  gadiforms  there  is  a  prolonged  pelagic  juvenile 
stage  which,  in  some  cases,  includes  symbiotic  association  with 
jellyfish  (Mansueti,  1963).  In  phycines,  for  example,  this  stage 
is  neustonic,  includes  a  pigmentation  pattern  different  from  both 
larval  and  benthic  juveniles,  and  is  characterized  by  a  dense 
concentration  of  melanophores  on  the  dorsal  surface.  In  morids, 
some  pelagic  juveniles  have  been  described  as  new  genera,  such 
as  Svetovidovia  Cohen,  1973  (=Gargilius  Koefoed,  1953). 

We  are  not  aware  of  any  gadiform  that  can  be  shown  not  to 
possess  a  pelagic  juvenile.  In  fact,  it  appears  that  life-history 
neoteny  has  occurred  several  times  and  adults  have  retained  the 
pelagic  habitat  (bregmacerotids,  melanonids,  the  gadines  Gad- 
iculus  and  Micromesistius,  and  some  macrourids).  The  pelagic 


adult  has  clearly  evolved  independently  more  than  once.  Even 
within  a  single  family,  Macrouridae,  it  has  apparently  happened 
at  least  three  times  and  Hubbs  and  Iwamoto  (1977)  have  called 
attention  to  this  form  of  neoteny  with  the  generic  name,  Me- 
sobius  ("middle  life"). 

Pelvic  fins.— The  gadiform  pelvic  fin  shows  two  major  ontoge- 
netic sequences.  In  the  phycines,  Urophycis  and  Phycis,  larvae 
initially  form  3  or  4  rays  and  ontogenetically  reduce  or  resorb 
the  innermost  ray  to  produce  the  adult  count  of  2  or  3  (Markle, 
1982).  During  the  course  of  this  study,  we  have  also  found 
ontogenetic  pelvic  fin  ray  reduction  in  the  morid  Svetovidovia 
vitellius.  One  transforming  specimen,  55  mm  SL  (MCZ  59773), 
has  two  large  pelvic  fin  rays  and  2  or  3  very  minute  remnants 
of  inner  pelvic  fin  rays.  Smaller  specimens  have  as  many  as  1 1 
rays  (Table  72).  Cohen  (1979)  has  previously  suggested  that 
Lotella  ma.xillaris  (10  pelvic  fin  rays)  may  be  the  young  of 
Laemonema  (1-3  pelvic  fin  rays).  Gadiforms  may  be  pre-adapt- 
ed  for  this  type  of  metamorphosis  since  even  in  species  with 
numerous  pelvic  fin  rays,  such  as  the  morid  Physiculus,  the 
external  fin  ray  nerves  appear  restricted  to  the  outer  two  rays 
(Freihofer,  1970;  Fig.  12). 

In  the  other,  presumed  ancestral,  ontogenetic  sequence,  pelvic 
fin  rays  increase  in  number.  Variation  is  seen  in  this  sequence 
in  the  speed  at  which  the  adult  complement  is  formed.  The  rays 
form  very  quickly  in  Meriuccius,  somewhat  more  slowly  in  Ele- 
ginus.  and  over  a  protracted  size  range  in  Ga/^ropsarw,?  (Markle, 
1982;  Dunn  and  Vinter,  1984,  MS). 

In  many  gadiforms,  such  as  some  macrourids,  Meriuccius, 
many  gadids  and  morids,  the  pelvic  fins  also  change  allometri- 
cally.  In  the  Krohnius  and  several  other  types  of  macrourid 
larvae  as  well  as  in  morids,  the  pelvics  are  greatly  expanded 
over  their  relative  size  in  any  known  adult.  In  some  phycines 
the  pelvics  are  not  necessarily  relatively  longer,  but  are  wider 
and  fan-like  as  opposed  to  filamentous  in  adults.  This  allometry 
favoring  a  relatively  large,  fan-like  pelvic  in  the  young  would 
seem  to  be  a  device  to  aid  flotation.  It  is  noteworthy,  however, 
that  only  bregmacerotids  among  the  pelagic  gadiforms  have 
retained  enlarged  pelvic  fins  as  adults. 

In  addition  to  elongation,  prominent  pigmentation  of  pelvic 
fins  characterizes  many  genera,  including  most  phycines.  The 
precise  extent  and  location  of  pigment  on  the  pelvic  fin  is  often 
an  important  identifying  feature  in  these  larvae.  For  example, 
it  is  absent  in  Urophycis  regia,  restricted  to  the  tips  of  the  fins 
in  most  other  Urophycis  and  densely  covers  the  fin  membrane 
in  U.  tenuis,  Enchelyopus,  Gaidropsarus  and  Raniceps. 

Gadines,  as  previously  mentioned,  show  a  clear  departure  in 
the  sequence  of  fin  formation.  Instead  of  forming  first,  pelvic 
fins  form  last.  In  Meriuccius,  whose  condition  may  be  an  evo- 
lutionary precursor  to  the  gadine  condition,  pelvic  fins  form 
second  in  the  sequence  after  the  caudal  (Table  74). 

Pectoral  fins.  — \s  is  the  case  with  most  teleosts,  pectoral  fin  rays 
form  late,  although  they  may  form  before  the  late-forming  cau- 
dal in  morids.  As  with  the  pelvic  fin,  the  pectoral  fin  is  often 
elongate  and/or  fan-shaped  in  some  morids  and  macrourids  (i.e., 
Gadella  and  Hymenocephalus)  but  this  fin  is  not  prominently 
pigmented  in  any  member  of  the  order  except  some  species  of 
Meriuccius. 

Dorsal  and  anal  fins.  — \n  the  development  of  all  gadiforms  de- 
scribed, vertical  fins  form  in  their  adult  positions  and  there  is 
no  evidence  of  fin  base  migration.  The  dorsal  fin  origin  in  gad- 


280 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  76.    Distribution  of  Caudal  Rays  on  Supporting  Bones  of  the  Calidal  Fin  in  Selected  Species  of  Gadiformes.  "Inferior  Hypurals" 
do  not  include  the  parhypural  in  this  listing.  See  Table  72  for  primary  sources  of  data. 


Epu- 
rals 

Hypu- 
rals 

XY 

bones 

Number  of 
vertebrae 
associated 

with 
caudal  Iin 

Dorsal 

Number  of  rays 

Superi- 
or         Infenor 
hypu-        hypu- 
ral(s)          ral(s) 

Ventral 

Taxa 

(Un- 
branched) 

(Branched) 

(Branch- 
ed) 

(Un- 
branched) 

caudal 
rays 

Muraenolepididae 

Muraenolepis  sp. 

2 

2 

See  text 

See  text 

2-3 

4 

2 

1-2 

8-10 

Bregmacerotidae 
Bregmaceros 
balhymaster 

2 

2 

Present 

5-6 

(10-12) 

8 

2 

(10-12) 

34-36 

(?) 

Melanonidae 

Meianonus  zugmayeri 

2 

(4)  5  (6) 

Absent 

13 

23-25 

6 

3 

22-25 

55-60 

Fam.  Incertae  sedis 

Euclichlhys  sp. 

2 

2 

X  present 
Y  absent 

10 

17 

4 

3 

17 

41 

Moridae 

Antimora  rostrata 

Brosmiculus  sp. 
Eretmophorus 
kleinenbergi 
Gadella  sp. 
Gadella  maraldi 
Laemonema  barbatula 
Laemonema  longipes 
Lepidwn  eques 
Lepidion  lepidion 
Mora  mora 
Lotella  fernandeziana 
Physiculus  neniatopus 

2 

2 
2 

2 
2 
2 
2 

2 
2 
2 
2 
2 

2 

2 
2 

2 

5(6) 

5 

5 
5 
5 
5 
5 
5 
5 
5 
5 

5 

5 
5(6) 

5 

Present 

Present 
Present 

Present 
Present 
Present 
Present 
Present 
Present 
Present 
Present 
Present 

Present 

Present 
Present 

Present 

8 
9 

7 

7 
4-5 

5-6 

7 
5 

7(?) 

(5-6) 

(6-7) 
(6-8) 
(4-6) 

(6-8) 

14 
15 

16 

8-9 
11-13 

15 

7 

(7) 
8-9 

(2-3) 

(6-7) 
(3-4) 

(3-5) 

5 
6 

5 

6 

6 

24- 

6 
5-6 

13- 

5-6 

14- 

6 
6 

6 

4 
4 

4 

3-4 

3 

(5-6) 

12 
13 

13 

12-15 
15-17 

(6-7) 

(8-9) 
(6-7) 
(5-7) 

(7-8) 

(30) 
35 
38 
26 

38 

20-24 
27-31 
22-25 
34-36 

22 

3-4 
4 

(5-6) 
(0-1) 

34-38 
22-26 

Physiculus 

rastrelliger 
Salilota  australis 
Triplerophycis 

gilchrisli 
Svetovidovia 

4 

(1-3) 

27-32 

4 
4 

3 

15 
5 

(12) 
16-18 

40 
22 

(28) 
30-33 

Lotinae 

Brosme  brosme 

2 

2 
2 

2 

2 
2 

Absent 

Absent 

Absent 

(variable) 

14 
13 

19-20 

4 

3 

22-24 

45-48 

Lota  lota 
Molva  molva 

(4-5) 

20-22 
22-23 

(14-15) 

4  2 

5  3 

(12-13) 

21-25 
25 

(4-6) 

42-50 
52-53 

Phycinae 

Enchelyopus  cimbrius 

2 

2 

Present 

8-9 

(5) 

12-15 

(9) 

(4) 
5 

2 

(6) 

13-15 

(5) 

31-35 

2 
2 

2 
2 
2 
2 
2 

2 

2 

2 
2 
2 
2 
2 

Present 
Present 

Present 
Present 
Present 
Present 
Present 

8-9 
7-8 

6 

7 
8 

(6-8) 

(3) 

(7-8) 

(9) 

-24 

2 

* 

3 

(3-4) 

18-21 
14-15 

14-17 

(6-8) 

(4) 

(7-8) 

(9) 

Gaidropsarus  ensis 
Gaidropsarus 

mediterraneus 
Phycis  chesteri 
Phycis  blennioides 
Urophycis  regia 
Urophycis  chuss 
Urophycis  tenuis 

17-20 
13-15 

12-14 

11-12 
12-13 

13-15 

(5-6) 

(4-5) 
(5) 

5 

(4) 
5 
6 

6 
6 

(5) 
6 

41-46 
31-36 

32-37 

28 
30-32 
29-34 
33-39 

3 
3 
3 

(2) 
(3) 

13-14 
10-11 

13-14 

Gadinae 

Arctogadus  borisovi 
Boreogadus  saida 

Eleginus  gracilis 
Gadiculus  argenteus 

2 
2 

2 

2 
2 

2 

Absent 
Absent 

Absent 

15 
13-14 

15-16 

(15) 
(7-8) 

21-26 
21-25 

22-25 

(8) 

4 
4 

5 
20- 

1 
2 

(3) 

2 

(6) 

22-25 
21-25 

23-26 

(16) 
(7-8) 

47-55 
46-54 

50-56 

35-36 

(?) 

FAHAY  AND  MARKLE:  GADIFORMES 

Table  76.    Continued. 


281 


Epu- 
rals 

Hypu- 
rals 

XY 

bones 

Number  of 
vertebrae 
associated 

with 
caudal  tin 

Dorsal 

Num 

Superi- 
or 
hypu- 
ral(s) 

ber  of  rays 

inferior 
hypu- 
raKsl 

Ventral 

To 

cau 

ra 

tal 

Taxa 

(Un- 
branched) 

(Branched) 

(Branch- 
ed) 

(Un. 
branched) 

dal 

Gadus  morhua 
Gadus  ogac 
Gadus  macrocephalus 
Melanograminus 

aegtefimis 
Microgadus  tomcod 
Microgadus  proximus 

2 
2 

2 
2 

2 
2 

2 
2 
2 

2 

2 
2 

Absent 
Absent 
Absent 
Absent 

Absent 
Absent 

13-14 

14 
11-13 

14 

13-14 
14-15 

(12-13) 
(15-16) 

(9) 

22 
21 

25- 

21 

■yi. 

-26 
-23 

-27 

-22 
-75 

(11) 
(11) 

(12) 

(11-13) 

4 
4 

4 
5 

4 
(4) 

2 

2 
2 
2 

1 

(9) 
(9) 

(10) 

(9-11) 

(13-14) 
(8-9) 

23- 

21- 
26- 

21- 

-27 
-24 

-27 

-24 
-''4 

(12-13) 
(15) 

(10) 
(11-12) 

(16-17) 
(11-13) 

49- 
46- 
51- 

57- 

46- 
49- 

-57 
-51 

-53 
-60 

-50 

-56 

Micromesistius 

poulassou 
Poltachius  virens 
Theragra 

chalcogramma 

2 

2 
2 

2 

2 
2 

Absent 

Absent 
Absent 

15 

15-16 
12-13 

(15- 
(12- 

-16) 
-14) 

21- 
30- 

-22 
-32 

(15-16) 
(9-10) 

5 
5 

5 
4 

2 

2 
2 

23- 
31- 

-24 
-33 

50 

66-70 

Merluciidae 

Merlucaus  albidus 
Merlucinis  bilinearis 

2 
2 

2 
2 

Present 
Present 

10 
9 

(10) 
(8) 

16 
13-15 

(6) 
(5-6) 

6 
6 

3 

(2) 

3 

2-3 

2 

(4) 
(5-6) 

18 
15-17 

(10) 
(7-8) 

40 
34-37 

Merluccius  productus 
Macruronus 
novaezelandiae 

2 
2 

2 

Present 
See  text 

9-10 

2 

(10- 

-11) 

4? 

(6-7) 

6 
3 

(4-6) 

2- 

-3? 

(10-11) 

42- 
10 

-43 
-12? 

•  Included  in  ventral  count. 


iforms  vaiies  from  occipital  to  slightly  behind  the  pectoral  fin- 
tips  and  is  long-based,  reaching  almost  to  the  caudal  fin  (and 
confluent  with  the  caudal  in  Muraenolepis  and  Macruronus). 
The  anal  fin  originates  close  behind  the  anus  (except  in  some 
macrourids  and  steindachneriids)  and  also  extends  posteriorly 
to  near  the  caudal  fin  (confluent  with  caudal  in  Muraenolepis 
and  Macruronus).  The  caudal  peduncle  is,  as  a  result,  very  short, 
being  longest  in  Merluccius.  Variations  are  fijund  in  morids 
(where  the  dorsal  fin  may  extend  farther  posteriorly  than  the 
anal,  accommodating  an  asymmetrical  caudal  fin),  and  in  Stetn- 
dachneria  and  the  Macrouridae  (where  the  lack  of  a  caudal  fin 
results  in  dorsal  and  anal  rays  tapering  until  they  meet  at  the 
tip  of  the  tail). 

Gadiform  dorsal  and  anal  fin  rays  usually  form  after  pelvic 
fin  rays  begin  (with  the  exception  of  the  anterior  dorsal  ray  in 
Bregmaceros.  which  may  form  very  early).  Usually,  the  vertical 
fins  ossify  together,  so  that  caudal,  dorsal  and  anal  fin  rays 
appear  at  about  the  same  time.  In  some  genera  with  one  long 
dorsal  fin  (or  a  short  first  dorsal  fin  preceding  a  longer  second 
dorsal)  fin  rays  in  the  longer  fin  form  from  two  or  more  centers 
of  ossification,  for  example,  Molva  (Schmidt,  1906b)  and  Ater- 
luccius  (Ahlstrom  and  Counts,  1955).  This  may  indicate  either 
preadaptation  for  the  multiple  dorsal/anal  condition  or  second- 
ary loss  of  multiple  dorsals/anals. 

In  cases  where  a  long  second  dorsal  fin  is  preceded  by  a  rel- 
atively short  first  dorsal  fin,  development  of  the  first  is  usually 
delayed  and  is  often  the  site  of  the  last  fin  ray  formation,  as 
in  Gaidropsarus  {Gcmn.  1982)  and  tVop/jic/.v  (Hildebrand  and 
"^able,  1938).  In  gadines,  which  have  three  dorsal  fins,  the  first, 
again,  forms  last.  In  Macrouridae  and  Merluccius.  however,  first 
dorsal  fin  rays  appear  to  form  much  sooner  than  the  second 
(Merrett,  1978;  Ahlstrom  and  Counts,  1955).  In  the  morids 
Physiculus  and  Svetovidovia  the  first  and  second  dorsal  fins 
develop  together  and  divide  late  in  development  into  apparent 
first  and  second  fins.  Thus  in  the  gadiforms,  ossification  of  ver- 


tical fin  rays  is  not  uniformly  anterior  to  posterior,  or  vice-versa, 
or  from  the  middle  toward  both  ends,  but  instead  is  variable. 
The  relationship  of  dorsal  and  anal  fin  rays  to  centra  is  an 
important  characteristic  in  gadiforms  (Rosen  and  Patterson, 
1969;  Marshall  and  Cohen,  1973).  We  believe  the  primitive 
state,  as  exemplified  by  Muraenolepis,  involved  about  three  rays 
per  centrum.  The  major  evolutionary  trend,  as  identified  by 
Rosen  and  Patterson  ( 1 969),  is  for  this  ratio  to  change  anteriorly 
where  it  is  replaced  by  an  approximate  2:1  relationship  in  most 
families  of  gadiforms.  In  phycines,  two  derived  character  states 
are  found:  a  7:1  relationship  in  rocklings  (Cohen  and  Russo, 
1979;  Markle,  1982:  fig.  5B)  and  a  2:1  relationship  in  hakes 
(Markle,  1982:  fig.  5C). 

Caudal  fin.  — Y>e^X>^\t  its  absence  in  over  a  third  of  all  gadiforms, 
the  distinctive  caudal  skeleton  has  received  an  inordinate  amount 
of  attention.  In  the  present  context  it  is  doubly  important,  for 
its  presence  offers  identification  as  well  as  phylogenetic  infor- 
mation (Markle,  1982;  Dunn  and  Vinter,  1984,  MS).  Tailed 
gadiform  larvae  typically  have  a  symmetrical  fin  where  some 
"secondary"  rays  form  before  some  "primary."  As  used  here, 
"primary"  refers  to  those  rays  articulating  with  the  superior 
hypural  (hypurals  three  through  five)  while  rays  attached  to 
inferior  hypurals  (hypurals  one  and  two),  parahypurals,  epurals, 
accessory  bones  (the  X  and  Y  bones  of  some  authors),  or  to 
elongate  neural  and  haemal  spines  are  referred  to  as  "second- 
ary." Secondary  rays  also  include  those  originating  between 
neural  or  haemal  spines. 

We  have  summarized  the  distribution  of  gadiform  caudal  fin 
rays  in  Table  76.  Various  authors  lump  secondary  rays  together 
or  express  them  as  branched  or  unbranched.  We  have  included 
both  methods,  thus  the  sum  of  counts  from  different  sources  do 
not  always  correspond.  Several  things  are  apparent  from  this 
table.  One  involves  the  utility  of  using  the  distribution  of  caudal 
rays  both  in  intra-  and  intergeneric  comparisons  (Markle,  1982; 


282 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


CFR 


A    Muraenolepis 


Fig.  144.  Hypothesized  acquisition  of  gadoid  caudal  structure  from 
condition  in  Muraenolepis  (see  text).  X  and  Y  bones  shaded.  Abbre- 
viations: CFR:  caudal  fin  rays;  EP:  epurals;  HS:  haemal  spines;  IH: 
inferior  hypural;  NS;  neural  spine;  PH:  parhypural;  PU,:  first  preural 
centrum;  R;  radials;  SH;  superior  hypural;  U,:  first  ural  centrum;  U,; 
second  ural  centrum. 


Dunn  and  Vinter,  1984,  MS).  Primary  and  total  caudal  fin  ray 
counts  also  exhibit  some  difference  in  symmetry  and  patterns 
of  evolutionary  change.  Morids  are  the  only  group  of  tailed 
gadiforms  that  show  noticeable  asymmetry  in  superior  versus 
inferior  secondary  caudal  rays  (Table  76).  Morids  and  some 
phycines  have  relatively  low  total  caudal  fin  ray  counts  (20-38) 
and  numerous  groups  have  some  genera  with  six  primary  caudal 


fin  rays.  Markle  (1982)  interpreted  both  of  these  as  primitive 
states. 

The  caudal  oi Muraenolepis  differs  from  most  other  gadiforms' 
in  its  complete  continuity  with  both  dorsal  and  anal  fin  rays 
(Fig.  144A).  It  is  virtually  identical  to  that  of  the  ophidiiform, 
Brotula  (Monod,  1968),  differing  only  in  number  of  rays  sup- 
ported by  the  parhypural  (one  vs.  two).  The  typical  gadoid 
caudal  fin  skeleton  is  easily  derived  from  the  condition  in  Mu- 
raenolepis, which  we  identify  as  the  primitive  state.  The  primary 
requirement  is  the  acquisition  of  X  and  Y  bones  and  modified 
spines  of  the  first  preural  centrum,  both  sets  of  which  must  have 
cartilaginous  articulating  surfaces  entering  into  support  of  cau- 
dal rays.  X  and  Y  bones  are  present  in  Muraenolepis  as  the 
penultimate  radials  of  the  dorsal  and  anal  fins.  If  the  last  radials 
fuse  with  the  spines  of  the  first  preural  centrum,  both  sets  of 
preural  caudal  bones  (with  cartilaginous  articulating  surfaces) 
are  acquired. 

A  second  requirement  is  an  interspace  (lacking  rays  and  ra- 
dials) between  the  caudal  fin  and  the  dorsal  and  anal  fins.  This 
condition  could  have  been  satisfied  in  one  of  two  ways.  Rays 
(and  their  supporting  radials)  anterior  to  the  X  and  Y  bones 
might  have  been  lost,  and  subsequent  changes  in  caudal  ray 
numbers  would  then  involve  the  addition  of  secondary  rays 
lacking  radials.  A  less  parsimonious  scenario  involves  the  loss 
of  radials  (only)  anterior  to  the  X  and  Y  bones  which  leaves  a 
continuous  dorsal-caudal-anal  fin  including  some  anterior  un- 
supported rays.  In  this  case,  further  variation  in  numbers  of 
secondary  caudal  rays  would  involve  both  increases  and  de- 
creases. The  hypothesized  ancestral  gadoid  condition  is  shown 
in  Fig.  144B.  Presumably,  this  ancestor  would  have  had  16 
caudal  fin  rays  (one  each  on  X  and  Y  bones,  first  preural  neural 
and  haemal  spines,  each  epural  and  parhypural,  six  on  the 
superior  hypural  and  three  on  the  inferior  hypural).  This  total 
is  close  to  the  lowest  known  (and  presumably  most  primitive) 
counts  in  certain  morids  (Table  76)  and  corroborates  the  sug- 
gestion that  higher  counts  in  Melanonidae,  Gadidae  and  Mer- 
luccius  are  derived  states  brought  about  through  the  acquisition 
of  additional  rays  lacking  radial  support  (Fig.  144C).  In  Brosme 
this  acquisition  has  resulted  in  a  secondary  elimination  of  the 
caudal  peduncle  and  an  almost  continuous  dorsal-caudal-anal 
fin  (Markle,  1 982:  fig.  7C).  The  acquisition  of  rays  has  apparently 
occurred  asymmetrically  in  some  morids,  where  ventral  sec- 
ondary rays  outnumber  dorsal. 

Olfactory  lobes.— J\\e  position  of  olfactory  lobes  relative  to  na- 
sal organs  and  the  forebrain  has  been  used  as  a  systematic  char- 
acter in  gadiforms  by  Svetovidov  (1948,  1969)  and  Marshall 
( 1 965).  This  character  develops  during  ontogeny  since  the  bulbs 
are  close  to  the  forebrain  in  young  of  all  gadiforms  (Rass  in 
Svetovidov,  1948;  Marshall,  1965).  It  reaches  the  most  derived 
state  adjacent  to  olfactory  capsules  in  "nearly  all  of  the  Gadi- 
dae," "most  species  of  Macrouridae,"  Muraenolepis  (Marshall. 
1966b)  and  Merluccius  (Inada,  1981b).  Olfactory  lobes  are  be- 
tween the  forebrain  and  olfactory  capsules  in  Bregmaceros  and 
next  to  the  forebrain  in  other  merlucciids  and  Steindachneria 
(Marshall,  1966b). 

We  are  not  certain  how  to  interpret  the  available  information 


'  The  caudal  fin  of  Macruromis  novaezelandiae.  though  much  reduced 
in  over-all  size,  is  similar  to  that  of  Muraenolepis  \n  its  confluence  with 
dorsal  and  anal  fins. 


FAHAY  AND  MARKLE:  GADIFORMES 


283 


on  this  structure  since  an  undescribed  ontogenetic  sequence  is 
involved.  This  character  is  an  important  part  of  our  current 
concept  of  Merlucciidae,  thus  descriptions  of  its  ontogeny  could 
contribute  to  a  better  understanding  of  this  family's  interrela- 
tionships. 

Genilal papilla.— A.  genital  papilla  develops  precociously  in  most 
gadiforms.  It  is  most  pronounced  in  morids.  macrourids  and 
Steindachneria  (see  figures),  but  we  also  could  find  it  in  gadids 
and  Merluccius. 

Mental  barbels.  —  Mental  barbels  usually  develop  late  in  the  lar- 
val or  early  in  the  pelagic  juvenile  period.  They  are  found  in 
most  gadids  (being  lost  in  some  of  the  secondarily  pelagic  forms 
such  as  Pollachhis  pollachiiis  and  Micromesistiiis).  most  ma- 
crourids, muraenolepidids  and  morids.  Additional  fleshy,  snout 
barbels  are  found  in  phycine  rocklings.  The  propensity  to  de- 
velop snout  and  mental  barbels  seems  widespread  in  gadiforms 
and  can  also  be  found  in  some  ophidiiform  fishes;  it  appears  to 
have  a  strong  ecological  component  and  we  are  unable  to  attach 
phylogenetic  significance  to  its  presence  or  absence. 

Gadiform  Phylogeny 

A  framework  of  interrelationships  of  the  eight  gadiform  fam- 
ilies has  developed  from  among  others,  Marshall  ( 1 966b),  Gos- 
line  (1968),  Rosen  and  Patterson  (1969),  Okamura  (1970b)  and 
Marshall  and  Cohen  (1973).  A  concensus  on  minor  as  well  as 
some  major  points  does  not  exist,  and  we  therefore  follow  a 
modification  of  Rosen  and  Patterson  (1969)  and  Cohen  (this 
volume).  In  this  framework  muraenolepidids  are  the  most  prim- 
itive group,  showing  no  obvious  relationships,  and  are  the  pre- 
sumed sister  group  to  all  other  gadiforms.  Based  on  fossil 
evidence  (Danil'chenko,  I960),  bregmacerotids  are  thought  to 
be  related  to  a  group  composed  of  morids  and  melanonids. 
These  three  families  are  the  sister  group  of  macrourids  and 
together  form  a  principal  gadiform  lineage.  Steindachneria  and 
merlucciids  are  sister  groups  and  with  gadids  form  the  other 
principal  gadiform  lineage. 


On  the  basis  of  available  data,  we  can  identify  the  following 
early  developmental  characters,  their  denved  states,  and  known 
distribution  in  the  order.  In  many  cases  the  "holes"  in  our  data 
severely  reduce  the  weight  of  our  arguments.  (1)  Oil  globule  in 
egg— lost  — gadines;  (2)  Chorion  ornamentation  — honey- 
combed—macrourines;  (3)  Lateral  premaxillary  spines— pres- 
ent—muraenolepidids;  (4)  Pterotic  spines— present— some  phy- 
cines;  (5a)  Sequence  of  fin  formation— caudal  first— gadines  and 
Merluccius;  (5b)  Sequence  of  fin  formation  — pel  vies  last— gad- 
ines; (6)  Pelvic  fin  ontogeny— reduction  in  ray  number— phycine 
hakes  and  morids;  and  (7)  Larval  pectoral  fin— pedunculate— 
macrourids  and  Steindachneria.  To  this  list  we  can  add  onto- 
genetically  persistent  characters  taken  in  part  from  Rosen  and 
Patterson  ( 1 969),  Marshall  and  Cohen  (1973)  and  Markle  (1982). 
(8)  X  and  Y  bones— loss  in  forms  with  tails— melanonids,  gad- 
ines and  lotines;  (9)  Total  caudal  fin  rays— over  50  — melan- 
onids, gadines  and  lotines;  (10a)  Anterior  dorsal  fin  rays  to 
centra  ratio  — 7:1  —phycine  rocklings;  (10b)  Anterior  dorsal  fin 
rays  to  centra  ratio— ca.  1:1— gadines,  morids?,  macrourids  and 
merlucciids:  (1 1)  Precaudal  vertebrae— counts  greater  than  20— 
gadines,  lotines,  merlucciids  and  muraenolepidids;  (12)  Hypur- 
als— fusion  into  two  plates— muraenolepidids,  bregmacerotids, 
gadids  and  meriucciids;  (13)  Otophysic  connection  — present— 
morids;  and  (14)  Fin  diflferentiation— three  dorsals  and  two 
anals— gadines,  some  morids  (Merluccius  and  bregmacerotids 
to  a  lesser  degree). 

These  characters  generally  do  not  support  the  above  hypoth- 
eses of  relationships.  Notable  discrepancies  and  areas  for  ad- 
ditional investigation  are:  ( I )  whether  gadids  are  monophyletic, 
specifically  whether  phycines  belong  in  and  Merluccius  belongs 
out;  (2)  relationship,  if  any,  of  melanonids  to  gadines;  and  (3) 
relationships  of  Steindachneria. 

(M.P.F.)  National  Marine  Fisheries  Service,  Northeast 
Fisheries  Center,  Sandy  Hook  Laboratory,  High- 
lands, New  Jersey  07732;  (D.F.M.)  Huntsman  Marine 
Laboratory,  Brandy  Cove,  St.  Andrews,  New 
Brunswick  EGG  2X0  Canada. 


Gadidae:  Development  and  Relationships 
J.  R.  Dunn  and  A.  C.  Matarese 


LARVAE  of  the  fishes  of  the  family  Gadidae  have  received 
a  great  deal  of  study  through  the  years  and  because  of  the 
economic  value  of  the  family,  the  larvae  are  taxonomically  as 
well  known  as  those  of  most  families  of  teleosts.  Svetovidov's 
(1948)  classic  work  on  the  systematics  of  adult  gadid  fishes  is 
the  benchmark  of  knowledge  of  the  family.  He  considered  22 
genera  (including  Merluccius).  examined  osteological  characters 
of  representatives  of  all  genera,  and  based  his  classification 
scheme  mainly  on  the  structure  and  number  of  median  fins  (see 
also  Svetovidov,  1956).  Subsequent  workers  (Mujib,  1967,  1969; 
Marshall  and  Cohen,  1973)  have  extended  our  understanding 


of  the  relationships  of  certain  members  of  the  family,  but  a 
comprehensive  study  of  Gadidae,  including  early  life  history 
stages,  has  not  yet  been  accomplished.  Recently  Markle  (1982) 
examined  larval  and  adult  representatives  of  all  gadoid  families 
which  led  him  to  recognize  three  gadid  subfamilies:  Phycinae, 
Lotinae,  and  Gadinae. 

Our  purpose  here  is  to  summarize  available  knowledge  of  the 
taxonomy  of  eggs  and  larvae  of  the  family  Gadidae.  We  include 
observations  on  eggs,  larval  morphology  and  pigment  patterns, 
and  developmental  osteology.  Included  are  illustrations  of  lar- 
vae of  representatives  of  all  currently  recognized  gadid  genera 


284 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  77.    Summary  of  Egg  Characters  in  Genera  of  the  Family  Gadidae.  All  eggs  are  spherical  in  shape  with  a  homogeneous  yolk. 


Oil  globules 

Pelagic/ 

Size 

Taxon 

demersal 

(mm) 

Number 

Size  (mm) 

Pigment 

Lotinae 

Brosme 

P 

1.29-1.51 

1 

0.23-0.30 

Molva 

P 

0.97-1.13 

1 

0.28-0.31 

yes 

Lota 

D 

1.00-1.90 

1 

yes 

Phycinae 

Enchelyopus 

P 

0.66-0.98 

multiple  to  ! 

0.13-0.20 

yes 

Gaidropsarus 

P 

0.70-0.85 

multiple  to  i 

0.14-0.16 

yes 

Phyas' 

P 

0.76-0.79 

1 

Urophycis 

P 

0.63-0.97 

multiple  to  '. 

0.17-0.20 

yes 

Citiata 

P 

0.67-0.98 

multiple  to  1 

0.11-0.18 

yes 

Rankeps 

P 

0.75-0.91 

1 

0.14-0.19 

yes 

Gadinae 

Trisopterus 

P 

0.90-1.22 

0 

Merlangius 

P 

0.97-1.32 

0 

Pollachius 

P 

1.00-1.22 

0 

Melanogrammus 

P 

1.10-1.72 

0 

Gadus 

P.  D 

0.92-1.90 

0 

Gadiculus 

0.91-0.97- 

Microgadus 

D 

1.39-1.70 

probably  0 

Eleginus 

D 

1.00-1.70 

0 

Boreogadus 

P 

1.53-1.90 

0 

Arctogadus 

D 

Theragra 

P 

1.19-1.81 

0 

Micromesistius 

P 

1.04-1.27 

0 

Russell  (1976),  Schmidt  (1905b) 
Schmidt  (1905b) 

Brederand  Rosen  (1966),  Jude  (1982b),  Morrow 
(1980),  Schmidt  (1907a).  Snyder  (1979) 

Fahay  (1983),  Hardy  (1978a).  Russell  (1976) 

Dekhnik  (1973),  Russell  (1976) 

Wenner(1978) 

Fahay  (1983),  Hardy  (1978a) 

Russell  (1976) 

Kennedy  and  Fitzmaurice  (1969),  Russell  (1976) 

Russell  (1976) 

Dehknik  (1973),  Russell  (1976) 

Fahay  (1983),  Fridgeirsson  (1978),  Russell  (1976) 

Hardy  (1978a),  Russell  (1976) 

Mukhacheva  and  Zviagina  (1960),  Russell  (1976) 

Russell  (1976) 

Hardy  (1978a),  this  study 

Breder  and  Rosen  (1966),  Kozlov  (1952),  Mukhacheva 

(1957),  Mukhacheva  and  Zviagina  (1960) 
Pertseva  (1936).  Rass  (1968),  Russell  (1976) 
Zviagina  (1961) 

Gorbunova  (1954),  Yusa  (1954) 
Lisovenkoet  al.  (1982),  Russell  (1976),  Seaton 

and  Bailey  (1971) 

'  Applies  lo  P.  cfieslen  only. 
'  Ovarian  eggs  only. 


(except  Arctogadus).  Finally,  we  attempt  to  evaluate  the  rela- 
tionships of  the  subfamilies  of  gadid  fishes  based  on  early  life 
history  and  adult  characters. 

Methods 

We  have  examined  developmental  series  of  varying  com- 
pleteness of  representatives  of  all  gadid  genera  except  Phycis 
and  Arctogadus.  of  which  only  juvenile  specimens  were  avail- 
able to  us.  Measurements  were  taken  on  these  series  and  smaller 
series  were  differentally  stained  (Dingerkus  and  Uhler,  1977) 
for  study  of  developmental  osteology. 

Russell  (1976)  described  eggs  and  pigment  patterns  in  gadid 
larvae.  Matarese  et  al.  (1981)  modified  the  terminology  used  by 
Russell  in  describing  postanal  pigment  and  we  use  their  ter- 
minology here. 

Our  discussion  of  osteology  presented  here  is  limited  pri- 
marily to  features  of  the  pectoral  and  pelvic  girdle,  vertebral 
column,  and  median  and  paired  fins.  Svetovidov  (1948)  and 
Mujib  (1967,  1969)  have  described  cranial  osteology.  Matarese 
et  al.  (1981)  and  Markle  (1982)  have  discussed  the  significance 
of  median  fins  in  gadoid  fishes. 

Characteristics  of  Adult  and  Early 
Life  History  Stages 

Family  Gadidae.— Gadid  fishes  possess  four  to  six  pectoral  ra- 
dials;  the  posttemporal  is  attached  to  the  skull  in  adults  and 
possesses  a  ventral  branch  of  varying  length.  An  anterior  process 
of  varying  length  is  present  on  the  coracoid,  but  a  posterior 
process  is  lacking.  The  postcleithrum  is  variously  curved  and, 
in  some  genera,  possesses  an  expanded  distal  head.  The  pelvic 


basipterygia  have  a  posterio-lateral  process  of  varying  length. 
The  first  neural  spine  is  attached  to  the  supraoccipital  crest  in 
adult  fishes;  and  subsequent  anterior  neural  spines  (on  vertebrae 
2-10)  vary  in  length  and  are  oriented  vertically  or  posteriorly. 
One  or  two  predorsal  bones  are  present  in  some  genera  but  are 
absent  in  most.  One  to  three  dorsal  and  one  or  two  anal  fins  are 
present.  When  two  dorsal  fins  are  present,  the  first  may  be 
separate  from  the  second  (intemeural  bones  absent),  or  contin- 
uous (intemeural  bones  present);  when  three  dorsal  fins  are  pres- 
ent the  second  is  always  internally  continuous  with  the  third; 
when  two  anal  fins  are  present,  they  too  are  internally  contin- 
uous. The  distance  (number  of  intemeural  bones)  between  mul- 
tiple dorsal  and  anal  fins  varies  among  genera. 

The  caudal  fin  has  three  hypural  bones  (Matarese  et  al.,  1981) 
including  the  parhypural  (Markle,  1982),  and  four  to  six  "pri- 
mary" caudal  fin  rays  (those  articulating  with  the  superior  hy- 
pural bone  [Markle,  1982]).  Accessory  (.x  and  y)  bones  are  pres- 
ent or  absent,  two  epural  bones  are  present,  uroneural  bones  are 
absent,  two  ural  centra  are  present,  and  the  neural  and  haemal 
spines  on  preural  centra  one  are  broadly  spatulate  in  most  gen- 
era. Matarese  et  al.  ( 1 98 1 )  did  not  detect  fusion  of  hypural  bones 
during  ontogeny  of  Microgadus  proximus,  but  they  hypothe- 
sized that  hypural  2  represented  a  fusion  of  hypurals  2  and  3 
and  that  hypural  3  represented  a  fusion  of  hypurals  4-6.  because 
of  the  presence  of  three  inferior  and  three  superior  hypural 
elements  in  Moridae  (Fitch  and  Barker,  1972),  which  is  generally 
considered  a  more  primitive  family  than  Gadidae. 

Subfamily  Lotinae  (Tables  77-82.  Figs.  145-146.).  — Membtrs 
of  the  Lotinae  are  elongate  gadid  fishes  including  Brosme,  Mol- 


DUNN  AND  MATARESE:  GADIDAE 


285 


Fig.  145.  (A)  Preflexion  larva  of  Bros  me  brosme.  5.9  mm  SL  (Huntsman  Mar.  Lab.,  H-16260.  stored  at  NWAFC);  (B)  Flexion  larva  of  Molva 
moha.  8.2  mm  SL  (Inst.  Sci.  Tech.  Peches  Marit.,  Nantes,  stored  at  NWAFC);  and  (C)  Preflexion  larva  of  Lota  lota.  3.7  mm  SL  (Group  Interuniv. 
Res.  Oceanogr.,  Quebec,  stored  at  NWAFC). 


va,  and  Lota  (MarkJe,  1982).  Brosme  is  monotypic  and  occurs 
on  both  sides  of  the  North  Atlantic  Ocean.  Molva.  with  three 
nominal  species,  occurs  in  the  east  and  west  North  Atlantic 
Ocean  (Svetovidov,  1 948;  Leim  and  Scott,  1 966).  Lola  is  mono- 
typic and  two  subspecies  occur  in  fresh  and  brackish  waters  of 
Europe,  northern  Asia,  and  North  America  (Pivnicka,  1970). 

The  characteristics  of  the  subfamily,  based  on  Markle  (1982) 
and  this  study,  are  egg  diameter  relatively  large  (0.97-1 .90  mm); 
oil  globule  present  (0.2-0.3  mm  diameter);  vertebrae  numerous 
(62-66  total,  20-26  precaudal  in  specimens  examined);  pterotic 
spines  absent;  pelvic  ray  formation  prior  to  notochord  flexion 
but  acquisition  of  adult  complement  delayed;  x  and  v  bones 
usually  absent;  4-5  primary  caudal  fin  rays;  45-54  total  caudal 
fin  rays;  and  numerous  total  dorsal  and  anal  fin  rays  (77-108D 
and  59-75A). 

Eggs  and  larvae  of  lotines  are  reasonably  well  known  (Tables 
77-79).  Brosme  and  Molva  shed  planktonic  eggs  whereas  Lola 
deposits  nonadhesive,  demersal  eggs,  all  with  a  single  oil  globule. 
The  chorion  of  eggs  of  Brosme  has  deep  pits  visible  by  scanning 
electron  microscopy  (Markle,  pers.  comm.'). 

Lotine  larvae  hatch  at  moderate  sizes  (3-4  mm),  yolk  is  ab- 
sorbed at  around  5  mm,  and  notochord  flexion  is  delayed  (9- 
25  mm).  Size  at  transformation  is  large  and  the  duration  of  the 
pelagic  stage  is  extensive  (Table  78).  The  larvae  tend  to  be 
slender  to  moderately  slender  and  taper  toward  the  tail.  Pelvic 
fins  are  precocious  in  Brosme  and  Molva.  but  not  Lola. 

Head  pigment  in  larvae  is  generally  limited  to  the  mouth  and 
dorsal  area  of  the  head.  Gut  pigment  is  sparse,  initially  located 
only  on  the  dorsal  surface.  Brosme  and  Molva  have  pelvic  fins 


'  D.  F.  Markle,  Huntsman  Marine  Laboratory,  St.  Andrews,  New- 
Brunswick,  pers.  comm.,  25  February  1983. 


which  are  pigmented  distally.  Postanal  pigment  patterns  are 
similar  in  Brosme  and  Molva  (Table  79).  Brosme  larvae  have 
two  postanal  bars  and  distinctive  pigment  above  and  below  the 
urostyle  (Fig.  1 45 A).  Although  Molva  does  not  have  a  bar  pat- 
tern initially,  the  dorsal  and  ventral  pigment  eventually  coalesce 
into  two  postanal  pigment  bands,  the  characteristics  of  which 
are  of  taxonomic  value  in  differentiating  species  in  the  genus 
(Fig.  145B).  Preflexion  larvae  (3-7  mm)  of  L.  lota  (lacusiris?) 
in  North  American  waters  were  reported  by  Fish  (1932)  to  lack 
postanal  pigment.  Snyder  (1979),  however,  reported  finding  dor- 
sal and  ventral  postanal  pigment  in  preflexion  larvae  identified 
as  L.  lota,  as  we  did  in  those  we  examined  from  James  Bay, 
Canada  (Fig.  145C). 

Brosme  has  single  dorsal  and  anal  fins  with  a  slight  separation 
between  the  anal  and  caudal  fins  (Markle,  1982).  The  neural 
spine  on  preural  centrum  one  (PU,)  is  distally  flattened,  the 
haemal  spine  on  this  centrum  is  distally  rounded  (Table  82), 
and  \/y  bones  are  absent  (Fig.  146).  Molva  possesses  two  dorsal 
fins  with  only  a  slight  internal  separation.  The  haemal  spine  on 
PU,  is  distally  rounded  and  x/v  bones  are  present  or  absent 
(usually  absent).  Lola  also  possesses  two  dorsal  fins,  with  only 
slight  internal  separation,  and  a  single  anal  fin.  Both  the  neural 
and  haemal  spines  on  PU,  are  distally  flattened  and  the  species 
usually  lacks  x/v  bones,  but  a  reduced  x  and/or  y  bone  is  some- 
times present  (Markle,  1982). 

Subfamily  Phycinae  (Tables  77-82,  Figs.  147-148). -T^xt 
subfamily  Phycinae  was  resurrected  by  Markle  (1982)  who  ex- 
amined seven  species  of  Northwest  Atlantic  gadids  belonging 
to  four  genera:  Enchelyopus,  Gaidropsarus,  Phycis.  and  (Jro- 
phycis.  We  include  also  Ciliata  and,  arbitrarily,  Raniceps  as 
phycines.  Enchelyopus  and  Raniceps  are  each  monotypic;  the 
former  is  found  on  both  sides  of  the  North  Atlantic  Ocean,  the 


286 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  78. 


Summary  of  Morphological  Characters  of  Larvae  of  the  Family  Gadidae.  Proportions  are  expressed  as  percentages  of  standard 

length,  when  possible. 


Standard  length  (mm 

Morphomelncs  (%  SL) 

Eye  diameter 

Head  length 

Body  depth' 

Post- 

Pre- 

Pre- 

Posl- 

Pre- 

Post- 

Pre- 

Taxon 

Halching 

Rexion 

flexion 

juveniie 

Juvenile 

flexion 

flexion 

flexion 

flexion 

flexion 

Lotinae 

Brosme 

4 

14-25 

25-40 

40-60 

>60 

21- 

16 

Molva 

3 

9-14 

14-20 

20-80 

>80 

9 

7 

30 

24 

21 

Lota 

3 

14-19 

19-30 

8 

6 

21-25 

Phycinae 

Enchelyopus 

2 

5-7 

7-18 

18-45 

>45 

10-15 

4-5 

Gaidropsarus 

2 

5-7 

7-12 

12-48 

>48 

11-13 

9-13 

20-30 

20 

30 

Phycis 

7 

5-8 

8-12 

12-30 

>30 

9 

23 

Urophycis 

2 

4-5 

5- 

-40 

>40 

Ciliata 

2 

5-8 

13 

9 

28-32 

27 

25-30 

Raniceps 

3 

7-12 

11 

27 

32 

Gadinae 

Trisopterus 

3 

7-11 

>45 

8 

11 

26 

30 

22 

Merlangtus 

3 

9-13 

13-23 

9 

9 

28 

27 

20 

Pollachius 

3-4 

12-16 

>50 

9 

21 

26 

16 

Melanogrammus 

3-4 

10-16 

16-22 

>90 

10 

Gadus 

3-4 

10-17 

17-25 

25-35 

>35 

5-9 

7-10 

11-27 

22-26 

10-18 

Gadiculus 

? 

7-13 

13-30 

30-40 

>40 

10 

31 

Microgadus 

3 

8-15 

14-28 

28-46 

>46 

8 

8 

22 

32 

19 

Eleginus 

4 

11-17 

17-24 

24-27 

8 

9 

20 

26 

15 

Boreogadus 

6 

11-17 

17-30 

30-45 

>45 

8 

18 

14 

Arclogadus 

? 

28- 

-31 

3 

22-25 

16-19 

Theragra 

3-4 

10-17 

17-25 

25-40 

>40 

7- 

-8 

15 

33 

Micromesislius 

2 

8-13 

>32 

9 

9 

29 

32 

26 

'  At  pectoral  fin  base  when  possible. 

^  Data  between  columns  indicate  no  data  available  for  ditferences  in  pretlexion  and  flexion  larvae. 

'  Data  are  %  HL. 


latter  in  the  eastern  North  Atlantic,  as  are  the  two  nominal 
species  of  Ciliata  (Cohen  and  Russo,  1979).  Gaidropsarus  has 
about  14  nominal  species,  1 1  in  the  North  Atlantic  Ocean  and 
single  species  off  South  Africa,  New  Zealand,  and  Japan.  Phycis 
has  three  nominal  species  occurring  in  both  sides  of  the  North 
Atlantic  Ocean.  About  seven  species  of  Urophycis  are  presently 
recognized  in  the  western  Atlantic  Ocean  from  Canada  to  South 
America  (Svetovidov,  1948). 

Characteristics  of  phycines  according  to  Markle  (1982)  and 
this  study  include:  egg  diameter  small  (0.63-0.98  mm):  multiple 
oil  globules  that  eventually  coalesce  into  a  single  moderately 
sized  globule  (0.1 1-0.22  mm  diameter);  vertebrae  moderately 
numerous  (45-55  total,  14-17  precaudal);  pterotic  spines  pres- 
ent in  some  larvae  and  juveniles  of  Gaidropsarus,  Ciliata.  and 
Phycis;  initial  pelvic  fin  ray  formation  prior  to  flexion  but  ac- 
quisition of  adult  complement  delayed;  x  and  y  bones  present 
(sometimes  absent  in  Raniceps);  5-6  primary  caudal  fin  rays 
and  29-38  total  caudal  fin  rays;  and  moderate  numbers  of  total 
dorsal  and  total  anal  fin  elements  [D,  49-73,  not  including  spe- 
cialized rays  (e.g.,  Enchelyopus),  and  A,  40-57]. 

Taxonomic  problems  are  prevalent  in  this  group.  Specific 
identification  of  smaller  larvae  is  not  presently  possible  for  cer- 
tain species  of  Gaidropsarus.  Phycis.  and  Urophycis  (Russell, 
1976;  Markle,  1982).  Adultsof  some  species  are  easily  confused 
(Musick,  1973;  Svetovidov.  1982). 

Eggs  of  Enchelyopus  and  Ciliata.  some  Urophycis.  as  well  as 
those  oi  Gaidropsarus  that  are  known,  have  multiple  oil  globules 
in  the  earliest  stages,  which  coalesce  into  a  single  oil  globule; 


melanistic  pigment  is  present  on  both  the  embryo  and  oil  globule 
(Table  77).  In  Phycis.  only  ovarian  eggs  of  P.  chesteri  have  been 
described  (Wenner,  1978).  Multiple  oil  globules  have  not  been 
observed  in  U.  tenuis  (Markle-)  nor  reported  in  eggs  oi Raniceps 
raninus. 

Phycine  larvae  hatch  at  small  sizes  (1.5-3.0  mm),  yolk  is 
absorbed  quickly,  notochord  flexion  occurs  at  small  sizes  (about 
5-12  mm),  size  at  transformation  is  variable,  and  a  prejuvenile 
stage  is  present  in  most  genera  (Table  78).  Phycine  preflexion 
larvae  tend  to  be  deeper  bodied  (at  the  pectoral  fin  base)  than 
lotines  or  gadines,  but  with  development  they  become  mor- 
phologically diverse.  Pelvic  fins  are  precocious  and  pigmented, 
although  the  extent  and  duration  of  pigmentation  varies  among 
genera.  The  entire  fin  is  pigmented  in  Gaidropsarus.  whereas 
only  the  tip  is  pigmented  in  some  species  of  Urophycis.  Raniceps 
is  morphologically  the  most  divergent  phycine.  By  5  mm,  the 
preanal  portion  of  the  body  is  usually  high  in  relation  to  the 
postanal  region  and  at  a  length  of  about  7.5  mm  the  larvae  are 
"tadpole  shaped." 

Although  pigmentation  is  highly  variable  in  phycines,  most 
genera  possess  head  pigment  on  the  dorsal  part  (sometimes 
extending  to  the  nape),  and  on  the  snout  and  mouth.  In  addition, 
some  genera  may  have  pigment  near  the  eye  and  on  the  opercular 
area.  Gut  pigment  is  initially  located  along  the  dorsal  surface, 
with  some  genera  (e.g.,  Enchelyopus.  Phycis)  developing  more 
pigment  over  the  lateral  surface.  Postanal  pigment  is  variable. 


'  D.  F.  Markle,  pers.  comm.,  5  July  1983. 


DUNN  AND  MATARESE:  GADIDAE 


287 


Table  78.    Extended. 


Morphomelncs 

(%SL) 

Body  depth 

Preanal  length 

Approximate 

Posi- 

Pre- 

Posl- 

Lateral 

Precocious 

pelvic  fin 

flexion 

flexion 

flexion 

vent 

pelvic  fin 

fonnation 

35 

48 

yes 

yes 

6 

15 

43 

48 

yes 

5 

47 

yes 
yes 

no 
yes 

11 
3 

20 

60 

37 

45 

yes 
yes 
yes 

yes 
yes 
yes 

<5 

3-4 

<25 

44 

50 

yes 

yes 
yes 

3 

5 

26 

50 

no 

11 

48 

43 

no 

9 

20 

40 

no 
no 

12 

9 

14-20 

35-50 

47-54 

no 

13 

28 

48 

7-8 

41-48 

45 

yes 

no 

14 

20 

42 

49 

yes 

no 

12 

38 

42-46 

yes 

no 

13 

30 

50 

yes 

no 

14 

23 

51 

41 

no 

11 

Fahay  (1983),  Russell  (1976),  this  study 
Russell  (1976),  Schmidt  (1906b),  this  study 
Fish  (1932),  Jude  (1982b),  Snyder  (1979) 

Fahay  (1983),  Hardy  (1978a),  Russell  (1976) 

Demir  (1982),  Markle  (1982) 

D'Ancona  (1933a),  Fahay  (1983),  Russell  (1976) 

Fahay  (1983),  Hardy  (1978a) 

Russell  (1976),  this  study 

Russell  (1976),  Schmidt  (1907b).  this  study 

Russell  (1976),  Schmidt  (1905a,  1906a),  this  study 

Russell  (1976),  this  study 

Fahay  (1983),  Russell  (1976),  this  study 

Fahay  (1983),  Scott  (1982) 

Mukhacheva  and  Zviagina  (1960),  this  study 

Russell  (1976),  Schmidt  (1905a,  1906a),  this  study 

Matareseet  al.  (1981) 

Dunn  and  Vinter(1984) 

This  study 

Zviagina  (1961) 

Dunn  and  Vinter  (1984),  this  study,  (T.  Nishiyama,  pers. 

comm.  July  15,  1982) 
Russell  (1976),  Schmidt  (1905a),  Seaton  and  Bailey  (1971) 


both  between  and  within  genera  (Table  79,  Fig.  147A-F).  At 
some  size,  phycine  larvae  usually  have  a  single  postanal  pigment 
bar  located  about  midtrunk,  but  Ciliata  has  two  bars  (which 
disappear  during  ontogeny)  and  Raniceps  has  none.  Phycts  lar- 
vae less  than  4.3  mm  in  length  are  not  known,  but  larger  post- 
flexion  larvae  have  a  single  midtrunk  patch  of  pigment.  The 
location  of  the  pigment  bar  varies  among  species  of  Gaidrop- 
sarus.  Postanal  pigment  spots  along  the  ventral  body  midline 
occur  in  Raniceps  (anteriorly)  and  in  Ciliata.  Caudal  pigment 
can  be  present  or  absent  and  may  be  taxonomically  significant 
at  the  species  level. 

Phycines  have  two  dorsal  fins  and  one  anal  fin  (Svetovidov, 
1 948);  the  first  and  second  dorsal  fins  are  only  slightly  separated 
(Table  81).  A  predorsal  bone  is  present  in  Urophycis,  Phycis 
chestcri  (two  in  P.  blennoides)  and  Raniceps.  but  is  wanting  in 
the  other  genera.  X/Y  bones  are  present  (Fig.  148 A),  or  usually 
present  in  Raniceps  (Fig.  148B).  Neural  and  haemal  spines  on 
PU,  are  distally  flattened  except  in  Raniceps,  in  which  those 
bones  are  distally  rounded.  We  detected  evidence  of  ontogenetic 
fusion  of  the  hypural  bones  in  Raniceps.  as  hypurals  2  and  3 
are  bifurcate  distally  (Fig.  148),  and  a  sixth  hypural  bone  was 
found  in  one  larva. 

Subfamily  Gadinae  (Tables  77-82.  Figs.  149-1 5 D.  —  lhis 
subfamily  contains  the  "true  cods."  There  are  approximately 
30  nominal  species  presently  assigned  to  twelve  genera.  They 
are  found  in  the  North  Atlantic,  North  Pacific  and  Arctic  oceans 
except  for  a  single  species,  Micwmesistius  australis.  which  is 
distributed  in  the  western  South  Atlantic  Ocean  to  60°S  (Merrett, 
1963;Shust,  1978). 


The  subfamily  Gadinae  is  characterized  as  follows  [MarkJe 
(1982);  this  study]:  egg  diameter  relatively  large  (0.9  to  1.9  mm); 
no  oil  globule;  vertebrae  moderately  numerous  (39-64  total 
vertebrae,  1  7-26  precaudal);  pterotic  spines  absent;  pelvic  fin 
ray  formation  at  the  same  time  as  notochord  flexion;  .\  and  v 
bones  absent;  4-5  primary  caudal  fin  rays;  46-70  total  caudal 
fin  rays;  relatively  few  total  dorsal  and  total  anal  fin  elements 
(D,  45-67  and  A,  35-65). 

Eggs  of  gadines  are  well  known,  with  eggs  of  one  or  more 
species  of  each  genus  described,  except  for  Gadiculus  and  Arc- 
togadus  (Table  77).  Most  species  shed  small,  planktonic  eggs, 
but  demersal  eggs  are  deposited  by  a  number  of  species  (Gadus 
macrocephalus  and  both  species  of  Microgadus.  Eleginus.  and, 
presumably,  Arctogadus).  Characteristic  pigment  develops  on 
late  stage  embryos  which  aids  in  their  identification. 

Gadine  larvae  are  also  well  known  except  for  Arctogadus  (Ta- 
bles 78,  79).  Length  at  hatching  ranges  from  2  to  6  mm,  yolk 
absorption  (when  known)  occurs  relatively  early,  notochbrd 
flexion  occurs  from  about  7  to  1  7  mm,  and  transformation  to 
the  juvenile  stage  occurs  at  about  25-40  mm.  The  duration  of 
the  pelagic  state  is  moderate  to  long.  Preflexion  larvae  typically 
are  moderately  slender,  tapering  toward  the  tail,  while  flexion 
larvae  tend  to  be  more  robust  (Table  78). 

Head  pigment  is  more  diverse  and  diagnostically  more  im- 
portant in  this  subfamily  than  in  the  lotines  or  phycines.  Larvae 
of  most  genera  have  pigment  on  the  dorsal  head  and  on  the 
mouth  (usually  the  dentary).  In  some  genera,  the  presence  (e.g., 
Eleginus)  or  absence  (e.g.,  Boreogadus)  of  gular  and  isthmus 
pigment  is  important  in  identification.  Absence  of  ventral  gut 
melanophores  in  certain  size  larvae  (e.g.,  Boreogadus,  Melano- 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  146.  Caudal  fin  of  Brosme  hrosme.  45.2  mm  SL.  Hyl  =  Hypural  bone  1;  Hy2-3  =  Hypural  bones  2  and  3;  Hy4-6  =  Hypural  bones  4, 
5,  and  6;  EP,  =  Epural  bone  1;  EP,  =  Epural  bone  2;  U,  =  Ural  centrum  1;  U;  =  Ural  centrum  2;  PU,  =  Preural  centrum  1;  PU|„  =  Preural 
centrum  10.  (Huntsman  Mar.  Lab.,  H  9742,  stored  at  NWAFC). 


grammus)  or  the  presence  and  distribution  of  such  pigment  (e.g., 
Pollachius  virens.  P.  pollachius)  is  also  of  diagnostic  value  (Rus- 
sell, 1976;  Matarese  et  al.,  1981;  Dunn  and  Vinter,  1984). 

Pigment  in  the  postanal  region  is  also  diverse  and  usually  of 
value  in  discriminating  among  species  (e.g.,  Russell,  1976).  For 
purposes  of  discussion  here  we  divide  the  gadines  into  three 
groups  based  on  their  postanal  pigment  patterns:  those  genera 
without  postanal  pigment  bars  in  preflexion  larvae,  those  genera 
in  which  individual  species  may  or  may  not  possess  such  bars, 
and  those  genera  possessing  one  or  two  postanal  pigment  bars 
(Table  79). 

Merlangius  and  Melanogrammus  lack  postanal  pigment  bars 
(Fig.  149C,  D).  In  Merlangius.  postanal  pigment  develops  along 
the  dorsal  body  midline  and  extends  to  nearly  three-quarters 
the  length  of  the  body.  Ventral  pigment  consists  of  a  row  of 
melanophores  from  the  anus  to  the  caudal  fin.  Preflexion  larvae 
of  Afelanogratnmus  lack  the  dorsal  line  of  pigment,  but  possess 
the  continuous  ventral  line. 

Within  Trisopterns  and  Pollachius,  some  species  have  one  or 
two  postanal  pigment  bars,  whereas  others  lack  such  bars  [7'. 
esmarkii.  T.  minutus.  and  P.  pollachius (RusseW.  1976)].  In  those 
species  possessing  postanal  bars,  T.  luscus  (Fig.  149 A)  has  a 
single  pigment  bar,  with  dorsal  and  ventral  midline  pigment 
extending  to  about  one-half  the  postanal  body  (Russell,  1976). 
Pollachius  virens  (Fig.  149C)  has  two  postanal  pigment  bars,  the 


anterior  of  which  is  close  to  the  vent.  Of  those  without  pigment 
bars,  T.  minutus  and  P.  pollachius  possess  dorsal  and  ventral 
lines  of  pigment  extending  to  about  three-quarters  of  the  body 
length;  caudal  peduncle  pigment  may  be  present  in  certain  size 
larvae  of  the  former  species,  but  is  normally  lacking  in  the  latter. 
Gadiculus  has  one  postanal  pigment  bar  located  posterior  to 
the  midpoint  of  the  postanal  region  whereas  Micromesistius  has 
a  single  bar  near  the  midpoint  of  this  region  (Figs.  149F  and 
1 50E).  The  dorsal  stripe  is  slightly  longer  than  the  ventral  stripe. 
Mediolateral  pigment  between  the  dorsal  and  ventral  bars  de- 
velops during  ontogeny,  but  the  caudal  peduncle  area  is  not 
pigmented.  Gadus  (in  those  species  whose  larvae  are  known), 
Microgadus.  Eleginus,  Boreogadus,  and  Thcragra  have  two 
postanal  bars  of  pigment  (not  known  for  Arctogadus)  as  shown 
in  Figs.  149E  and  150A-C.  In  some  genera  (e.g.,  Boreogadus) 
the  dorsal  stripe  of  each  bar  is  longer  than  the  ventral  stripe;  in 
others,  the  ventral  stripe  is  longer  than  the  dorsal  (Gadus.  Mi- 
crogadus. Eleginus,  and  Theragra).  The  anterior  end  of  the  ven- 
tral stripe  may  be  near  the  anus  (e.g.,  Gadus),  or  some  distance 
from  It  (e.g.,  Boreogadus).  and  the  ventral  stripes  may  be  com- 
posed of  a  single  row  of  melanophores  on  each  side  of  the  body 
midline  (e.g.,  G.  macrocephalus),  a  double  row  on  each  side  of 
the  midline  (E.  gracilis),  or  on  the  ventral  midline  with  scattered 
pigment  on  each  side  of  the  body  {B.  saida).  Caudal  peduncle 
pigment  may  be  present  or  absent. 


DU>fN  AND  MATARESE:  GADIDAE 


289 


Table  79.    Selected  Pigmentation  Characters  Useful  in  Identifying  Preflexion  and  Flexion  Larvae  of  the  Family  Gadidae. 


Postanal  pigment 


Bars  Nuinber        Dorsal  Ventral 

Flexion  (any  of  stnpcs  smpes 

(mm)  size)  bars        continuous    continuous       Description  of  pigment 


Hypural 
inargin 


Pelvic 
fins 


Diagnostic 


Lotinae 
Brosme 

Molva 
Lota' 


Phycinae 
Enchelyopits 


Gaidropsarus 
Phycis 


Urophycis 

Ciliala 

Raniceps 

Gadinae 
Tnsopterus 


Merlangius 


Poltachius 


14-25         Yes 


9-14 


14-19 


Yes 


No 


May 


No 


5-7  Yes  1 


5-7  Yes  1 


5-8  No?2 


4-5  Yes 


5-8  Yes 


7-12         No 


7-1  1       Yes/no 


8-13         No 


11-16      Yes/no 


No 


Yes 


Yes  Yes 

(1  1  mm)  (14  mm) 


Melanogrammus     10-16         No 


Within  bars  only 

Above  and 

Yes 

Caudal,  2  bars, 

Fahay  (1983), 

below 

pelvics 

Russell 

(1976) 

May  occur  along 

May  occur 

Yes 

Caudal  when 

Russell  (1976). 

dorsal  midline. 

above 

present,  2 

Schmidt 

mediolateral 

and  be- 

bars, pelvics 

(1906b, 

within  bars 

low 

1907a) 

Along  dorsal  and 

No 

No 

Lack  of  pig- 

Fish (1932), 

ventral  midline. 

ment  in  lar- 

Hardy 

mediolateral 

vae  <6  mm, 
or  dorsal/ 
ventral  rows 

(1978a), 

Jude 

(1982b), 

Snyder 

(1979) 

Bar  at  mid-trunk 

Yes 

Yes 

Mid-trunk  bar, 
pelvics 

Fahay  (1983), 
Hardy 
(1978a), 
Russell 

(1976) 

Bar  location  vari- 

May 

Yes 

Bar  location. 

Demir(1982), 

able,  dorsolat- 

presence of 

Markle 

eral  at  bar 

caudal 

(1982) 

Several  spots 

Yes 

Yes 

Lack  of  dorsal 

D'Ancona 

along  ventral 

initially,  pel- 

(1933), Fa- 

midline 

vic 

hay  (1983), 

Hardy 

(1978a), 

Russell 

(1976) 

Variable,  but  mid- 

May 

Yes' 

Postanal  pat- 

Fahay (1983), 

trunk  patch,  me- 

tern, pelvic 

Hardy 

diolateral  at  bar 

tips 

(1978a) 

Bars  disappear. 

Yes 

Yes 

Loss  of  dorsal 

Russell  (1976), 

ventral  midline 

pigment,  pel- 

this study 

only 

vics 

Anterior  ventral 

No 

Yes 

LIpper  body 

Russell  (1976), 

midline,  upper 

pigment,  pel- 

Schmidt 

body 

vics 

(1907b) 

Variable  length 

No 

No 

Lack  of  bars 

Russell  (1976), 

dorsal,  ventral 

(some 

Schmidt 

midline,  medio- 

species) re- 

(1905a, 

lateral  within 

duced  dorsal 

1906a) 

bar 

midline 

Dorsal  (shorter) 

Yes 

No 

Length  of  dor- 

Russell (1976) 

and  ventral 

sal  midline 

midline  dorso- 

ventrolateral 

Dorsal,  and  dou- 

No 

No 

Unpigmented 

Fahay  (1983), 

ble  ventral  mid- 

posterior 1/4 

Fridgeirsson 

line,  extends 

body 

(1978), 

3/4,  some 

Hardy 

mediolateral 

(1978a), 

Russell 

(1976) 

Small,  double 

Yes 

No 

Ventral  mid- 

Fahay (1983), 

ventral  midline 

line,  nape 

Hardy 

(1978a), 

Russell 

(1976) 

290 


ONTOGENY  AND  SYSTEM ATICS  OF  FISHES -AHLSTROM  SYMPOSIUM 


Table  79.    Continued. 


Postanal  pigment 


Flexion 
(mm) 


Bars 

(any 
size) 


Number 
of 
bars 


Dorsal 

stnpes 

contmuous 


Ventral 

stnpes 

continuous 


Descnption  of  pigment 


Hypural 
margin 


Pelvic 

tins 


Diagnostic 


Gadus 


10-17 


Yes 


Yes 

{6  mm) 


Yes 

(6  mm) 


Initially  posterior 
stripes  longer, 
mediolateral 


Gadiculus 

7-13 

Yes 

1 

Bar  posterior  to 
mid-trunk 

Microgadm 

8-15 

Yes 

2 

Yes 
(15  mm) 

Yes 

(6  mm) 

Bars  anterior,  me- 
diolateral 

Eleginus 

11-17 

Yes 

2 

Yes 
(10  mm) 

Yes 

(7  mm) 

Ventral  stripes 
longer,  double 
ventral  row 
each  side  of 
midline,  medio- 
lateral 

Boreogadus 

11-17 

Yes 

2 

Yes 

Yes 

Ventral  stripe 

(7  mm)  ( 10  mm) 

shorter,  medio- 

lateral 

Arctogadus 

Dorsal,  ventrolat- 
eral margins, 
mediolateral 

Theragra 

10-17 

Yes 

2 

Yes 

(13  mm) 

May 

Posterior  ventral 
stripe  longer 
than  dorsal,  sin- 
gle ventral  row 
each  side  mid- 
line, medio- 
lateral 

Mkromesislius 

8-13 

Yes 

1 

Bar  mid-trunk, 
dorsal  stnpe 
longer,  nape 

May 


No 


No 


Yes 


No 


No 


No 


No 


No 


No 


May 


No 


No 


No 


Continuous 

Dunn  and 

stnpes,  ven- 

Vinter 

tral  gut 

(1984),  Ma- 

tarese  et  al. 

(1981),  Mu- 

khacheva 

and  Zvi- 

agina(1960) 

Schmidt 

(1906a) 

Posterior  loca- 

Russell (1976) 

tion  of  bar 

Schmidt 

(1905a, 

1906a) 

Bar  location. 

Matarese  et  al. 

ventral  gut. 

(1981) 

caudal 

Continuous 

Dunn  and 

stripes. 

Vinter 

mediolateral, 

(1984) 

ventral  gut 

Bars,  medio- 

Dunn and 

lateral 

Vinter 

(1984), 

this  study 

Zviagina 

(1961) 

Bar  location. 

Dunn  and 

length 

Vinter 

(1984),  Gor- 

bunova 

(1954),  Ma- 

tarese et  al. 

(1981) 

Bar  location, 

Lisovenko  et 

upper  body 

al.  (1982), 

Russell 

(1976), 

Schmidt 

(1905a), 

Weiss 

(1974) 

'  See  lext  for  discussion  of  pigmenlation  in  Lota 

'  In  Phycines.  Ihe  bar  occurs  early.  Specimens  of  Ph\ 

'  Pelvic  fins  not  pigmented  in  Li  regia. 


/.s  3-4  mm  not  available  to  us. 


Gadines  have  three  dorsal  fins  and  two  anal  fins.  The  distances 
(number  of  intemeural  bones)  between  dorsal  fins  2  and  3  and 
between  anal  fins  1  and  2  vary  among  genera.  In  all  genera,  the 
dorsal  and  anal  fins  are  separate  from  the  caudal  fin.  The  lower 
branch  of  the  posttemporal  tends  to  become  more  elongate  than 
in  lotines  or  phycines  (longest  in  Arctogadus  and  Micromesis- 
tius),  the  postcieithntm  always  has  an  expanded  distal  head,  and 


the  posterior  process  of  the  basipterygia  tends  to  be  short  or 
even  lacking.  Predorsal  bones  are  absent  in  Gadinae.  The  first 
and  second  dorsal  fins  are  not  usually  internally  continuous,  but 
the  second  and  third  dorsal  fins  are  always  internally  continuous. 
The  neural  and  haemal  spines  on  PU,  are  distally  flattened  in 
all  gadine  genera  and  only  three  hypural  bones  (including  the 
parhypural)  are  present  (Table  82,  Fig.  151). 


Fig.  147.  (A)  Preflexion  larva  of  Enchelyopus  cimbrius.  3.7  mm  SL  (Huntsman  Mar.  Lab.,  H-5388,  stored  at  NWAFC);  (B)  Flexion  larva  of 
Gaidropsarus  mediterraneus,  6.1  mm  SL  (from  Demir,  1982);  (C)  Flexion  larva  of  Phycis  blennoides.  4.3  mm  SL  (from  Russell,  1976);  (D) 
Flexion  larvae  of  Urophycis  sp.,  4.2  mm  SL  (Huntsman  Mar.  Lab.,  H- 16384,  stored  at  NWAFC);  (E)  Flexion  larva  of  Ciliata  sp.,  4.4  mm  SL 
(Zool.  Mus.  Copenhagen,  stored  at  NWAFC);  and  (F)  Preflexion  larva  of  Raniceps  raninus,  4.7  mm  SL  (Inst.  Sci.  Tech.  Peches.  Marit.  Nantes, 
stored  at  NWAFC). 


DUNN  AND  MATARESE:  GADIDAE 


291 


B 


292 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


^e-  Hy  4-6 


Fig.  148.  (A)  Caudal  fin  oi  Phyas  blennoides.  82.3  mm  SL  (British  Mus.  Nat.  Hist.  1976.  7.30.110-119);  Hyl  =  Hypural  bone  1;  Hy2-3  = 
Hypural  bones  2  and  3;  Hy4-6  =  Hypural  bones  4,  5,  and  6;  EP,  =  Epural  bone  1,  EP,  =  Epural  bone  2;  X  =  x  bone,  Y  =  y  bone;  U,  =  Ural 
centrum  1;  LI,  =  Ural  centrum  2;  PU,  =  Preural  centrum  1;  PU^  =  Preural  centrum  6;  (B)  Caudal  fin  oi  Raniceps  raninus.  44.4  mm  SL  (British 
Mus.  Nat.  Hist.  1971.2-16.640);  symbols  as  m  (A). 


Fig.  149.  (A)  Flexion  larva  of  Trisopterus  luscus.  7.5  mm  SL  (Inst.  Sci.  Tech.  Peches  Marit.,  Nantes,  stored  at  NWAFC);  (B)  Preflexion  larva 
of  Merlangius  merlangus.  5.0  mm  SL  (Inst.  Sci.  Tech.  Peches  Mant.,  Nantes,  stored  at  NWAFC);  (C)  Preflexion  larva  of  Pollachius  virens.  5.9 
mm  SL  (Huntsman  Mar.  Lab.,  H-8057,  stored  at  NWAFC);  (D)  Preflexion  larva  of  Melanogrammus  acglefinus.  6.1  mm  SL  (Huntsman  Mar. 
Lab.,  H-9473,  stored  at  NWAFC);  (E)  Preflexion  larva  of  Gadus  macrocephalus,  4.4  mm  SL  (from  Dunn  and  Vinter,  1984);  and  (F)  Preflexion 
larva  of  Gadiculus  argenteus.  3.7  mm  SL  (Zool.  Mus.  Copenhagen,  stored  at  NWAFC). 


DUNN  AND  MATARESE:  GADIDAE 


293 


B 


294 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  80. 


Summary  of  Osteological  Characters  of  the  Pectoral  and  Pelvic  Girdles,  Axial  Skeleton,  and  Median  Fins  in  Pre- 

TRANSFORMATION  LaRVAE  OF  REPRESENTATIVES  OF  THE  GeNERA  OF  THE  FAMILY  GaDIDAE. 


Relative  length  of  lower 
fork  of  posttemporal 


Shape  of 
poslcleithrum 


Length  of  postenor  process 
of  bastpterygia 


Number 

pre-      Number 
dorsal     dorsal 
bones        fins 


Lotinae 
Brosme  brosme 

Molva  dipterygia 

Lola  Iota' 

Phycinae 

Enchelyopus  cimbrius 

Gaidropsarus  sp. 
Phycis  blennoides' 
Vrophycis  sp. 
Ciliata  must  el  la 
Raniceps  raninus 

Gadinae 

Trisopterus  liiscus 

Merlangius  merlangus 

Pollachius  virens 

Melanogrammus 
aeglefinus 

Gadus  macrocephalus 

Gadiculus  argenteus 

Microgadus  proximus 
Eleginus  gracilis 
Boreogadus  saida 

Arctogadus  horisovi' 

Theragra  chalcogramma 

Micromesislius 
poutassou 


Very  short,  less  than  '/,o  length 
of  upper  fork 

Short,  less  than  'A  length  of 
upper  fork 

Short,  less  than  %  length  of 
upper  fork 


Short,  less  than  '/j  length  of 

upper  fork 
Short,  less  than  %  length  of 

upper  fork 
Short,  less  than  V,  length  of 

upper  fork 
Short,  less  than  %  length  of 

upper  fork 
Very  short,  less  than  y,,,  length 

of  upper  fork 
Short,  less  than  '/,  length  of 

upper  fork 


Moderately  long,  about  Vi 

length  of  upper  fork 
Moderately  long,  about  % 

length  of  upper  fork 
Moderately  long,  about  V^ 

length  of  upper  fork 
Moderately  long,  about  % 

length  of  upper  fork 

Moderately  long,  about  '/, 

length  of  upper  fork 
Long,  about  %  length  of  upper 

fork 

Long,  about  %  length  of  upper 

fork 
Moderately  long,  about  % 

length  of  upper  fork 
Long,  about  'A  length  of  upper 

fork 

Long,  about  -/,  length  of  upper 

fork 
Long,  about  y,  length  of  upper 

fork 
Long,  about  -/,  length  of  upper 

fork 


Long,  thin,  slightly  curved, 
distal  head  slightly  expand- 
ed 

Long,  thin,  slightly  curved, 
distal  head  slightly  expand- 
ed 

Long,  thin,  slightly  curved, 
distal  head  wanting 

Moderately  long,  thin,  not 

curved,  no  distal  head 
Long,  thin,  not  curved,  no 

distal  head 
Long,  thin,  slightly  curved,  no 

distal  head 
Long,  thin,  curved,  no  distal 

head 
Short,  thin,  no  distal  head 

Lower  bone  long,  thin,  point- 
ed; upper  bone  short,  ob- 
long 

Long,  thin,  slightly  curved, 
slightly  expanded  head 

Long,  thin,  straight,  expanded 
distal  head 

Long,  thin,  curved,  with  ex- 
panded distal  head 

Long,  thin,  slightly  curved, 
slightly  expanded  distal 
head 

Long,  thin,  strongly  curved, 
expanded  distal  head 

Short,  relatively  wide,  pointed 
distally,  slightly  expanded 
head 

Long,  straight,  recurved  dis- 
tally, with  expanded  head 

Long,  straight,  recurved  dis- 
tally, with  expanded  head 

Long,  thin,  slightly  recurved 
distally,  moderately  ex- 
panded head 

Long,  thin,  striaght,  expanded 
head 

Long,  thin,  recurved,  expand- 
ed distal  head 

Long,  thin,  recurved,  expand- 
ed distal  head 


Long,  about  -/j  length  anterior      0 
process 

Moderate,  about  Vi  length  an-      0 
terior  process 

Short,  about  '/,  length  anterior     0 
process 

Very  long,  about  2  x  length  0 

anterior  process 
Long,  about  -/,  length  anterior      0 

process 
Very  long,  about  1 '/.  x  length        2 

anterior  process 
Very  long,  about  4  x  length  1 

anterior  process 
Very  long,  about  I  'A  x  length        1 

anterior  process 
Long,  about  74  length  anterior       1 

process 


Absent 


Very  short,  less  than  '/lo 
length  anterior  process 


Very  short,  less  than  '^j  0 

length  anterior  process 
Absent  0 


0 


Moderate,  about  '/;  length  an-      0 

terior  process 
Short,  about  '/,  length  antenor      0 

process 

Very  short,  about  '/,o  length  0 

anterior  process 
Short,  about  %  length  anterior     0 

process 
Short,  about  %  length  anterior     0 

process 

Short,  about  %  length  ante-         0 

rior  process 
Short,  about  '/i  length  anterior     0 

process 
Absent  0 


'  Juvenile  specimens  only  examined. 


Comments  on  the  Systematic  Relationships 

OF  Subfamilies  in  the 

Family  Gadidae 

Gadoid  fishes  comprise  a  coinplex  and  rather  confusing  array 
of  teleosts  possessing  both  relatively  primitive  and  apparently 
derived  character  states  (Cohen,  this  volume;  Fahay  and  Markle, 
this  volume).  In  our  analyses  of  character  states  we  generally 
follow  Markle  (1982),  Fahay  and  Markle  (this  volume)  and 


Cohen  (this  volume).  For  outgroup  comparisons,  we  have  ex- 
amined the  osteology  of  representatives  of  a  limited  array  (14 
families)  of  gadiform  and  non-gadiform  fishes  whose  utility  is 
limited  because  we  lack,  in  many  cases,  ontogenetic  series.  We 
contrast  here  the  characters  of  the  three  recognized  subfamilies 
(Markle,  1 982)  with  Merluccius.  insofar  as  possible,  as  the  genus 
is  variously  considered  primitive  in  the  gadid-merlucciid  lineage 
(Danil'chenko,  1947,  1950;  Rosen  and  Patterson,  1969;  Cohen, 


Table  81. 


Summary  of  Osteological  Characters  of  the  Median  Fins  in  Pretransformation  Larvae  of  Representatives  of  the  Genera 

OF  THE  Family  Gadidae. 


Relative  dislance  between 
dorsal  fins  one  and  two 


Relative  distance  between 
dorsal  fins  two  and  three 


Relative  distance  between 

postenormost  dorsal  and 

caudal  fin 


Number 

of  anal  Relative  distance  between 

fins  anal  fins  one  and  two 


Lotinae 
Brosme  brosme 

Molva  dipten'gia 
Lota  lota' 

Phycinae 

Enchelyopus  cimbrius 

Gaidropsarus  sp. 

Phycis  blennoides' 

Urophycis  sp. 
Ciliala  muslella 

Raniceps  raninus 


Gadinae 

Tnsopterus  luscus 

Merlangius  merlangus 

Pollachius  virens 


Melanogra  mmiis 
aeglefinus 

Gadus  macrocephatus 


Gadiculus  argenteus 


Microgadus  proximus 


Eleginus  gracilis 


Boregadus  saida 


Arclogadus  borisovi' 


Theragra 
chalcogramma 


Micromesislius 
poutassou 


Very  close,  about  '/,  in- 
temeural  space;  no  in- 
temeural  bones 

Very  close,  about  '/,  in- 
teraeural  space;  1  or  0 
intemeural  bones 


Wide,  about  2  inter- 
neural  spaces;  no  in- 
temeural bones 

Very  close,  about  'A  in- 
temeural space;  no  in- 
temeural bones 

Very  close,  about  '/j  in- 
temeural space;  1  in- 
temeural bone 

Nearly  continuous;  I  in- 
temeural bone 

Very  close,  about  '/,  in- 
temeural space;  I  in- 
temeural bone 

Close,  about  1  inter- 
neural  space;  I  re- 
duced intemeural 
bone 


Close.  '/;-l  intemeural 
space;  0  or  1  inter- 
neural  bones 

Close,  usually  '/,-l  inter- 
neural  space;  2  inter- 
neural  bones 

Close,  about  'A  inter- 
neural  space;  no  inter- 
neural  bones 

Close,  about  '/,  inter- 
neural  space;  0-2  in- 
temeural bones 

Close,  about  '/,  inter- 
neural  space;  0  or  1 
intemeural  bones 

Moderately  wide,  about 
I  intemeural  space; 
no  intemeural  bones 

Close,  about  '/,  inter- 
neural  space;  0  or  I 
intemeural  bones 

Moderately  wide,  about 
I '/,  intemeural  spaces; 

0  or  1  intemeural 
bones 

Wide,  about  2  inter- 
neural  spaces;  0  or  I 
intemeural  bones 

Wide,  about  2'/,  inter- 
neural  spaces;  no  in- 
temeural bones 

Moderately  wide,  about 

1  Vj  intemeural  spaces; 
0  or  I  intemeural 
bones 

Wide,  about  2  inter- 
neural  spaces;  no  in- 
temeural bones 


Very  close  together,  0-2 
intemeural  bones 

Close  together.  2-4  in- 
temeural bones 

Close  together,  3-4  in- 
temeural bones 

Close  together,  2-3  in- 
temeural bones 

Close  together,  2-3  in- 
temeural bones 

Moderately  wide,  about 
5-6  intemeural  bones 

Moderately  wide,  4-7 
intemeural  bones 

Moderately  wide,  5-7 
intemeural  bones 


Wide,  6-7  intemeural 
bones 

Wide,  5-8  intemeural 
bones 

Wide.  7-9  intemeural 
bones 


Very  wide,  20-2 1  inter- 
neural  bones 


Close,  about  1  inter- 
neural  space 

Close,  about  I  'A  inter- 
neural  spaces 

Close,  about  1  inter- 
neural  space 


Close,  about  1  inter- 
neural  space 

Close,  about  1  inter- 
neural  space 

Close,  about  I  inter- 
neural  space 

Close,  about  1  inter- 
neural  space 

Close,  about  1  inter- 
neural  space 

Close,  about  1  inter- 
neural  space 


Wide,  2-2 'A  intemeural 
spaces 

Wide.  2-2',  intemeural 
spaces 

Wide,  about  2  inter- 
neural  spaces 

Wide,  about  2  inter- 
neural  spaces 

Wide,  about  3  inter- 
neural  spaces 

Close,  about  1  'A  inter- 
neural  spaces 

Wide,  about  3-3'/2  inter- 
neural  spaces 

Wide,  about  3  inter- 
neural  spaces 


Wide,  about  3  inter- 
neural  spaces 

Wide,  about  4  inter- 
neural  spaces 

Wide,  about  3-3'/,  inter- 
neural  spaces 


Wide,  about  2'/,  inter- 
neural  spaces 


Very  close,  from  0-2 
intemeural  bones 

Very  close,  from  0-2 
intemeural  bones 

Close,  from  2-3  inter- 
neural  bones 

Close,  usually  2  inter- 
neural  bones 

Moderately  wide,  4  or 
5  intemeural  bones 

Moderately  wide,  4  or 
5  intemeural  bones 

Moderately  wide,  about 
4  intemeural  bones 

Wide,  6  or  7  intemeu- 
ral bones 


Wide.  5-7  intemeural 
bones 

Wide,  5  or  6  intemeu- 
ral bones 

Wide,  4-6  intemeural 
bones 


2         Very  close,  1  intemeu- 
ral bone 


'  Juvenile  specimens  only  examined. 


296 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES- AH LSTROM  SYMPOSIUM 


B 


D 


DUNN  AND  MATARESE:  GADIDAE 


297 


Table  82.    Summary  of  Osteological  Characters  of  the  Caudal  Fin  in  Pretransformation  Larvae  of  Representatives  of  the  Genera 

OF  the  Family  Gadidae. 


Taxon 

Relalive  distance  between  postenor 
margin  of  anal  to  caudal 

Shape  neural  spine  on 
preural  centrum  one 

Shape  haemal  spine  on 
preural  centrum  one 

X.  Y  bones 

Lotinae 

Brosme  brosme 

Very  close,  about  '/,  intemeural 

Distally  flattened 

Rounded  distally 

Absent 

Molvu  dipterygia 

space 
Moderate,  about  2  intemeural 

Distally  flattened 

Rounded  distally 

Usually  absent 

Lola  tola' 

spaces 
Close,  about  1  intemeural 
space 

Distally  flattened 

Distal  -/j  slightly  flat- 
tened 

Usually  absent 

Phycinae 

Enchelyopus  cimhhus 

Close,  about  1  intemeural 

Distally  flattened 

Distally  flattened 

Present 

Gaidropsarus  sp. 

space 
Close,  about  1  intemeural 

Distally  flattened 

Distally  flattened 

Present 

Phycis  blennoides' 

space 
Close,  about  1  intemeural 

Distally  flattened 

Distally  flattened 

Present 

Urophyas  sp. 

space 
Close,  about  1  intemeural 

Distally  flattened 

Distally  flattened 

Present 

Ciliala  mustelta 

space 
Close,  about  1  intemeural 

Distally  flattened 

Distally  flattened 

Present 

Ramceps  raninus 

space 
Close,  about  1  intemeural 
space 

Distal  '/,  rounded 

Distal  ■/,  rounded 

Usually  pres- 
ent 

Gadinae 

Trisoplerus  luscus 

Wide,  about  2-2'/,  intemeural 

Distally  flattened 

Distally  flattened 

Absent 

Merlangius  merlangus 

spaces 
Moderate,  about  1  '/,-2  inter- 

Distally  flattened 

Distally  flattened 

Absent 

Pol  lac  hi  us  virens 

neural  spaces 
Wide,  about  2-2'/,  intemeural 

Distally  flattened 

Distally  flattened 

Absent 

Melanogrammus 

aeglefinus 

Gadus  macrocephalus 

spaces 
Moderate,  about  2  intemeural 

spaces 
Wide,  about  3  intemeural 

Distally  flattened 
Distally  flattened 

Distally  flattened 
Distally  flattened 

Absent 
Absent 

Gadiculus  argenleus 

spaces 
Close,  about  1 '/,  intemeural 

Distally  flattened 

Distally  flattened 

Absent 

Microgadus  proximus 

spaces 
Close,  about  1 '/,  intemeural 

Distally  flattened 

Distally  flattened 

Absent 

Eleginus  gracilis 

spaces 
Wide,  about  3  intemeural 

Distally  flattened 

Distally  flattened 

Absent 

Boreogadus  saida 

spaces 
Wide,  about  3  intemeural 

Distally  flattened 

Distally  flattened 

Absent 

Arctogadus  borisovi' 

spaces 
Wide,  about  3  intemeural 

Distally  flattened 

Distally  flattened 

Absent 

Theragra  chalcogramma 

spaces 
Wide,  about  3  intemeural 

Distally  flattened 

Distally  flattened 

Absent 

Micromesistius 

spaces 
Moderate,  about  2  intemeural 

Distally  flattened 

Distally  flattened 

Absent 

poutassou 

spaces 

'  Juvenile  specimens  only  examined. 


this  volume),  a  basal  gadid  (Mujib,  1 967),  a  medial  gadid  related 
to  gadines  (Svetovidov,  1948,  1969)  or  of  questionable  rela- 
tionship (Fahay  and  Markle,  this  volume).  We  present  here  our 
interpretation  of  the  relationships  of  subfamilies  of  gadid  fishes. 
Egg  diameter  is  largest  in  lotines  and  gadines,  smallest  in 
phycines  (Table  77).  A  single  oil  globule  is  present  in  lotines; 
multiple  oil  globules,  which  duiing  development  coalesce  into 


one,  are  found  in  most  phycines  (not  yet  reported  to  occur  in 
Ramceps).  and  are  absent  in  gadines.  Merlucctus  has  a  mod- 
erately sized  egg  (0.8-1.2  mm)  with  a  single  oil  globule  (Ahl- 
strom  and  Counts,  1955;  Russell,  1976;  Fahay,  1983).  Markle 
( 1982)  considered  small  ( <  1  mm)  eggs,  possessing  an  oil  globule, 
the  primitive  state.  We  agree,  but  also  consider  multiple  oil 
globules,  which  coalesce  into  one  the  most  primitive  state. 


Fig,  150.  (A)  Preflexion  larva  of  Microgadus  proximus.  3.6  mm  SL  (from  Matarese  et  al.,  1981);  (B)  Preflexion  larva  of  Eleginus  gracilis,  5.0 
mm  SL  (from  Dunn  and  Vinter,  1984);  (C)  Preflexion  larva  of  Boreogadus  saida.  6.3  mm  SL  (from  Dunn  and  Vinter,  1984);  (D)  Preflexion  larva 
of  Theragra  chalcogramma.  6.2  mm  SL  (from  Matarese  et  al.,  1981);  and  (E)  Flexion  larva  of  Micromesistius  poulassou.  8.0  mm  SL  (Zool.  Mus. 
Copenhagen,  stored  at  NWAFC). 


298 


ONTOGENY  AND  SYSTEM ATICS  OF  FISHES- AHLSTROM  SYMPOSIUM 


Hy4-6 


Fig.  151.  Caudal  fin  o(  Microgadus  proximus,  41.1  mm  SL.  Hyl  =  Hypural  bone  1;  Hy2-3  =  Hypural  bones  2  and  3;  Hy4-6  =  Hypural  bones 
4,  5,  and  6;  EP,  =  Epural  bone  1;  EP,  =  Epural  bone  2;  U,  =  Ural  centrum  1;  U,  =  Ural  centrum  2;  PU,  =  Preural  centrum  1;  PUk,  =  Preural 
centrum  10  (after  Matarese  et  al.,  1981). 


Lotinae  lai-vae  are  relatively  elongate  and  somewhat  narrow 
at  the  pectoral  fin  base;  the  former  state  is  partially  due  to  their 
numerous  vertebrae  (Table  78).  In  contrast,  phycines  are  shorter 
and  stockier  in  appearance,  deep  bodied  at  the  pectoral  fin  base, 
and  morphologically  somewhat  resemble  scorpaeniform  larvae. 
Ramceps  larvae  are  morphologically  the  most  divergent,  ap- 
pearing tadpole  shaped  due  to  their  depth  at  the  pectoral  fin 
base.  Gadines  are  somewhat  shorter  in  appearance,  and  deeper 
bodied,  than  lotines,  but  morphologically  intermediate  between 
phycines  and  lotines.  Merluccius  larvae  are  similar  to  gadines 
in  overall  shape  (Fig.  143D  in  Fahay  and  Markle,  this  volume). 

Length  at  hatching  is  smallest  in  most  phycines  and  somewhat 
larger  in  Raniceps.  lotines,  and  gadines  (Table  78).  Notochord 
flexion  occurs  at  quite  small  sizes  in  phycines  (except  in  Ran- 
iceps), relatively  larger  sizes  in  lotines,  and  intermediate  sizes 
in  gadines.  A  silvery  prejuvenile  stage  is  present  in  phycines 
(not  recorded  for  Raniceps),  but  a  pelagic  stage  of  varying  du- 
ration (Table  78)  is  probably  present  in  all  gadid  larvae  (Fahay 
and  Markle,  this  volume).  Merlucciiis  hatches  at  moderate  lengths 
(2.6-3.8  mm;  Ahlstrom  and  Counts,  1955;  Russell,  1976;  Fahay, 
1983),  notochord  flexion  begins  at  about  9  mm,  and  transfor- 
mation begins  at  20-25  mm,  somewhat  similar  to  gadines.  Elon- 
gate pelvic  fins  develop  precociously  in  all  lotines  (except  Lota) 
and  phycines,  but  not  in  gadines.  Pelvic  fins  in  Merluccius  are 
shorter  than  in  phycines.  but  longer  than  in  gadines.  Fahay  and 
Markle  (this  volume)  noted  similarities  in  fin  development  be- 
tween Merluccius  and  gadines  in  that  the  caudal  develops  first. 
In  gadines,  however,  the  pelvic  fins  develop  last.  In  Merluccius. 
it  is  the  second  fin  to  develop. 

Pigment  patterns  are  shared  by  Brosme  and  Molva.  but  not 
Lota,  whose  pigment  resembles  certain  gadines  (e.g.,  Pollachius 
pollachius,  Trisopterus  minutus).  Two  kinds  of  pigment  patterns 
have  been  identified  for  phycines:  either  dorsal  saddles  of  pig- 
ment in  the  postanal  region  (Enchelyopus  cimbrius,  Gaidrop- 


sarus  mediterraneus.  Urophycis  chuss)  or  a  ventral  series  of  me- 
lanophores  (Phycis  blennoides,  Ciliata.  and  Raniceps).  Gadines 
have  either  one  or  two  postanal  bars  or  dorsal  and  ventral  lines 
of  pigment.  In  gadines  the  pelvic  fins  lack  pigment  such  as  that 
present  in  lotines  and  phycines.  In  comparison,  Merluccius  has, 
in  certain  species  (e.g.,  M.  product  us.  Ahlstrom  and  Counts, 
1955),  a  single  postanal  band  of  pigment,  but  two  in  M.  albidus 
and  M.  hilinearis  (Fahay,  1983)  and  three  lateral  melanophores 
(M.  merluccius.  Russell,  1976).  Pelvic  fins  are  pigmented  in 
some  Merluccius. 

Lotines  have  one  (Brosme)  or  two  (Moha.  Lota)  dorsal  fins; 
when  two  fins  are  present,  they  are  internally  continuous  (Mar- 
kle, 1982).  Phycines  have  two  dorsal  fins,  the  first  specialized 
(Cohen  and  Russo,  1979),  and  are  internally  continuous  (Mar- 
kle. 1982).  Gadines  have  three  dorsal  fins,  of  which  the  second 
and  third  are  always  internally  continuous.  The  posterior  mar- 
gins of  the  dorsal  and  the  anal  fins  are  close  to  the  procurrent 
rays  of  the  caudal  fin  in  lotines  and  phycines,  whereas  in  gadines 
these  fins  are  generally  some  distance  from  the  caudal  fin,  as  is 
the  case  in  Merluccius  (Fahay  and  Markle,  this  volume). 

Certain  trends  in  osteological  structures  can  be  noted  in  the 
family  Gadidae.  Transient  pterotic  spines  are  present  in  some 
phycines  (some  f/!i'ai.  Gaidropsarus.  Markle,  1 982)  and  Ciliata 
(this  study),  and  are  lacking  (so  far  as  is  known)  in  other  phy- 
cines, lotines,  gadines,  and  in  Merluccius.  The  distribution  of 
branchiostegal  rays  varies  among  gadid  genera.  The  seven  bran- 
chiostegal  rays  in  Brosme  are  carried  on  the  outer  surface  of  the 
ceratohyal,  whereas  in  Gadus  and  Lota  (Mujib,  1967,  1969).  as 
well  as  in  Theragra  (Dunn  and  Vinter,  MS),  the  three  anterior 
rays  are  internal  and  the  posterior  four  are  external.  Raniceps 
has  one  branchiostegal  ray  on  the  epihyal,  a  character  considered 
primitive  and  shared  (so  far  as  known)  with  Urophycis  chuss 
and  Merluccius  (Mujib,  1967;  Inada,  1981b),  whereas  gadines 
have  all  seven  branchiostegal  rays  on  the  ceratohyal  (this  study). 


DUNN  AND  MATARESE:  GADIDAE 


299 


Phycinae 


Gadinae 


Raniceps 


/ 


Fig.  152.     Proposed  relationships  of  gadid  subfamilies. 


The  ventral  branch  of  the  posttemporal  is  shortest  in  lotines 
and  phycines  (moderately  long  in  Merluccius)  and  longest  in 
gadines.  The  gadids  we  examined  all  have  four  pectoral  radials, 
except  for  a  single  specimen  of  Ciliata  with  five  radials  on  one 
side.  Phycines  lack  an  expanded  distal  head  on  the  postcleithra. 
Raniceps.  however,  has  two  postcleithra;  the  upper  is  oblong  in 
shape,  the  lower  is  distally  pointed  as  in  phycines.  Among  the 
lotines,  the  distal  end  of  the  postcleithrum  is  slightly  expanded 
in  Brosme  and  Molva  whereas  the  postcleithrum  is  distally 
pointed  in  Lota.  In  Merluccius  it  is  moderately  expanded  while 
in  gadines,  the  postcleithrum  is  considerably  expanded  at  its 
tip,  which  we  infer  is  a  derived  condition.  Predorsal  bones  are 
present  in  some,  but  not  all,  phycines  (Urophycis.  Phycis.  and 
Raniceps).  but  are  lacking  in  lotines,  gadines,  and  Merluccius: 
the  loss  is  considered  an  advanced  state.  The  posterior  process 
of  the  basipterygia  is  quite  long  in  some  phycines  (and  in  Mer- 
luccius), moderately  long  in  lotines,  and  shortest  in  gadines,  and 
the  latter  state  is  considered  derived. 

The  shape  of  the  neural  and  haemal  spine  on  PU,  varies 
among  genera.  The  neural  spine  in  Raniceps  is  distally  rounded 
(a  primitive  condition),  but  this  spine  is  flattened  in  all  other 
gadids,  as  it  is  in  Merluccius.  Raniceps.  Brosme,  and  Molva 
have  a  rounded  haemal  spine  on  PU,,  in  contrast  to  the  flattened 
tip  on  all  other  gadids  (and  Merluccius):  x/y  bones  are  present 
in  all  phycines  (usually  present  in  Raniceps)  and  Merluccius, 
but  are  absent  {Brosme)  or  usually  absent  {Molva,  Lota)  in 
lotines  (Markle,  1982;  this  study)  and  are  absent  in  gadines.  All 
gadids  and  Merluccius  (Ahlstrom  and  Counts,  1955;  Inada. 
1981b)  have  three  hypural  bones  (including  the  parhypural); 
Raniceps  alone,  among  the  gadids  examined  by  us,  showed  evi- 
dence of  ontogenetic  reduction  by  fusion  from  six  hypural  bones 
to  three.  As  noted  by  Markle  ( 1 982)  and  Fahay  and  Markle  (this 
volume),  lotines  and  gadines  have  four  or  five  primary  caudal 
fin  rays,  while  phycines  have  five  or  six  such  rays.  Merluccius 
and  Raniceps  each  have  six  primary  rays  (Inada,  1981b;  this 
study). 


We  consider  Raniceps  a  basal  gadid  considering  the  following 
characters:  eggs  small  with  a  single  oil  globule;  larvae  tadpole- 
or  liparid-shaped;  one  branchiostegal  ray  on  the  epihyal;  two 
postcleithra  present;  a  predorsal  bone  present;  the  neural  and 
haemal  spines  on  PU,  distally  rounded;  six  hypural  bones  which 
fuse  into  three  during  ontogeny;  x/y  bones  usually  present;  and 
six  primary  rays  on  the  superior  hypural  bone. 

We  further  consider  phycines  to  be  a  more  primitive  group 
than  lotines  based  on  the  following  characters:  eggs  small,  with 
multiple  oil  globules  which  coalesce  into  one  during  develop- 
ment; larvae  stocky  and  deep  bodied  (at  the  pectoral  fin  base); 
elongate  and  precocious  pelvic  fins  present;  postcleithrum  with- 
out an  expanded  head;  one  or  more  predorsal  bones  present; 
elongate  pelvic  process;  and  x/y  bones  present.  Until  the  pres- 
ence or  absence  of  transient  pterotic  spines  is  established  in  all 
phycine  larvae,  the  most  parsimonious  explanation  is  that  their 
presence  represents  a  derived  character  state. 

Lotines,  as  presently  constituted,  appear  to  us  to  possess  a 
number  of  primitive  and  intermediate  characters,  as  well  as 
some  rather  specialized  traits:  eggs  moderately  large  with  a  single 
oil  globule;  larvae  elongate,  relatively  shallow  at  the  pectoral  fin 
base;  pelvic  fins  precocious,  elongate  and  with  the  posterior 
process  of  the  basipterygium  moderately  long;  postcleithnam 
with  slightly  expanded  head;  predorsal  bones  absent;  x/y  bones 
usually  absent;  and  three  hypural  bones  present.  Brosme  has 
both  apparently  primitive  (e.g.,  all  branchiostegal  rays  carried 
on  the  outside  surface  of  the  ceratohyal  and  a  rounded  haemal 
spine  on  PU,)  and  derived  characters  (e.g.,  x/y  bones  always 
lacking);  its  single  dorsal  fin  was  considered  primitive  by  Sve- 
tovidov  (1948)  or  derived  (within  Lotinae,  sensu  Svetovidov, 
1948)  by  Mujib  (1969).  As  noted  by  Markle  (1982),  high  total 
dorsal  and  anal  fin  ray  counts  may  be  primitive  for  the  order 
Gadiformes. 

Gadines  seem  to  us  a  relatively  homogenous  group,  charac- 
terized by  reductive  (or  lost)  and  apparently  derived  characters. 
The  former  include:  eggs  without  an  oil  globule;  posterior  pro- 
cess of  the  pelvic  bone  reduced  in  length  or  wanting;  predorsal 
and  x/y  bones  absent;  and  three  hypural  bones  present.  The 
latter  characters  include:  eggs  moderate  in  size;  larvae  morpho- 
logically uniform  in  appearance;  lower  branch  of  posttemporal 
relatively  long;  postcleithrum  with  expanded  head;  and  three 
dorsal  and  two  anal  fins  present,  with  the  anal  fins  and  dorsal 
fins  two  and  three  internally  continuous. 

Our  hypothesis  of  relationships  of  gadid  subfamilies  is  pre- 
sented in  Figure  152.  The  relationships  of  a  number  of  genera, 
such  as  Brosme  and  Raniceps.  and  the  relationships  of  Phycis, 
Gaidropsarus,  and  Ciliata  to  other  phycines  still  remain  con- 
fused. Based  on  early  life  history  characteristics  and  osteology, 
we  consider  Merluccius  a  gadid  related  to,  but  more  primitive 
than.  Gadinae  and,  following  Svetovidov  (1948,  1969),  restrict 
Merlucciinae  to  this  genus.  The  relationship  of  all  nominal  gadid 
subfamilies  requires  further  study. 

Northwest  and  Alaska  Fisheries  Center,  2725  Montlake 
Boulevard  East,  Seattle,  Washington  981 15. 


Bregmacerotidae:  Development  and  Relationships 
E.  D.  HouDE 


THE  codlets  are  small,  gadiform  fishes  of  pelagic  habit  found 
in  neritic  and  oceanic  water  of  tropical  and  subtropical 
seas.  The  family  Bregmacerotidae  (Gill,  1872)  includes  the  single 
genus  Bregmaceros  (Thompson,  1840),  in  which  there  are  sev- 
eral species.  In  recent  reviews  six  (Belyanina,  1974)  or  seven 
(D'Ancona  and  Cavinato,  1965)  valid  species  have  been  rec- 
ognized. The  systematics  remain  confused,  although  Belyanina 
(1974)  has  partly  clarified  species  relationships.  Larvae  often 
are  among  the  ten  most  common  families  occurring  in  both 
oceanic  and  coastal  ichthyoplankton  surveys  in  subtropical  and 
tropical  waters  (e.g.,  Ahlstrom,  1971;  Moser  et  al.,  1973;  Houde 
etal.,  1979;  Loeb,  1979;  Richards,  1981).  The  species  are  mor- 
phologically similar  but  most  have  distinctive  meristics,  from 
which  specific  identifications  usually  are  possible.  Differences 
in  vertebral  number  and  median  fin  ray  counts  serve  to  distin- 
guish larval  to  adult  stages  while  pigmentation  differences  and 
the  size  at  appearance  of  the  single,  first  dorsal  fin  ray  serve  to 
identify  small  larvae.  Larval  characters,  particularly  those  of  the 
smallest  individuals  (1.5-3.0  mm  SL),  often  are  the  best  char- 
acters for  identification  purposes.  A  careful  examination  of  on- 
togenetic evidence  indicates  that  some  species  are  still  unde- 
scribed  and  that  misidentified  Bregmaceros  frequently  have  been 
reported  in  the  literature.  Based  on  evidence  from  larval  char- 
acteristics there  may  be  ten  or  more  valid  species  in  the  world 
oceans. 

Species  distributions.  —  Larvae  of  Bregmaceros  commonly  occur 
between  latitudes  40°S  and  40°N  (Table  83).  D'Ancona  and 
Cavinato  (1965)  and,  more  recently,  Belyanina  (1974),  have 
reviewed  distribution  data  on  the  known  species.  Centers  of 
abundance  have  been  observed  in  the  western  Indo-Pacific  and 
Indian  Oceans  (Munro,  1950;  D'Ancona  and  Cavinato,  1965; 
Kotthaus,  1969;  Belyanina,  1974),  in  the  eastern  Pacific  (Ahl- 
strom. 1971;  Belyanina,  1974)  and  in  the  Caribbean  Sea  and 
Gulf  of  Mexico  (Belyanina  and  Lopes,  1974;  Milliken,  1975; 
Belyanina,  1980;  Houde,  1981).  Bregmaceros  macclellandi  is 
circumtropical  with  areas  of  apparent  high  abundances  in  the 
Caribbean  Sea,  western  Indian  Ocean  and  Indo-Malayan  region. 
It  also  occurs  in  the  eastern  Pacific.  Bregmaceros  atlanticus. 
including  the  closely  related  Pacific  Ocean  form  B.  japonicus 
(D'Ancona  and  Cavinato,  1965)  also  is  circumtropical  with  an 
apparent  center  of  abundance  in  the  western  Atlantic. '  The  latter 
sometimes  occurs  in  neritic  waters.  Several  neritic  species  are 
known,  including  B.  nectahanus,  B.  arahicus,  B.  rarisquamosus, 
B.  bathymaster,  B.  caw/or;  (Milliken  and  Houde,  1984)  and  the 
Type  A  larva  described  by  Houde  (1981). 


Neritic  species  vary  in  the  breadth  of  their  distributions.  It 
now  seems  certain  that  the  Indo-Pacific  B.  nectabanus  does  not 
occur  in  the  western  Atlantic  and  its  occurrence  in  the  eastern 
Atlantic  Ocean  is  uncertain.  The  species  B.  cantori.  described 
by  Milliken  and  Houde  (1984),  is  the  most  common  bregma- 
cerotid  in  the  western  Atlantic.  It  occurs  in  the  Caribbean  Sea 
and  Gulf  of  Mexico  (Milliken,  1975;  Houde,  1981),  in  the  south- 
west Atlantic  Ocean  off  Brazil-  and  along  the  East  Coast  of  the 
United  States.'  The  common  bregmacerotid  in  the  Gulf  of  Ca- 
riaco,  initially  referred  to  as  B.  atlanticus  (Mead,  1963)  and 
subsequently  as  B.  nectabanus  (Baird  et  al.,  1973,  1974;  Bely- 
anina and  Lopes,  1974)  and  that  referred  to  as  B.  nectabanus 
from  the  Caribbean  Sea  and  Gulf  of  Mexico  (Belyanina,  1980) 
is  B.  canton  (M\\\\ken,  1975;  Houde,  1981;  Milliken  and  Houde, 
1984).  Bregmaceros  bathymaster  has  been  collected  only  in  the 
eastern  Pacific.  It  is  abundant  in  the  Gulf  of  Panama  (D'Ancona 
and  Cavinato,  1965)  and  in  the  Gulf  of  California  (Moser  et  al., 
1973).  Bregmaceros  rarisquamosus  occurs  in  the  Indian  Ocean, 
Bay  of  Bengal,  Arabian  Sea  and  western  Pacific  Ocean.  It  also 
is  present  in  the  Persian  Gulf''  where  it  occurs  with  B.  necta- 
banus and  B.  arabicus.  Previously,  B.  arabicus  had  been  re- 
ported from  the  Arabian  Sea,  Bay  of  Bengal  and  East  China  Sea. 
Larvae  of  an  undescribed  species,  B.  Type  A,  have  been  collected 
in  the  western  North  Atlantic  (Houde,  1981)'  ^ 

Bregmacerotids  reportedly  occur  from  the  surface  to  depths 
of  approximately  4,000  m,  but  are  most  common  in  the  upper 
300  m.  Larvae  generally  occur  from  surface  to  600  m  depth, 
neritic  species  tending  to  be  closer  to  the  surface  than  oceanic 
species  (D'Ancona  and  Cavinato,  1965).  Some  reported  catches 
from  great  depths  may  be  in  error.  Adults  and  subadults  of  some 
Bregmaceros  undertake  extensive  vertical  migrations  and  one 
species  (B.  cantori)  inhabits  anoxic  water  during  a  part  of  the 
day  (Mead,  1963;  Wilson.  1972;  Baird  et  al.,  1973;  Milliken, 
1975). 

Family  characteristics.  —  CharaclcTs  defining  Bregmacerotidae 
were  summarized  briefly  by  Nelson  ( 1 976)  and  more  extensively 
by  D'Ancona  and  Cavinato  (1965)  and  by  Belyanina  (1974). 
Fahay  and  Markle  (this  volume)  have  tabulated  meristic  data 
and  discussed  ontogenetic  characters  of  Gadiformes,  including 


'  Late  larvae  and  juveniles  that  I  examined  from  the  eastern  Pacific 
appeared  to  be  typical  B.  atlanticus  but  small  larvae,  which  may  have 
been  younger  specimens  of  this  species,  did  not  resemble  typical  B. 
atlanticus  from  the  Atlantic.  The  eastern  Pacific  specimens  were  less 
pigmented,  with  a  prominent  melanophore  on  the  ventral  midline,  be- 
tween the  anus  and  the  lip  of  the  tail.  Specimens  were  provided  by  Dr. 
H.  G.  Moser,  Southwest  Fisheries  Center,  National  Marine  Fisheries 
Service,  La  Jolla,  California. 


-  I  examined  specimens  of  B.  cantori  from  coastal  waters  of  Brazil, 
collected  from  latitudes  of  22°S  to  27°S,  provided  to  me  by  Dr.  Y. 
Matsuura,  Instituto  Oceanografico,  Universidade  de  Sao  Paulo,  Brazil. 

'  1  examined  specimens  from  R/V  DOLPHIN  cruises,  taken  from 
Florida  to  the  Carolinas,  provided  to  me  by  M.  P.  Fahay,  Sandy  Hook 
Laboratory,  National  Marine  Fisheries  Service,  Highlands,  New  Jersey. 

■■Houde.  E.  D.,  J.  C.  Leak,  S.  Al-Matar,  and  C.  E.  Dowd.  1981. 
Ichthyoplankton  abundance  and  diversity  in  the  weslem  Arabian  Gulf. 
Kuwait  Institute  for  Scientific  Research,  Mariculture  and  Fisheries  De- 
partment, Final  Report,  Project  MB- 16,  3  volumes.  (This  report  was 
not  available  for  distribution  at  the  time  the  present  paper  was  written.) 

'  The  Type  A  larva  was  present  in  collections  from  two  R/V  AL- 
BATROSS cruises  into  the  Caribbean  Sea.  1  examined  larvae  provided 
by  Dr.  W.  J.  Richards,  Southeast  Fisheries  Center,  National  Marine 
Fisheries  Services,  Miami,  Florida. 


300 


HOUDE:  BREGMACEROTIDAE 


301 


Bregmacerotidae.  Bregmacerotids  are  small  fishes,  the  largest 
species,  B.  macclellandi,  rarely  exceeding  120  mm  SL.  They 
have  two  dorsal  fins,  the  first  a  single,  elongate  ray  on  the  occiput. 
The  second  dorsal  fin  and  the  anal  fin  are  long  with  median  rays 
much  reduced,  giving  the  fins  a  divided  appearance.  The  caudal 
fin  is  separated  from  the  dorsal  and  anal  fins.  Pelvic  fins  are 
j  ugular  and  consist  of  5  (usually  )-7  rays,  the  outer  three  elongate. 
The  olfactory  nerves  pass  through  a  broad  canal,  wider  than 
that  in  Gadidae.  The  sacculus  is  very  large.  The  swimbladder 
does  not  contact  the  auditory  capsules.  There  are  a  few  pyloric 
caeca.  The  vomer  is  toothed.  A  lateral  line  is  present  under 
the  second  dorsal  fin.  Chin  barbels  are  absent. 

Development 

Spawning.  — Size  at  maturity  is  variable  but  generally  <30  mm. 
In  one  species,  B.  rarisquamosus.  maturity  is  attained  at  <  1 5 
mm  (D'Ancona  and  Cavinato,  1965).  Larvae  occur  in  the  tropics 
and  subtropics  dunng  all  months,  indicating  protracted  spawn- 
ing, although  seasonality  is  apparent  for  individual  species  in 
some  areas. 

£■^^5.  — Eggs  are  presumed  to  be  pelagic.  Excepting  a  single  re- 
port, the  fertilized  eggs  and  embryos  of  Bregmaceros  species 
have  not  been  described.  Pertseva-Ostroumova  and  Rass  (1973) 
described  fertilized  eggs,  attributable  to  B.  atlaniicus.  as  pelagic 
with  smooth  chorion,  small  perivitelline  space  and  homoge- 
neous yolk  containing  an  oil  globule.  They  reported  the  egg 
diameter  to  be  1.1  mm  and  the  oil  globule  diameter  to  be  0.20 
mm.  In  my  opinion,  it  is  unlikely  that  Bregmaceros  eggs  are 
that  large  because  newly-hatched  larvae  are  only  1.5  mm  long. 
Ahlstrom's''  unpublished  notes  give  diameters  of  Z?.  bathyniaster 
eggs  as  0.84-1.00  mm  and  indicate  that  a  single  oil  globule  is 
present. 

Ten  eggs  with  well-developed  embryos  that  I  examined,  iden- 
tified as  B.  bathymaster  by  E.  H.  Ahlstrom.  collected  in  the 
mouth  of  the  Gulf  of  California^  ranged  from  0.88-1 .00  mm  in 
diameter  (.v  =  0.94  mm)  and  had  a  single  oil  globule  0.22-0.28 
mm  in  diameter  (.v  =  0.24  mm).  The  chorion  was  smooth,  per- 
ivitelline space  narrow  and  yolk  homogeneous.  The  oil  globule 
was  situated  in  the  posterior  part  of  the  yolk  mass.  Several  small 
melanophores  were  scattered  on  the  head  and  dorsal  side  of  the 
embryo  and  on  the  ventral  side  of  the  tail. 

Larvae  —  The  larvae  are  not  unusual.  Their  general  morphology 
is  similar  to  that  of  other  gadiform  larvae  but  bregmacerotids 
are  not  likely  to  be  confused  with  them  or  with  larvae  of  other 
tropical-subtropical  fishes  with  which  they  occur.  In  bregma- 
cerotids, metamorphosis  is  gradual  and  direct. 

Newly-hatched  larvae  are  small,  approximately  1.5  mm  NL, 
a  fact  often  not  appreciated  when  collecting  nets  with  >333- 
^m  meshes  have  been  used.  The  smallest  larvae  usually  have 


Table  83.    Geographic  Distribution  Information  and  Some  Meris- 

Tic  Data  of  Bregmaceros  Adults  and  Larvae  >  8  mm  SL.  Numbers 

in  parentheses  are  the  most  common  counts  for  a  species.  For  additional 

meristic  data,  see  Fahay  and  Markle,  this  volume. 


'  Ahlstrom,  E.  H.  Personal  Notes.  "Gadiformes."  Notes  on  file  at 
National  Marine  Fisheries  Service,  Southwest  Fisheries  Center,  La  Jolla, 
California,  USA. 

'  The  eggs,  identified  as  B.  bathymasler.  were  provided  by  Dr.  H.  G. 
Moser,  Southwest  Fisheries  Center,  National  Mannc  Fisheries  Service. 
La  Jolla,  California.  They  were  collected  on  10  June  1957,  Station 
I45G.40,  Cruise  5706-S,  near  the  mouth  of  the  Gulf  of  California.  I 
could  not  confirm  that  the  eggs  were  those  of  Bregmaceros.  although 
embryo  myomere  numbers  were  in  the  reported  range  for  B.  bathy- 
master. 


Myomeres 

Dorsal 

Anal 

Species 

Distnbution' 

(vertebrae) 

fin  rays 

fin  rays 

B.  macclellandi 

CTO 

52-59 

57-65 

58-69 

(54-55) 

(58-61) 

(62-66) 

B.  atlanticus 

CTO(N) 

50-55 

47-56 

49-58 

(52-53) 

(50-54) 

(52-55) 

(B.  japonicusl) 

WPO 

56-58 

51-60 

56-63 

B.  nectabanus 

I,  IP,  WPN 

47-52 

42-55 
(47-50) 

43-55 
(50-52) 

B.  cantori 

WAN 

45-48 

45-48 

45-49 

B.  rarisquamosus 

I.  IP.  WPN 

43-48 
(43-46) 

34-41 

36-43 

B  bathymaster 

EPN 

48-51 

44-50 

45-52 

B  arabicus 

I.  IP,  WPN 

50-54 

50-60 
(52-54) 

50-63 
(56-57) 

B.  Type  A 

WAN 

44-47 

40-44 

42-46 

'  CT  =  circumlropical;  O  =  oceanic;  N  =  nentic;  WP  =  weslem  Pacific;  I  =  Indian;  IP  =  Indo- 
Pacific;  WA  =  weslem  Atlanlic;  EP  =  eastern  Pacific. 


not  been  described,  although  it  is  dunng  that  stage  when  specific 
pigmentation  is  unique  and  identification  easiest.  Small  speci- 
mens (1.5-3.1  mm  NL)  of  eight  species  are  illustrated  (Figs.  153 
and  154).  Larvae  of  3.0-6.0  mm  SL  may  be  most  difficult  to 
identify  because  pigment  patterns  are  in  transition  and  fin  rays 
have  not  developed  sufliciently  to  be  diagnostic.  At  lengths  >  6.0 
mm  identification  becomes  easier,  based  on  pigmentation  char- 
acteristics (Figs.  1 55  and  I  56)  and  on  complete  (or  nearly  com- 
plete) counts  of  median  fin  rays  and  myomeres.  For  larvae  >10- 
1 1  mm,  diagnostic  meristics  usually  are  complete  and  illustra- 
tions/descriptions in  D'Ancona  and  Cavinato  (1965)  and  Be- 
lyanina  (1974)  usually  will  lead  to  correct  identifications.  Use 
of  information  on  larval  pigmentation,  meristics  and  size  at 
occipital  ray  development  allow  all  described  species  to  be  iden- 
tified. 

Occipital  ray  (Table  84).— The  size  at  appearance  of  the  single 
occipital  ray  varies  among  species.  It  is  the  first  fin  ray  to  develop 
in  B.  macclellandi  and  in  B.  Type  A  (Houde,  1981),  appearing 
when  larvae  are  2.0-2.5  mm  in  length.  In  all  other  species  the 
ray  develops  at  lengths  of  5.0-7.5  mm,  usually  at  approximately 
6.5  mm.  The  occipital  ray  ofB.  macclellandi  is  long  and  delicate, 
often  extending  to  near  the  middle  of  the  second  dorsal  fin  in 
specimens  <  1 0  mm,  but  subsequently  declining  in  relative  length. 
In  other  species,  the  occipital  ray  is  shorter,  never  reaching  the 
second  dorsal  fin. 

Pigiyientation  (Table  84) —'Larvae  of  the  oceanic  species  B. 
macclellandi  and  B.  atlanticus  are  darkly  pigmented.  Larvae  of 
the  neritic  species  are  lightly  pigmented.  All  have  heavy  internal 
pigment  over  the  visceral  mass.  The  most  distinctive  pigment 
is  present  on  the  smallest  larvae  (Figs.  153  and  154)  and  all 
described  species  can  be  identified  using  pigmentation  patterns 
for  larvae  of  1.5-3.0  mm  SL.  The  amounts  of  pigment,  and 
particularly  the  diagnostic  melanophore  patterns,  tend  to  be  lost 
or  reduced  as  larvae  grow.  External  pigment  tends  to  migrate 
internally  with  growth,  the  tendency  being  most  apparent  in  the 
neritic  Indo-Pacific  species  B.  nectabanus.  B.  arabicus  and  B. 
rarisquamosus.  At  the  smallest  lengths,  the  closely  related  B. 
nectabanus  and  B.  cantori  have  obviously  different  pigmenta- 


Fm   1 53      Urvae  of  Bregmaceros  in  the  length  range  2. 1  to  3. 1  mm  NL.  (A)  B  macMland,.  3.0  mm.  2701 5'N,  084»28'W;  (B)  B  allanUcus. 
2.9  mm   27W  084»2Tw  (C)  B.  Type  A.  3.1  mm.  26°00'N.  083«53'W;  (D)  B.  balhymasWr.  2.1  mm,  22°55'N,  108-40'W. 


HOUDE:  BREGMACEROTIDAE 


303 


Fig.  154.     Larvae  of  Bregmaceros  in  the  length  range  2.1  to  3.1  mm  NL.  (A)  B.  nectabanus.  2.8  mm.  26°18'N.  052°00'E;  (B)  B.  canlori,  2.6 
mm;  27°00'N.  084°2rW;  (C)  B.  arabicus.  2.5  mm,  29°26'N,  048°00'E;  (D)  B.  rarisquamosus.  2.5  mm,  25°52'N,  OSS'SS'E. 


B 


Fig   1 55      Larvae  of  Bregmaceros  in  the  length  range  7.0  to  10.0  mm  SL.  (A)  B  macclellandi.  7.0  mm,  1  3°00'N,  060°00W,  (B)  B.  attanucus. 
9.0  mm,  24''34'N,  082°56'W;  (C)  B.  Type  A,  8.5  mm,  27"'00'N,  084°22'W;  (D)  B.  balhymaster.  9.5  mm.  13°I2'N.  09r5rw. 


HOUDE:  BREGMACEROTIDAE 


305 


Table  84.    Size  at  Appearance  of  Occipital  Ray  and  Pigmentation  Characteristics  of  Bregmaceros  Larvae  in  Two  Length  Ranges. 
See  Figures  153-156.  In  addition  to  pigment  described,  all  Bregmaceros  larvae  have  internal  pigment  on  dorsal  surface  of  visceral  mass. 


Species 


Size  at 

appearance  of 

occipital  ray 

(mm  SL) 


^  .1  mm  SL 


Distinctive  pigmentatii 


5-10  mm  SL 


B.  macclctlandi 


B.  atlanticus 


B.  nectabanus 


B.  canlori 


B.  ransquamosus 


B  bathyniastcr 


B.  arabicus 


B.  Type  A 


2.0-2.5         Melanophore  at  angle  of  jaw  and  tip  of  lower  jaw; 
scattered  melanophores  on  head  and  at  base  of 
pectoral  fin.  A  few  large,  internal  stellate  mela- 
nophores in  double  row  on  side  of  body  and 
tail.  Melanophores  on  ventral  surface  of  viscer- 
al mass. 
-5.0-5.5         Melanophore  at  angle  of  jaw  and  tip  of  lower  jaw. 
Scattered  melanophores,  on  head  and  over 
midbrain  and  at  base  of  pectoral  fin.  Scattered, 
large  internal  stellate  melanophores  on  side  of 
body.  Diffuse  melanophores,  some  dendritic, 
on  surface  of  trunk  and  tail. 

6.0-7.0         Single  melanophores  at  angle  of  jaw,  over  hind- 
brain,  on  nape  and  just  anterior  to  anus.  Dif- 
fuse melanophores  in  short,  double  row  on  side 
of  tail  and  also  in  dorsal  and  ventral  finfolds 
directly  above  and  below  the  double  row.  Me- 
lanophore on  ventral  side  of  tail,  just  anterior 
to  notochord  tip. 

6.0-7.0         Melanophore  at  angle  of  jaw.  A  few  small  mela- 
nophores on  ventral  surface  of  visceral  mass. 


6.0-7.0         Melanophore  at  angle  of  jaw  and  on  lower  jaw 

tip;  also  over  hindbrain.  Few  scattered  melano- 
phores on  ventral  surface  of  visceral  mass.  Dif- 
fuse melanophores  in  three  patches  on  side  of 
tail.  Melanophore  on  dorsal  surface  of  tail  just 
antenor  to  notochord  lip. 

6.5-7.5  Melanophore  at  angle  of  jaw  and  at  anus.  A  few 
scattered  melanophores  on  ventral  surface  of 
visceral  mass.  A  row  of  5-7  melanophores  on 
the  ventral  side  of  the  tail. 


6.0-7.0         Melanophore  at  tip  of  lower  jaw  and  on  ramus  of 
lower  jaw  (elongate  melanophore).  Often  a  few 
scattered  melanophores  on  ventral  surface  of 
visceral  mass.  Diffuse  pigment  in  three  patches 
on  side  of  tail.  Melanophore  on  side  of  tail  just 
anterior  to  notochord  tip.  Single  melanophore 
over  forebrain. 

2.0-2.5         Melanophore  at  angle  of  jaw  and  on  tip  of  lower 
jaw.  Scattered  melanophores  over  hindbrain. 
Melanophores  on  occipital  ray. 


Small,  scattered  melanophores  over  surface  of  head 
and  body  but  not  on  posterior  part  of  tail.  Sev- 
eral, large,  internal  stellate  melanophores  in  a 
double  row  on  side  of  body  and  tail. 


Many  melanophores  over  surface  of  head  and 
body,  including  dorsal,  anal  and  caudal  fins.  Lar- 
va more  or  less  "completely"  pigmented. 


Melanophore  at  angle  of  jaw.  A  few  melanophores 
on  tail  just  anterior  to  its  lip  and  sometimes  one 
or  two  melanophores  at  base  of  caudal  fin.  Inter- 
nal melanophores  on  side  of  body,  between 
origins  of  dorsal  and  anal  fins  and  also  in  tail 
midway  between  origins  of  those  fins  and  lip  of 
tail. 

Melanophore  at  angle  of  jaw.  A  large  melanophore 
often  present  over  forebrain.  Internal  pigment 
visible  near  otoliths  and  just  antenor  to  insertion 
of  pectoral  fins.  Internal  pigment  sometimes  visi- 
ble along  developing  vertebral  column. 

Melanophore  at  angle  of  jaw.  Large,  intense  group 
of  melanophores  in  caudal  fin.  Scattered  melano- 
phores on  anterior,  ventral  surface  of  visceral 
mass.  Some  internal  melanophores  along  devel- 
oping vertebral  column  in  tail,  just  antenor  to  its 
lip. 

Melanophore  at  angle  of  jaw  and  on  ramus  of  lower 
jaw.  Melanophores  on  snout  and  on  surface  over 
fore-  and  midbrain.  Melanophore  at  anus  and 
two  or  more  melanophores  on  dorsum  just  under 
base  of  anterior  third  of  second  dorsal  fin.  In- 
tense group  of  melanophores  in  caudal  fin.  Inter- 
nal pigment  along  developing  vertebral  column 
in  peduncle  region. 

Melanophore  at  tip  of  lower  jaw  and  an  elongate 
melanophore  on  ramus  of  lower  jaw.  Several  me- 
lanophores in  base  of  caudal  fin  and  a  few  scat- 
tered melanophores  on  tail  just  anterior  to  caudal 
fin. 


Melanophore  at  angle  of  jaw.  Scattered  melano- 
phores over  fore-,  mid-  and  hindbrain  and  on 
nape.  Melanophores  on  occipital  ray  and  in  pel- 
vic fins.  Scattered  melanophores  on  ventral  sur- 
face of  anterior  half  of  visceral  mass  sometimes 
present. 


tion.  The  smallest  B.  macclellandi  and  B.  atlanticus  larvae  po- 
tentially could  be  confused,  based  on  pigmentation  alone.  Breg- 
maceros macclellandi  has  less  external  pigment  and  internal 
pigment  on  tail  and  body  is  more  clearly  organized  into  two 
rows  than  that  of  B.  atlanticus. 

As  larvae  grow  pigmentation  becomes  less  reliable  as  a  means 
to  identify  them.  Nevertheless,  the  patterns  are  distinctive  enough 
to  allow  tentative  identification  (Figs.  1  55  and  1 56),  which  can 
be  confirmed  by  considering  meristic  characters.  Because  of 
identification  errors,  there  are  erroneous  descriptions  of  pig- 


mentation in  the  literature.  For  example,  larvae  of  fi.  nectabanus 
<  10  mm  do  have  a  melanophore  at  the  jaw  angle,  although  the 
review  literature  indicates  that  it  is  absent  (D'Ancona  and  Cav- 
inato,  1965;  Belyanina,  1974). 

Meristic  characters  (Table  83).  — Excepl  for  B.  arabicus.  neritic 
species  have  lower  myomere,  vertebrae  and  median  fin  ray  counts 
than  do  B.  macclellandi  or  B.  atlanticus.  The  lowest  meristics 
occur  in  B.  rarisquamosus  and  the  highest  in  B.  macclellandi. 
The  neritic  B.  arabicus  has  meristics  similar  to  those  of  B. 


Fig.  156.     Larvae  of  Bregmaceros  in  the  length  range  7.0  to  10.0  mm  SL.  (A)  B.  nectabanus.  10.0  mm.  25''28'N,  053°50'E;  (B)  B.  canton.  8.0 
mm,  27°15'N,  083°53'W;  (C)  B.  arabicus.  8.9  mm,  29°00'N.  048°29'E;  <D)  B.  ramquamosus.  7.0  mm,  27°4rN,  049°45'E. 


HOUDE:  BREGMACEROTIDAE 


307 


atlanticus  but  larvae  of  the  two  species  are  easily  separated  by 
pigmentation  differences.  There  is  slight  overlap  in  meristics  of 
B.  nectabanus  and  B.  canton,  although  B.  canton  generally  has 
lower  counts.  The  very  wide  range  in  dorsal  and  anal  fin  ray 
counts  attributed  to  B.  nectabanus  possibly  has  resulted  from 
identification  errors. 

Adult  complements  of  median  and  of  caudal  fin  rays  are 
present  at  7.5-9.5  mm  SL.  Three  or  four  pelvic  fin  rays  develop 
early  in  larvae,  most  precociously  in  B.  macclellandi  and  B. 
Type  A,  just  after  appearance  of  the  occipital  ray.  In  other 
species  pelvic  rays  appear  at  3.5-4.5  mm  length  prior  to  ap- 
pearance of  the  occipital  ray.  As  larval  development  proceeds 
an  additional  2-3  pelvic  rays  ossify,  giving  the  adult  comple- 
ment of  5-7  rays. 

Relationships 

Family  relationships.  — The  bregmacerotids  are  gadiform  fishes 
(Fahay  and  Markle,  this  volume)  of  uncertain  affinities  and  with 
no  obvious  close  relatives  (Cohen,  this  volume),  but  generally 
thought  to  be  most  closely  related  to  the  Muraenolepidae,  Mor- 
idae  and  Melanonidae  (Nelson,  1976;  Fahay  and  Markle,  this 
volume).  Although  affinities  are  unclear,  bregmacerotids  are 
clearly  gadiforms.  They  have  high  vertebral  numbers  (Table  83), 
a  long  tail  and  long  median  fins  with  numerous  rays  (Cohen, 
this  volume;  Fahay  and  Markle,  this  volume).  A  well-developed 
caudal  fin,  separate  from  the  dorsal  and  anal  fins,  is  present. 
Accessory  (x  and  \)  bones,  believed  to  be  a  primitive  character 
in  Gadiformes,  are  present  in  the  caudal  complex.  But  the  num- 
ber of  hypurals  has  been  reduced  to  two  inferior  elements  and 
a  platelike  superior  element,  believed  to  represent  fusion  of 
hypural  elements  3-5  (Markle,  1982;  Cohen,  this  volume).  The 
caudal  fin  of  bregmacerotids  is  the  most  symmetrical  in  the 
Gadiformes.  Both  Ahlstrom"  and  Markle  ( 1 982)  have  illustrated 
the  caudal  skeleton  of  a  Bregmaceros  sp.;  Markle's  specimen  is 
undoubtedly  B.  macclellandi.  based  on  meristics  that  he  gives. 
The  number  of  principal  (branched)  caudal  rays  is  12,  among 
the  lowest  in  gadiform  fishes.  Procurrent  (unbranched)  rays  are 
numerous,  20-24  in  number,  equally  divided  between  the  dorsal 
and  ventral  sides  of  the  caudal  complex.  One  principal  ray  is 
associated  with  each  inferior  hypural,  8  are  associated  with  the 
superior  hypural  plate  and  one  is  associated  with  each  epural 
bone.  No  uroneural  is  illustrated  by  Ahlstrom'  but  Markle  ( 1 982) 
illustrated  one  and  noted  that  its  presence  is  unique  among 
gadoid  fishes.  Six  vertebral  centra  appear  to  be  involved  in 
caudal  fin  ray  support.  The  first  dorsal  fin,  which  consists  of  a 
single,  elongate  ray,  is  located  on  the  occiput,  a  unique  condition 
in  gadiform  fishes.  The  pelagic,  tropical-subtropical  distribution 
of  bregmacerotids  is  unusual  among  gadiforms. 

Species  relationships.  —  The  species  of  Bregmaceros  are  remark- 
ably similar.  They  have  wide  geographic  distributions  with  little 
apparent  tendency  to  differentiate  over  their  ranges  of  occur- 
rence. Belyanina  (1974)  discussed  the  evolution  and  dispersal 
of  Bregmaceros.  She  believed  that  the  family  originated  in  the 
Indo-Malayan  Archipelago  from  which  it  dispersed  with  little 
morphological  modification.  The  present-day  richness  of  species 
in  the  Archipelago  and  the  adjacent  northern  Indian  Ocean  lends 
credence  to  that  hypothesis.  Five  species  (B.  macclellandi,  B. 
atlanticus,  B.  nectabanus,  B.  rarisquamosus  and  B.  arabicus) 
presently  occur  in  the  proposed  area  of  origin.  Three  species, 
{B.  bathymaster,  B.  contort  and  B.  Type  A)  do  not  occur  there. 
The  first  two  of  these  resemble  B.  nectabanus  and  may  be  de- 
rived from  it.  The  western  Atlantic  B.  Type  A  is  enigmatic 


Bregmaceros 
Proposed  Species  Relationships 


B.  rtectabanus 


.B  can  fori 
.B.  bathymaster 


B  arabicus 
B.  rarisquamosus 


B.  macclellandi 
B.  atlanticus 

B.  japonlcus 


'<>^\  /Bregmacerotidae 


Fig.  157.  Proposed  species  relationships  of  the  Bregmacerotidae. 
The  possible  relationships,  indicated  by  the  branching  points,  are  based 
on  interpretations  of  species  distributions  and  on  meristic  characters 
and  larval  pigmentation. 


because  it  differs  substantially  from  all  described  species.  Be- 
lyanina (1974,  1980)  believed  that  B.  nectabanus  was  the  com- 
mon neritic  Bregmaceros  in  the  western  Atlantic  but  subsequent 
research  (Milliken,  1975;  Houde,  1981;  Milliken  and  Houde, 
1984)  has  demonstrated  that  the  western  Atlantic  species,  B. 
cantori,  differs  substantially  in  modal  vertebral  numbers  and 
median  fin  ray  counts,  and  also  that  the  larvae  differ  significantly 
in  pigmentation  characteristics. 

Based  on  the  species  characteristics  and  known  distributions, 
possible  relationships  among  species  are  proposed  in  Fig.  157. 
Belyanina  ( 1974)  believed  that  the  two  oceanic  species,  B.  mac- 
clellandi and  B.  atlanticus,  evolved  from  neritic  species.  It  seems 
equally  probable  that  the  neritic  species  evolved  from  the  two 
circumtropical,  oceanic  species.  B.  macclellandi  and  B.  atlan- 
ticus are  very  similar.  They  have  relatively  high  meristic  counts 
and  are  darkly  pigmented.  Their  larvae  are  heavily  pigmented 
and  tend  to  be  deeper-bodied  than  larvae  of  neritic  species.  The 
neritic  species,  except  B.  arabicus.  have  vertebral  numbers  and 
median  fin  ray  counts  much  lower  than  those  of  B.  macclellandi 
and  B.  atlanticus.  As  larvae  the  neritic  species  are  relatively 
thin-bodied  and  lightly  pigmented  (Table  84,  Figs.  1 5  3  and  1 54). 

Bregmaceros  nectabanus.  B.  arabicus  and  B.  rarisquamosus 
overlap  broadly  in  their  ranges  of  occurrence,  as  do  B.  mac- 
clellandi and  B.  atlanticus  and,  to  a  lesser  extent,  B.  cantori  and 
B.  Type  A.  Species  frequently  are  collected  together  as  larvae 


308 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


in  ichthyoplankton  surveys.  Only  B.  hathymaster  appears  to 
live  in  relative  isolation  from  other  species  of  Bregmaceros.  In 
the  Indo-Pacific,  B.  ransquamosus  (small-size,  early  matura- 
tion, low  meristics)  and  B.  arabkus  (high  meristics)  possibly 
were  derived  from  a  B.  nectabanus  stock  intermediate  in  me- 
ristic  characteristics.  The  basic  B.  nectabanus  stock  also  may 
have  given  rise  to  B.  bathymaster  and  B.  cantori.  Detailed  study 
of  eastern  Atlantic  B.  cantori-\'\V.e  larvae  may  help  to  resolve 
questions  about  dispersal  and  evolution  of  species. 

Bregmaceros  Type  A  is  curious.  Like  B.  macclellandi.  its  lar- 
vae develop  the  occipital  ray  at  <2.5  mm  (Table  84).  Yet,  it 
bears  little  resemblance  to  B.  macclellandi  in  other  meristic  or 
pigmentation  characters.  It  has  the  lowest  vertebral  and  median 
fin  ray  counts  of  any  Bregmaceros  except  B.  ransquamosus  (Ta- 
ble 83).  Larvae  of  Type  A  generally  occur  over  the  deep  shelf 
and  slope,  occasionally  in  oceanic  waters,  and  often  co-occur 
with  B.  cantori  and  B.  atlanticus  (Houde,  1981). 

The  status  of  B.  japonicus  is  unclear  although  this  form  may 


be  a  western  Pacific  variety  or  subspecies  of  B.  atlanticus  (Be- 
lyanina,  1974).  A  recent  reexamination  of  the  holotype  (Masuda 
and  Ozawa,  1979)  indicated  that  its  vertebral  and  median  fin 
ray  counts  exceeded  or  were  at  the  upper  extreme  of  ranges 
reported  for  B.  atlanticus  (Table  83).  There  is  a  need  for  critical 
examination  of  B.  atlanticus  and  B.  japonicus  specimens  from 
the  tropical  Pacific  Ocean.  Juveniles  and  adults  that  I  examined' 
from  the  eastern,  tropical  Pacific  appeared  to  be  typical  B.  at- 
lanticus but  none  of  the  small  larvae  had  typical  B.  atlanticus 
pigmentation.  A  moderately  heavily-pigmented  larva  was  pres- 
ent in  tropical  Pacific  collections  that  may  be  an  undescribed 
species.  Its  status  and  its  possible  relationship  to  the  B.  atlan- 
ticus/B.  japonicus  systematics  problem  need  to  be  determined. 

University  of  Maryland,  Center  for  Environmental  and 
EsTUARiNE  Studies,  Chesapeake  Biological  Laboratory, 
Solomons,  Maryland  20688. 


Ophidiiformes:  Development  and  Relationships 
D.  J.  Gordon,  D.  F.  Markxe  and  J.  E.  Olney 


THE  order  Ophidiiformes  contains  300-400  species  occu- 
pying mostly  benthic  habitats  over  a  broad  range  of  depth 
and  salinity.  These  are  elongate,  tapering  fishes  with  or  without 
a  caudal  fin.  The  dorsal  and  anal  fins  are  long,  sometimes  con- 
fluent, without  spines  and  with  pterygiophores  more  numerous 
than  adjacent  vertebrae.  The  pelvic  fins,  if  present,  are  located 
far  forward  and  are  reduced  to  one  or  two  rays,  sometimes  with 
a  small  spine. 

Cohen  and  Nielsen  (1978)  summarized  the  present  under- 
standing of  the  systematics  of  ophidiiform  fishes,  presented  keys 
to  the  genera,  and  provided  a  useful  framework  on  which  to 
base  a  discussion  of  the  order.  The  presence  or  absence  of  vi- 
viparity defines  two  suborders,  Bythitoidei  and  Ophidioidei. 
Bythitoidei  contains  the  live-bearing  "brotulids"  and  is  divided 
into  two  families,  Aphyonidae  and  Bythitidae.  The  oviparous 
Ophidioidei  contains  Ophidiidae  and  Carapidae.  Ophidiidae 
includes  the  cusk-eels  (Ophidiinae)  and  the  oviparous  "brotu- 
lids," previously  allied  with  the  bythitoids  in  the  family  Bro- 
tulidae. 

Aphyonidae,  reviewed  by  Nielsen  (1969),  contains  18  species 
in  five  genera.  These  ovoviviparous  fishes  are  benthopelagic  and 
found  worldwide.  Bythitidae  contains  over  80  species  in  28 
genera.  Most  species  of  this  family  occur  either  in  shallow  trop- 
ical waters,  including  coral  reefs,  or  in  waters  of  intermediate 
depths  on  the  continental  shelf  and  slope.  Some  deeper-dwelling 
slope  species  occur  at  higher  latitudes,  a  few  species  inhabit 
abyssal  waters  and  some  are  found  in  freshwater.  Carapidae 
contains  about  30  species  divided  into  two  subfamilies  (Pyra- 
modontinae,  Carapinae)  and  six  genera,  all  possessing  a  vexil- 
lifer  larva  (Olney  and  Markle,  1979;  Markle  and  Olney,  1980; 


Markle  et  al.,  1983).  Some  species  are  free-living  while  others 
are  inquilines  within  the  body  cavities  of  invertebrate  hosts 
(Trott,  1 970;  Trott,  1981).  Ophidiidae,  as  defined  by  Cohen  and 
Nielsen  (1978),  includes  oviparous  ophidiiform  fishes  lacking  a 
vexillifer  larva  and  possessing  a  supramaxillary  bone.  The  fam- 
ily is  divided  into  four  subfamilies:  Brotulinae,  Brotulotaeni- 
inae,  Neobythitinae  and  Ophidiinae.  Brotulinae  contains  one 
genus  (Brotula)  with  at  least  five  species  (Cohen  and  Nielsen, 
1978).  Adult  Brotula  are  benthic  and  circumtropical  on  the 
continental  shelf  Brotulotaeniinae  contains  the  single  genus 
Brotulotaenia  with  four  midwater,  tropical  representatives  (Co- 
hen, 1974).  Neobythitinae  is  a  morphologically  diverse  group 
containing  38  genera  and  over  135  species  with  worldwide  dis- 
tribution and  a  wide  depth  range,  but  mostly  deep  sea.  Ophi- 
diinae consists  of  about  60  nominal  species  with  several  un- 
described forms  (Lea,  1980),  mostly  in  shelf  waters. 

Development 

Knowledge  of  the  early  life  history  of  ophidiiform  fishes  varies 
considerably  among  major  taxa.  Larvae  of  the  live-bearing  species 
are  infrequently  collected  and  larvae  of  deep  water  forms  are 
even  rarer.  The  incomplete  state  of  knowledge  of  the  taxonomy 
of  bythitoid  fishes  renders  identification  of  most  of  their  larvae 
tentative.  On  the  other  hand,  carapid  and  ophidiine  larvae  are 
common  to  abundant  in  tropical  plankton.  Carapid  larvae  are 
relatively  well  known  and  have  proven  to  be  of  systematic  value 
(Olney  and  Markle,  1979;  Markle  and  Olney,  1980).  Though 
the  larvae  of  only  a  small  percentage  of  the  species  of  ophidiines 
are  known,  these  larvae  provide  useful  characters  for  under- 
standing relationships  within  the  group  (Gordon,  1982). 


GORDON  ET  AL.:  OPHIDIIFORMES 


309 


Fig.  158.  Larvae  of  Bythitoidei.  (A)  Larva  of  Brosmophyas  marginala.  12.5  mm  NL.  NMFS-SWFC.  CalCOFI  7207  .A.x  Sta.  63.52.  (B) 
Unidentified  bythitid  larva,  21.9  mm  SL,  HML  H  4086,  40°34'N,  66''00'W.  (C)  Exterilium  larva  tentatively  assigned  to  Neobythitinae,  29.5  mm 
SL,  MCZ-WHOl,  Oceanus  22,  JEC  771 1,  0°00'N,  37''40'W. 


Eggs  and  embryos— Ophidiiform  eggs  are  poorly  known.  The 
pelagic  eggs  of  Gcnypterus  capensis  (Ophidiidae)  are  moderately 
large,  spherical  and  contain  a  single  oil  globule  (Brownell,  1 979). 
The  few  known  carapid  eggs  are  pelagic,  ellipsoidal,  and  possess 
a  single  oil  globule.  Early  developmental  stages  may  be  con- 
tained in  a  mucilaginous  raft.  Eggs  have  been  described  for 
Carapus  acus  {Emtry,  1880;  Padoa,  1956j).  Echtodon  dentatus 


(Sparta,  1926),  E.  c/n/wwow^/ (Kennedy  and  Champ,  1971),  E. 
rendahli  (Robertson,  1975b),  and  unidentified  carapid  species 
from  the  North  Atlantic  (Ryder.  1884)  and  South  Africa  (Brow- 
nell, 1979). 

Aphyonid  larvae  have  not  been  reported  from  plankton  tows 
but  late  embryos  taken  from  ovarian  tissue  were  illustrated  by 
Nielsen  (1969). 


310 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


5sSSS««*§^'*5''^^^^^^^5^5S'?S?5S^^ 


S*^!«S^!5SS!«Ssi5aS5 


■"■■■    --^^^<^^^^^^^^^^^^<^^^^^^^ -,s=,^,s.*j.^*^-^' 


GORDON  ET  AL.:  OPHIDIIFORMES 


311 


D 

Fig.  1 60.  Larvae  of  the  genus  Lepophidiuln.  (A)  Lepophidium  negropinna.  5. 1  mm  NL.  NMFS-SWFC,  P-28  1 20.35.  (B)  Lepophidiumjeannae, 
1 1  mm  SL.  UMML  CI  71 13  Sla.  62,  26°30'N.  83°00'W.  (C)  Lepophidium  staurophor.  12  mm  SL.  UMML,  CI  71 13  Sta.  81,  27°00'N,  84''05'W. 
(D)  Lepophidium  Type  1,  7.8  mm  NL,  UMML,  CI  71  14  Sta.  127,  28''15'N,  84°50'W. 


Larvae.  — The  reproductive  biology  of  three  bythitid  species  has 
been  discussed  (Wourms  and  Bayne,  1973;  Wourms  and  Cohen, 
1975;  Suarez,  1975).  Aboussouan  (1972a)  described  a  larva  at- 
tributable to  Oligopus  longhursti  and  Leis  and  Rennis  (1983) 
have  illustrated  a  larval  Dinematichlhys.  A  larva  of  Brosino- 
phycis  marginata  from  the  eastern  Pacific  and  an  unidentified 
bythitid  from  the  North  Atlantic  (Fig.  1  58)  are  illustrated  here. 


Larvae  of  a  number  of  carapid  and  ophidiine  species  have  been 
described,  but  few  larvae  of  other,  generally  deeper-dwelling, 
ophidioid  taxa  are  known.  Leis  and  Rennis  (1983)  illustrated  a 
larval  Brolula.  Aboussouan  (1980)  described  a  large,  ribbon- 
shaped  larva  which  he  attributed  to  Bwtidotaenia.  A  specimen 
of  Spectrunctulus  grandts  (56  mm  SL)  is  illustrated  and  dis- 
cussed by  Nielsen  and  Hureau  (1980).  Larvae  of  the  neobythi- 


Fig  159.  Urvae  of  tribe  Ophidimi.  (A)  Olophidium  omosligmum.  8.3  mm  NL,  UMML,  CI  71  14  Sta.  126,  28°15'N,  84°25'W.  (B)  Ophidian 
Type  1,  7.6  mm  NL,  UMML,  CI  7308  Sta.  60,  26°3rN,  82°28'W.  (C)  OphidionType  2,  7.0  mm  NL,  UMML,  CI  7303  Sta.  94,  27°30'N,  83°29'W. 
(D)  Ophidian  nocomis.  24  mm  SL,  NMFS-SEFC,  Ore  II  7343  Sta.  160.  (E)  Ophidian  setenops.  24  mm  SL,  UMML,  CI  71 13  Sta.  95,  27''31'N, 
83°46W.  (F)  Parophidion  schmidli.  17  mm  SL,  MCZ-WHOI,  RV  Chain  60  RHB  1315,  25°46'N,  79°47'W. 


312  ONTOGENY  AND  SYSTEMATICS  OF  FISHES- AHLSTROM  SYMPOSIUM 

Table  85.    Meristic  Variation  in  Western  North  Atlantic  Species  of  Cusk-Eels.  Sample  size  is  indicated  in  parentheses  below  the  range. 


Vertebrae 

Fin  i^ys 

Species 

Precaudal 

Caudal 

Total 

Dorsal 

Anal 

Pectoral 

Source 

Otophidium  omostigmum 

14 

42-44 

56-58 

99-108 

84-87 

16-18 

Gordon.  1982;  Bohlke 

(10) 

(10) 

(10) 

(10) 

(10) 

(20) 

and  Robins.  1959 

Otophidiuin  dormilator 

14,  15 

49-51 

64,65 

111-117 

94-96 

15-16 

Gordon.  1982;  Bohlke 

(11) 

(7) 

(10) 

(14) 

(14) 

(20) 

and  Robins.  1959 

Otophidium  chickcharney 

13 

50-52 

63-65 

111-116 

98-102 

16,  17 

Bohlke  and  Robins. 

(6) 

(6) 

(6) 

(9) 

(8) 

(16) 

1959;  onginal 

Parophidion  schmidti 

15 

52 

67 

115-126 

98-106 

17-19 

Bohlke  and  Robins. 

(1) 

(1) 

(1) 

(24) 

(24) 

(25) 

1959;  original 

Ophidian  selenops 

15,  16 

62-65 

77-81 

132-138 

123-129 

15-16 

Gordon.  1982;  Robins 

(14) 

(14) 

(14) 

(8) 

(8) 

(18) 

and  Bohlke,  1959 

O.  nocomis 

17 

67-70 

84-87 

144-153 

132-139 

14-17 

Robins  and  Bohlke, 

(7) 

(7) 

(7) 

(42) 

(42) 

(74) 

1959 

O.  holbrooki 

15,  16 

50-53 

66-69 

117-132 

97-109 

19-21 

Gordon.  1982 

(12) 

(12) 

(12) 

(12) 

(12) 

O.  beani 

15-17 

50-53 

65-69 

111-133 

94-103 

18-21 

Gordon.  1982 

(45) 

(45) 

(45) 

(9) 

(9) 

O.  grayi 

16 

48-50 

64-66 

131-145 

99-113 

20-23 

Gordon.  1982 

(11) 

(11) 

(11) 

(11) 

(11) 

O.  marginatum 

15 

53-54 

68-69 

147-158 

118-124 

— 

Miller  and 

(4) 

(4) 

(4) 

(4) 

(4) 

Jorgenson,  1973 

O.  welshi 

15,  16 

50-52 

67,68 

128-150 

105-122 

21 

Gordon,  1982 

(6) 

(6) 

(6) 

(15) 

(15) 

O.  lagochila 

— 

— 

— 

123-125 

103-105 

17-19 

Bohlke  and  Robins, 

(3) 

(3) 

(6) 

1959 

Lepophidium  graellsi 

14-16 

55-57 

69-73 

124-133 

101-109 

20-22 

Gordon.  1982 

(40) 

(39) 

(32) 

(38) 

(37) 

L.  marmoratum 

14,  15 

55-60 

70-75 

121-136 

103-112 

21-24 

Gordon,  1982 

(31) 

(31) 

(32) 

(30) 

(31) 

L.  jeannae 

14,  15 

58-60 

73-75 

131-140 

112-117 

20-21 

Gordon,  1982; 

(9) 

(9) 

(9) 

(11) 

(11) 

(12) 

Robins,  1960 

L.  staurophor 

15 

65-67 

80-82 

140-147 

122-127 

22,23 

Robins.  1958 

(3) 

(3) 

(3) 

(4) 

(4) 

(4) 

L.  profundorum 

15-17 

58-61 

73-78 

131-140 

110-121 

22-24 

Gordon.  1982 

(14) 

(14) 

(14) 

(14) 

(14) 

L.  kallion 

15 

59 

74 

133 

108 

23,24 

Robins,  1959 

(1) 

(1) 

(1) 

(1) 

(1) 

(4) 

L.  aporrhox 

13 

52.53 

65.66 

109-114 

96-99 

21-23 

Robins,  1961 

(5) 

(5) 

(5) 

(7) 

(7) 

(14) 

L.  pheromystax 

14.  15 

54-57 

69-72 

125-132 

104-110 

20-22 

Robins,  1960 

(20) 

(20) 

(20) 

(41) 

(41) 

(86) 

tine  genus  Benthocometes  {Pteridum)  were  illustrated  by  Padoa 
(1956i).  Exterilium  larvae  (Fraser  and  Smith,  1974;Moser,  1981) 
may  be  larvae  of  deep-dwelling  neobythitine  species  (Figure 
158C). 

Larvae  of  six  ophidiine  genera  are  known.  Padoa  (19561) 
described  the  larvae  of  Parophidion  vassali  and  Ophidian  bar- 
baturn  from  the  Mediterranean  Sea.  Aboussouan  (1972a)  de- 
scribed the  larvae  of  Ophidian  barbatum  from  the  eastern  At- 
lantic. Brownell  (1979)  reared  early  stages  of  the  larvae  oi  Gen- 
ypterus  capensis.  Larval  stages  of  Ophidian  marginatum  were 
illustrated  in  Scotten  et  al.  (1973)  and  were  reproduced  in 
Fritzsche  (1978).  Figure  178b  in  Fntzsche  (1978)  is  probably  a 
stichaeid  and  not  an  ophidioid  species.  Larval  stages  of  Oto- 
phidium omostigmum  (Fig.  1 59A),  Ophidion  selenops  (Fig.  1 59E), 
Lepophidium  jeannae  (Fig.  160B)  and  Lepophidium  staurophor 
(Fig.  160C)  from  the  Gulf  of  Mexico  were  described  by  Gordon 
(1982).  Larvae  of  Chilara  taylari  and  Ophidian  scrippsae  were 


described  by  Ambrose  et  al.  ( 1 983).  Larvae  of  Ophidion  nocomis 
(Fig.  159D)  and  Lepophidium  negropinna  (Fig.  160A)  are  illus- 
trated in  this  study. 

Carapid  larvae  have  been  described  (or  illustrated)  for  six 
genera  and  12  species:  Carapus  acus  (Psidoa.  I956j);  C.  imberbis 
(Aboussouan,  1972a);  C.  berniudensis  {Dawson,  1971b;  Olney 
and  Markle,  1979);  Echiodon  dentatus  (Emery,  1880;  Sparta, 
1926;  Padoa.  1947;  Maul.  1976);  E.  drummondi  (\ihrtinha\im, 
1905-1909;  Kennedy  and  Champ,  1971);  Echiodon  rendahli 
(Robertson,  1975b);  E.  dawsoni  (Olney  and  Markle,  1979);  E. 
c.xsilium  (Trott,  1970;  Olney  and  Markle,  1979);  Encheliophis 
jordani  (Tron,  1970);  Onuxodon  margaritiferae  (Gowoni  et  al., 
1984);  Snyderidia  canina  (Strasburg,  1965;  Markle  and  Olney, 
1980);  and  Pyramodon  wnlralis  (Markle  and  Olney,  1980).  In 
some  cases,  larval  identifications  are  unsubstantiated  and  cau- 
tion should  be  employed,  especially  in  the  older  literature.  For 
example,  Padoa  ( 1 956j)  confuses  larval  E.  dentatus  (plate  XLIV, 


GORDON  ET  AL.:  OPHIDIIFORMES 


313 


Table  86.    Meristic  Variation  in  Eastern  North  Pacific  Species  of  Cusk-Eels.  Sample  size  is  indicated  in  parentheses  below  the  range. 


Vertebrae 

Fin  rays 

Species 

Precaudal 

Caudal 

Toial 

Dorsal 

Anal 

Pectoral 

Source 

Otophidium  indefatigabile 

13-15 

45-49 

59-64 
(17) 

106-115 

(15) 

88-96 

(14) 

18-19 

Lea, 

1980 

Chilara  taylori 

18-19 

68-72 

86-91 
(66) 

187-229 
(50) 

150-181 
(50) 

24-26 

Lea, 

1980 

Ophidian  costaricense 

14-16 

50-54 

65-69 

(45) 

130-153 
(44) 

107-128 
(43) 

23-26 

Ua, 

1980 

O.  fulvum 

13-15 

49-55 

63-69 
(39) 

137-160 
(34) 

112-136 
(34) 

23-26 

Lea, 

1980 

O.  galeoides 

14-17 

47-49 

61-64 
(52) 

123-143 
(42) 

97-114 

(41) 

21-23 

Lea, 

1980 

O.  imitator 

15-16 

55-60 

70-76 
(20) 

135-163 
(18) 

112-139 
(18) 

25-28 

Lea, 

1980 

O.  ins 

16-17 

53-56 

69-73 
(67) 

121-148 
(65) 

98-122 
(63) 

22-24 

Lea, 

1980 

O.  moche 

15-16 

55-58 

71-74 
(6) 

142-148 
(6) 

118-126 
(6) 

24-25 

Lea, 

1980 

O.  scrippsae 

14-16 

50-54 

65-69 
(102) 

124-153 
(102) 

99-126 
(100) 

20-23 

Ua, 

1980 

Lepophidium  prorates 

14-16 

55-58 

70-73 

124-133 

106-113 

21-24 

Robins,  1962 

(63) 

(59) 

(59) 

(59) 

(60) 

(87) 

L.  pardale 

15 

57 

71 

128,  132 

106,  109 

22,23 

Robins,  1962 

(2) 

(2) 

(2) 

(2) 

(2) 

(2) 

L.  stigmatislium 

14.  15 

55,56 

70 

125,  130 

103,  109 

19-21 

Robins,  1962 

(2) 

(2) 

(2) 

(2) 

(2) 

(4) 

L.  microlepis 

14-16 

51-62 

66-77 

117-141 

97-121 

21-26 

Robins  and  Lea,  1978 

(83) 

(83) 

(83) 

(81) 

(82) 

(95) 

Cherublemma  emmelas 

13,  14 

42-44 

55-58 

99-113 

81-93 

24-26 

Robins,  1961 

(14) 

(14) 

(14) 

(33) 

(32) 

(58) 

L.  negropinna 

15,  16 

59-61 

75,77 

138-148 

116-121 

21-24 

Robins,  1962 

(5) 

(5) 

(5) 

(12) 

(12) 

(22) 

Fig.  14;  plate  XLV,  Figs.  3  and  4)  with  C.  acus  (plate  XLV,  Fig. 
5). 

Prejuveniles. —Ophidiine  lai^ae  are  pelagic,  and  the  develop- 
ment of  most  species  proceeds  directly  without  an  abrupt  tran- 
sition period.  The  larvae  of  Ophidian  selenops,  however,  are 
extremely  elongate  and  attain  a  length  of  40  mm  SL  before 
reduction  of  intervertebral  spaces  causes  a  reduction  in  total 
length  to  about  24  mm  (Gordon,  1982).  Soon  after  this  trans- 
formation the  body  shape  approaches  the  juvenile  form  and  the 
larvae  become  benthic.  Elongate  larvae  identified  here  as  O. 
nocomis  have  a  similar  morphology  (Fig.  1 59D).  Most  ophidiine 
species  probably  become  benthic  at  about  25-30  mm  SL.  Chi- 
lara taylori  and  Farophidion,  however,  have  extended  nektonic 
prejuvenile  stages  (Lea,  1980).  The  prejuvenile  stage  of  C  tay- 
lori reaches  80  mm  SL  and  was  described  as  Ophidion  nova- 
culum  by  Harry  (1951). 

A  specialized  prejuvenile  stage,  known  as  a  tenuis  larva,  has 
been  described  for  some  Carapus  species  and  Echiodon  dawsoni 
and  is  characterized  by  the  absence  of  a  vexillum  and  an  initial 
lengthening  and  subsequent  reduction  in  total  length  (Emery, 
1880;  Arnold,  1956;  Padoa,  1956j;  Strasburg,  1961;  Hipeau- 
Jacquotte.  1967;Gustato,  1976;Trott,  1981;  Williams  and  Shipp, 
1982).  The  stage  is  poorly  known  and  has  been  reported  as  an 
obligate  inquiline  parasite  (Trott,  1981),  a  free-living  benthic 
form  (Trott.  1981),  and  a  pelagic  form  sometimes  attracted  to 
nightlights  (Smith  et  al.,  1981). 


Meristic  characters.  — MensUc  characters  that  are  observable  in 
ophidiine  larvae  include  myomere  number,  vertebral  formula 
(precaudal  plus  caudal  vertebrae),  dorsal  fin  ray  number  and 
anal  fin  ray  number.  Pectoral  fin  ray  number  and  gill  raker 
development,  which  are  important  taxonomically  in  adults,  can- 
not be  considered  complete  in  pelagic  larvae.  Meristic  characters 
show  large  variation  within  species.  In  many  cases,  published 
ranges  for  these  characters  are  based  upon  too  few  specimens 
to  accurately  depict  the  range  of  variation.  In  addition,  meristic 
data  show  broad  range  overlap  between  several  species  and 
caution  should  be  employed.  Positive  identification  of  larvae 
based  solely  on  meristic  characters,  however,  can  be  made  for 
some  western  Atlantic  species,  including  Ophidion  selenops.  O. 
nocomis,  Otophidium  omostigmum.  Oto.  dormitator.  Oto. 
chickcharney  and  Lepophidium  staurophor.  In  the  eastern  Pa- 
cific, larvae  of  Chilara  taylori  and  Otophidium  indefatigabile 
are  identifiable  based  on  meristics.  Ranges  of  meristic  characters 
for  western  Atlantic  ophidiines  are  given  in  Table  85  and  for 
eastern  Pacific  ophidiines  in  Table  86.  Several  species  of  the 
genera  Ophidion  and  Lepophidium  from  the  western  Atlantic 
are  presently  undescribed.  and  taxonomic  questions  remain  to 
be  resolved  (C.  R.  Robins,  pers.  comm.). 

Development  of  the  ophidiine  vertebral  column  and  fins  was 
described  by  Gordon  ( 1 982).  Total  myomeres  in  larvae  compare 
closely  with  total  vertebrae  in  adults.  The  number  of  preanal 
myomeres  present  prior  to  coiling  of  the  gut  is  usually  greater 
than  the  number  of  precaudal  vertebrae  in  adults,  because  the 


314 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  87.  Meristic  Variation  in  Selected  Species  of  Carapid  Fishes.  Abbreviations  used  are:  N  — number  of  specimens  examined;  D,o— 
number  of  dorsal  rays  whose  bases  lie  anterior  to  31st  vertebra;  Ajo— number  of  anal  rays  whose  bases  he  anterior  to  31st  vertebra;  P,  — pectoral 
rays;  P^— pel  vie  rays;  PCV—precaudal  vertebrae;  NVD— number  of  vertebrae  to  dorsal  origin;  NVA— number  of  vertebrae  to  anal  origin;  ARDO— 

number  of  anal  rays  to  dorsal  ongin;  NA— not  applicable. 

Species  N  D^  A^  P^  p]  PCV  NVD  NVA  ARDO 


Pvramodon  ventralis 

10 

48-52 

46-53 

27-29 

1 

15-18 

5-8 

6-8 

NA 

Snydehdia  canina 

4 

49-51 

44-46 

25 

abs 

14-15 

6-7 

9-10 

NA 

Carapus  acus 

2 

37-39 

58 

abs 

18 

11-12 

3-4 

20-21 

Carapus  inourlani 

3 

38 

56-57 

18 

abs 

15-16 

12-13 

2-3 

22 

Carapus  homei 

2 

33-37 

56-60 

17-19 

abs 

16-17 

14 

3-4 

24-26 

Carapus  pan^ipinnis 

4 

35-38 

50-53 

15-19 

abs 

16 

12-13 

4-5 

17-19 

Echiodon  drummondi 

6 

42-45 

47-49 

15-17 

abs 

25-29 

8-9 

6-8 

5-7 

Echiodon  cryomargarites 

24 

37-40 

46-50 

19-21 

abs 

25-29 

11-12 

6-8 

9-12 

Echiodon  dawsoni 

1 

28 

39 

17-18 

abs 

22 

12 

6 

12 

Onuxodon  parvibrachium 

4 

44-46 

44-48 

14-15 

abs 

16-17 

6-7 

6-8 

1-2 

Onuxodon  margantiferae 

6 

46-47 

46-50 

16 

abs 

19-22 

6-8 

5-6 

3-4 

Encheliophis  vermicularis 

3 

25-26 

39-41 

abs 

abs 

21-22 

16-18 

4-5 

18-21 

Encheliophis  gracilis 

4 

28-31 

45-50 

15-17 

abs 

26-31 

16-17 

3-7 

23-26 

gut  migrates  forward  by  2-4  myomeres  during  formation  of  the 
gut  coil  (Gordon,  1982).  By  8-10  mm  NL,  the  haemal  arches 
are  closed  allowing  accurate  determination  of  the  vertebral  for- 
mula in  cleared  and  stained  larvae.  Rays  of  the  dorsal  and  anal 
fins  develop  from  posterior  to  anterior.  Development  begins  at 
7-10  mm  NL  and  is  complete  by  15-20  mm  SL.  The  adult 
complement  of  nine  caudal  rays  and  seven  branchiostegal  rays 
is  present  by  10  mm  NL  in  most  species.  These  structures  do 
not  appear  until  1 5  to  20  mm  SL  in  the  elongate  O.  selenops. 
The  number  of  pectoral  fin  rays  ranges  from  16-28,  with  sizes 
at  which  the  first  rays  appear  ranging  from  1 3-20  mm  SL.  The 
pectoral  fin  is  complete  in  some  species  by  18-20  mm  SL. 

Traditionally,  meristic  characters  have  not  been  widely  used 
in  adult  descriptions  of  carapids  (Arnold,  1956)  and  conse- 
quently, some  easily  observed  and  useful  characters  such  as 
pectoral  fin  ray  counts  (Cohen  and  Nielsen,  1978;  OIney  and 
Markle,  1979)  can  not  be  obtained  from  the  literature.  Inno- 
vative meristics,  partly  borrowed  from  eel  systematics  (Nielsen 
and  Smith,  1978),  are  useful  aids  to  identification  of  larval  and 
adult  carapids  (OIney  and  Markle,  unpublished  data).  Table  87 
summarizes  some  of  these  meristic  characters  for  selected  species 
of  carapids. 

Morphology.  —  Except  for  the  larvae  of  O.  selenops  and  O.  no- 
comis,  known  ophidiine  larvae  show  little  variation  in  size, 
shape  and  development  (Figs.  159,  160).  The  larvae,  which 
hatch  at  2-3  mm  NL,  are  moderately  elongate  and  taper  slightly 
from  the  head  to  the  end  of  the  notochord.  The  eyes  are  round 
and  conspicuous;  the  mouth  is  oblique.  Larvae  become  more 
laterally  compressed  with  growth.  In  all  species  examined  by 
Gordon  (1982),  the  gut  is  straight  at  hatching  and  develops  a 
single  coil  at  5-7  mm  NL  as  a  downward  loop  twists,  displacing 
the  more  posterior  portion  of  the  gut  to  the  right. 

Carapid  vexillifers  (Figs.  161-163)  are  elongate  larvae  with  a 
moderately  sized  head,  large  eye  and  nasal  rosette,  coiled  gut, 


short  preanal  length,  and  tapered  body  frequently  ending  in  a 
broken  filament,  and  an  elongate  larval  dorsal  fin  ray  (vexillum) 
in  front  of  the  first  adult  dorsal  fin  ray  (OIney  and  Markle,  1979; 
Govoni  et  al.,  1 984).  Larvae  of  Pyramodon  and  Snyderidia  (Fig. 
1 63)  have  a  somewhat  deeper  head  and  trunk,  shorter  pre-dorsal 
distance,  relatively  long  anal  fin  pterygiophores,  and  more  pec- 
toral fin  rays  than  other  carapid  larvae  (Markle  and  OIney, 
1980).  Variations  in  gross  morphology  in  carapine  larvae  are 
limited  to  variation  in  gut  shape  and  fin  ray  or  vexillum  position 
(Figs.  161,  162;  Table  87). 

Pigmentation.  — VxgmenXdiXion  of  ophidiine  larvae  is  useful  for 
identifying  species  and  species  groups  though  care  must  be  taken 
since  ontogenetic  changes  occur  (Gordon,  1 982).  Head  pigmen- 
tation typically  consists  of  two  or  three  melanophores  present 
distally  on  the  suspensorium  near  the  articulation  with  the  lower 
jaw.  Abdominal  pigmentation  is  usually  variable  within  species 
and  consists  of  melanophores  scattered  ventral  to  the  gut.  Pig- 
mentation on  the  posterior  half  of  the  body  is  the  most  useful 
for  taxonomic  purposes.  All  Lepophidium  larvae  have  2-10 
large  spots  placed  medially  along  the  base  of  the  anal  finfold 
and  1-2  spots  dorsally  in  the  caudal  peduncle  region.  Unlike 
Lepophidium,  larvae  of  Ophidion,  Otophidium  and  Chilara  have 
patterns  of  small  stellate  melanophores  present  laterally  on  the 
body.  Several  species  of  these  genera  can  be  recognized  on  the 
basis  of  postanal  pigmentation.  Some  species  have  larvae  which 
are  very  similarly  pigmented:  O.  selenops  and  O.  nocomis;  O. 
welshi  and  O.  marginatum.  Larvae  of  other  species  cannot  be 
distinguished  using  pigmentation  or  meristic  characters  (Gor- 
don, 1982):  Ophidion  Type  1  (Fig.  159B)  which  represents  O. 
holbrooki,  O.  beani  and  unidentified  Ophidion  species;  Ophidion 
Type  2  (Fig.  1 59C)  O.  welshi  and  O.  grayi;  and  Lepophidium 
Type  1  (Fig.  160D)  L.  graellsi  and  L.  jeannae. 

Carapid  vexillifers  are  sparsely  to  moderately  pigmented  (Figs. 
161-163).  Red  chromatophores  have  been  noted  in  fresh  ma- 


Fig.  161 .  (A)  Urva  of  Echiodon  drummondi,  76.5  mm  TL,  ZMUC  uncat.,  DANA  St.  8371,  5  r29'N,  1 2''50'W.  (B)  Urva  of  Echiodon  rendahli. 
72  mm  TL,  CSIRO  uncat.,  Warreen  Cruise,  Sta.  266/39,  36°17'S,  150°25'E.  (C)  Dorsal  fin  and  vexillar  supports  of  Echiodon  dawsoni.  4.2  mm 
HL  and  10  mm  HL.  Abbreviations  used  are  vex  — vexillum,  fdp— first  dorsal  pterygiophore.  (D)  Dorsal  fin  and  vexillar  supports  of  Echiodon 
drummondi.  4.6  mm  HL  and  24.3  mm  HL.  Abbreviations  as  in  Fig.  16 IC. 


GORDON  ET  AL.:  OPHIDIIFORMES 


315 


42mm  HL 


fdp 


46  mm  HL 


D 


/ 


fdp 


10 mm  HL 


24  3  mm  HL 


316 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  162.     Larva  of  Carapus  sp.,  (top)  uncat.,  Anton  Bruun,  CH'S,  65°03'E,  27-28  May  1964.  Larva  of  Onuxodon  panihrachium.  (bottom) 
ZMUC  uncat.,  Dana  St.  3768  XVI,  400  mwo,  1°20'S,  138°42E,  0315  hrs.  25  July  1929. 


GORDON  ET  AL.;  OPHIDIIFORMES 


317 


■'^■'^^^--^■<'«l/4'0^///////.'^^ 


///y/^/^y^^y. 


^■s-^--^. 


\\\\\V^wvv\\\^^n;^v;x;n^\^\\xx^-^v.-xx- 


•^y'^^r^- 


Fig.  163.     Larva  of  Pyramodon  ventralis.  (top)  uncat.,  21°20-30'N,  158°20-30W,  19  Dec.  1977.  Larva  oi  Snydehdia  bothrops.  (bottom)  MCZ 
uncat.,  RHB  1263,  Chain  60,  12°58'N,  73°34W,  29  May  1966,  0-120  m,  IKMT. 


terial  of  Echiodon  dentatus  (Padoa,  1956j)  and  E.  dawsoni  (Ol- 
ney  and  Markle,  1979)  but  are  not  normally  retained  in  pre- 
served material.  Meianophores  are  variously  present  at  the 
symphysis  of  the  lower  jaw,  on  the  snout,  head,  vexillum,  swim- 
bladder,  trunk,  and  tail.  Preliminary  studies  (Padoa,  1 956j;  Rob- 
ertson, 1975b;  OIney  and  Markle,  1979;  Markle  and  Olney, 
1980)  indicate  that  pigmentation  may  be  regionally  useful  as  an 
aid  to  identification  but  seems  problematic  as  an  indicator  of 
higher  relationships. 

Osteology.— The  placement  of  the  pelvic  fins,  which  defines  the 
subfamily  Ophidiinae,  shows  marked  ontogenetic  change  (Gor- 
don, 1982).  In  early  larvae,  the  cleithra  lack  the  forward  exten- 
sion and  the  pelvic  fins  (appearing  by  about  7  mm  NL)  are 
supported  in  a  jugular  position.  By  20  mm  SL,  the  bony  exten- 
sion of  the  cleithra  develops  and  begins  to  elongate  anteriorly. 
The  pelvic  fins,  which  are  supported  at  the  symphysis,  migrate 
forward  and  are  present  in  the  characteristic  mental  position  in 
the  juveniles.  The  presence  of  pelvic  fins  in  the  jugular  position 
has  occasionally  caused  the  confusion  of  early  larvae  with  other 
ophidioids. 

The  general  structure  of  the  vexillar  ray  is  described  by  Olney 
and  Markle  (1979)  and  Govoni  et  al.  ( 1 984).  External  variations 
of  vexilla  are  in  length,  ornamentation,  pigmentation,  and  po- 
sition. Some  variation  such  as  length  and  ornamentation  ap- 


pears to  be  an  artifact  (Govoni  et  al.,  1984).  In  several  species, 
the  vexillar  pigmentation  and  ornamentation  are  curiously  re- 
peated in  the  caudal  filament  (Fig.  162).  Variation  in  the  sup- 
porting proximal  radial  is  seen  in  its  shape,  its  position  relative 
to  the  first  adult  dorsal  fin  ray  and  to  vertebrae,  and  in  fusion 
with  the  proximal  radial  of  the  first  dorsal  fin  ray.  In  addition, 
the  supporting  proximal  radial  may  or  may  not  be  retained  in 
adults  (Fig.  161C,  D).  Its  retention  provides  a  means  of  iden- 
tifying the  location  of  the  vexillum  and  can  aid  in  larval  iden- 
tification. Its  absorption,  however,  appears  to  have  occurred 
independently  in  several  genera.  In  the  pyramodontines  (Markle 
and  Olney,  1980)  and  Carapus  (Olney  and  Markle,  1979)  there 
is  also  an  accessory  cartilage  in  front  of  the  second  neural  spine. 
Its  origin  and  function  are  unknown.  Carapus  (and  presumably 
Encheliophis)  and  the  pyramodontines  also  have  the  most  for- 
wardly  placed  vexilla,  usually  above  or  in  front  of  the  first  anal 
fin  ray.  Carapus  (and  presumably  Encheliophis)  differs  from  the 
pyramodontines  and  all  other  carapids  in  displacement  of  the 
first  adult  dorsal  fin  ray  far  posteriad  of  the  vexillum  (Fig.  162). 
Modified  ribs  on  the  anterior  vertebral  centra  of  carapids  and 
ophidiines  are  associated  with  sound  production/reception  (Rose, 
1961;  Courtenay  and  McKittrick,  1970;  Courtenay,  1971)  and 
develop  in  early  stages  in  carapids  (Olney  and  Markle,  1979). 
In  Carapidae,  the  first  two  ribs  are  movable  and  all  subsequent 
ribs  are  rigid  (Markle  et  al.,  1983).  A  simple  recurved  third  rib 


318 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


is  found  in  Echiodon  and  pyramodontines  while  an  expanded 
third  rib  is  found  in  Carapus  (Olney  and  Markle,  1979;  Markle 
and  Olney,  1980;  Williams  and  Shipp,  1982)  and  Oimxodon 
(Courtenay  and  McKittrick,  1 970).  The  sexually  dimorphic  and 
interspecific  differences  in  swimbladder  morphology  of  ophi- 
diines  appear  only  in  juveniles  and  adults  and  are  not  useful  in 
distinguishing  larvae. 

The  visceral  cradle,  formed  from  the  criss-crossing  elongate 
proximal  pterygiophores  of  the  anterior  anal  fin  rays,  is  a  unique 
specialization  of  Pyramodon  (Markle  and  Olney,  1980).  Its  pre- 
sumed precursor,  non-crossing  elongate  proximal  pterygio- 
phores, is  found  in  larval  Snydehdia.  The  elongate  proximal 
pterygiophores  found  in  pyramodontines  are  conspicuous  in 
larvae. 

The  pectoral  fin  of  carapids  is  a  variable  structure  and  po- 
tentially useful  in  the  study  of  relationships  as  well  as  for  iden- 
tification. Adults  of  some  species  of  Encheliophis  completely 
lack  a  pectoral  fin  while  pyramodontines  have  a  well-developed 
fin  with  up  to  29  rays.  Most  cleared  and  stained  carapid  and 
ophidiine  larvae  have  an  elongate,  cartilaginous,  ventral  process 
of  the  coracoid  (VPC).  In  the  carapid  "exterilium"  larvae  (Fig. 
16  IB,  see  also  Robertson,  1975b)  the  development  of  the  VPC 
has  been  carried  to  an  extreme.  The  hanging  or  trailing  gut  of 
this  larva  is  supported  by  a  skeleton  of  the  two  VPC's  which 
intertwine  with  the  intestine.  Support  of  a  trailing  gut  by  VPC's 
is  not  unique  since  we  have  also  seen  it  in  the  ophidioid  "ex- 
terilium" (Fraser  and  Smith,  1974;  Moser,  1981)  and  Symphii- 
rus  minor  (unpublished  data). 

The  dentition  of  carapids  is  useful  for  adult  identification 
purposes  (Arnold,  1956)  and  enlarged  canines  as  well  as  the 
dentary  diastema  have  been  used  to  separate  Carapus  and 
Echiodon  larvae  (Olney  and  Markle,  1979). 

Relationships 

Intra-ordinal  relationships.— T\\t  classification  of  Ophidi- 
iformes  proposed  by  Cohen  and  Nielsen  (1978)  differs  most 
significantly  from  earlier  classifications  in  the  use  of  mode  of 
reproduction  as  a  subordinal  character.  Previous  classifications 
recognized  the  highly  specialized  carapids  as  either  one  or  two 
families  (Carapidae  and  Pyramadontidae)  and,  based  on  the 
position  of  the  pelvic  fins,  divided  the  remaining  ophidiiforms 
into  two  groups,  the  ophidiids  (ophidiines,  pelvics  mental)  and 
the  brotulids  (pelvics  absent  or  jugular). 

Relationships  within  the  Bythitoidei  remain  unclear.  The 
aphyonids  share  a  number  of  neotenic  characters  serving  to 
define  the  family.  This  may  be  a  polyphyletic  group,  however, 
with  common  character  states  reflecting  convergent  trends  (Co- 
hen and  Nielsen,  1 978).  Comparisons  ofembryonic  adaptations, 
such  as  trophotaeniae  (Wourms  and  Cohen,  1975),  may  prove 
useful  in  resolving  systematic  problems  within  Bythitidae.  Two 
subfamilies  (Bythitinae  and  Brosmophycinae)  are  presently  de- 
fined on  the  basis  of  confluence  of  anal  and  dorsal  fins  with  the 
caudal  fin,  though  neither  subfamily  has  been  adequately  stud- 
ied. 

Ophidioidei  is  defined  by  the  presence  of  oviparity  and  the 
anterior  nostril  (in  most  genera)  well  above  the  upper  lip.  The 
relationships  of  the  ophidioid  subfamilies  are  also  uncertain  and 
the  suborder  may  be  paraphyletic.  Carapidae  and  subfamily 
Ophidiinae  each  seem  to  form  natural  groupings  based  upon 
well-defined  synapomorphies.  Further  study  of  the  neobythi- 
tines  may  reveal  several  natural  groupings  (Cohen  and  Nielsen, 
1978).  The  relationships  of  Brotulotaeniinae  and  Brotulinae  are 
unknown. 


Two  tribes  of  Ophidiinae  can  be  defined  on  the  basis  of  squa- 
mation  and  the  presence  of  pyloric  caecae.  Lepophidiini  (im- 
bricate scales;  pyloric  caecae  present)  contains  three  genera:  the 
monotypic  Cherublemma  emmelas.  Genypierus,  and  Lepophi- 
diiim.  Lea  (1980)  has  proposed  the  elevation  of  Genypterus  to 
the  level  of  tribe.  The  Ophidiini  (anguilloid  squamation;  pyloric 
caeca  absent)  contains  the  genera  Ophidian,  Otophidium.  Chi- 
lara,  Raneya  and  Parophidion.  These  genera,  established  on  the 
basis  of  meristics.  morphometries,  swimbladder  morphology 
and  squamation,  are  not  well-defined  and  require  further  study. 

A  comparative  study  of  the  development  of  ophidiine  larvae 
of  three  nominal  genera,  Ophidion.  Otophidium  and  Lepophi- 
dium,  suggests  that  body  shape,  development  of  the  caudal  fin 
and  pigmentation  can  provide  useful  taxonomic  characters 
(Gordon,  1982).  The  body  shape  and  development  of  Lepo- 
phidium  larvae  may  represent  the  primitive  state  for  the  subfam- 
ily. Otophidium  omostigmum  and  most  Ophidion  species  retain 
this  morphology,  as  does  Parophidion  (Fig.  1 59F;  Padoa,  1 956i). 
The  morphology  and  development  of  O.  selcnops  and  O.  no- 
comis.  however,  differ  markedly  from  that  of  other  ophidiine 
larvae.  The  possession  of  an  elongate  larva  is  a  derived  character 
uniting  these  two  species. 

Robins  and  Bohike  (1959)  recognized  the  close  relationship 
between  O.  selenops  and  O.  nocomis  based  upon  the  shared 
possession  of  a  well-developed  rostral  spine,  similar  to  that 
found  in  Lepophidium,  and  the  tendency  for  the  dorsal  fin  to 
originate  relatively  far  back  on  the  body.  The  larvae  of  Chilara 
taylori  are  slightly  more  elongate  than  typical  ophidiine  larvae, 
but  bear  no  close  similarity  to  the  larvae  of  O.  selenops. 

A  character  shared  by  all  Lepophidium  larvae  examined  by 
Gordon  ( 1 982)  is  the  presence  of  an  elongate  cartilaginous  epural 
which  ossifies  by  15  mm  SL.  All  larvae  of  the  tribe  Ophidiini 
develop  a  short  cartilaginous  epural  by  10  mm,  but  the  epural 
never  ossifies  and  is  not  visible  by  15  mm  SL.  The  presence  of 
an  epural  in  the  caudal  skeleton  of  the  adults  is  presumably  the 
primitive  character  state  for  the  subfamily. 

The  shared  pigmentation  pattern  oi  Lepophidium  larvae  unites 
these  species.  This  character  may  not  extend  throughout  the 
tribe,  however.  Brownell  (1979)  illustrates  a  Genypterus  larva 
(day  eight)  that  has  a  pigmentation  pattern  similar  to  that  of 
Lepophidium.  Other  stages  resembled  Ophidion,  however,  with 
stellate  melanophores  scattered  laterally  on  the  body.  The  pos- 
session of  similarly  pigmented  larvae  by  closely  related  species 
in  Ophidion  argues  for  the  validity  of  pigmentation  as  a  character 
to  show  phyletic  relationships.  The  larval  pigmentation  of  O. 
selenops  and  O.  nocomis  shows  only  slight  differences  as  does 
larval  pigmentation  of  O.  welshi  and  O.  marginatum.  If  the 
proposed  identities  of  Ophidion  Type  I  and  Type  2  and  Le- 
pophidiumType  1  are  correct  (Gordon,  1982),  then  species  which 
these  types  represent  are  presumably  closely  related. 

Adult  carapids  are  morphologically  conservative  and  present 
some  difficulty  in  identification  and  elucidation  of  phylogenetic 
relationships.  Larvae,  on  the  other  hand,  are  reasonably  well- 
known  for  all  genera,  fall  into  fairly  distinct  morphological  groups 
and  provide  morphological  diversity  which  is  potentially  useful 
in  understanding  intra-familial  relationships  (Olney  and  Mar- 
kle, 1979;  Markle  and  Olney,  1 980).  Robins  and  Nielsen  ( 1 970) 
and  Cohen  and  Nielsen  (1978)  recognized  a  single  family,  Carap- 
idae, consisting  of  two  subfamilies:  the  Pyramodontinae  with 
two  genera,  Pyramodon  and  Snyderidia:  and  the  Carapinae  with 
four  genera,  Carapus,  Echiodon,  Encheliophis  and  Onu.xodon. 
However,  Gosline  ( 1 960b)  and  Trott  (1981)  considered  the  Pyr- 
amodontidae  a  separate  family  while  Arnold  (1956)  ignored  this 


GORDON  ET  AL.:  OPHIDIIFORMES 


319 


group  in  his  revision  of  carapids.  Williams  { 1 984)  in  his  synopsis 
considered  it  as  a  subfamily.  The  common  possession  of  a  vex- 
illifer  larva  is  the  most  convincing  evidence  that  the  genera  of 
Carapidae  are  monophyletic,  thus  we  recognize  one  family. 

The  genera  Pyramodon  and  Snyderidia  were  considered  closely 
related  by  Robins  and  Nielsen  (1970),  and  Markle  and  Olney 
(1981),  on  the  basis  of  osteological  and  larval  characters,  added 
further  support  to  this  presumed  relationship.  It  now  appears 
that  many  of  the  character  states  of  these  genera  are  primitive. 
The  pelvic  fins,  lost  in  all  other  carapids,  are  obviously  a  prim- 
itive state  since  they  are  widely  present  in  all  other  ophidiiforms. 
Similarly,  the  dorsal  origin  is  over  or  in  advance  of  the  anal  fin 
in  all  non-carapid  ophidiiforms  as  well  as  in  Pyramodon  and 
Snyderidia.  The  posterior  placements  of  the  first  dorsal  fin  ray 
or  vexillum  can  therefore  be  considered  advanced  states.  Thus, 
the  anterior  placement  of  the  vexillum  relative  to  first  anal  ray 
(a  primitive  state)  in  combination  with  a  posteriorly  placed  first 
dorsal  fin  ray  (advanced  state)  appears  to  define  larvae  of  Car- 
apiis  (Fig.  162)  and  presumably  Encheliophis.  The  genera  pos- 
sess further  derived  states  such  as  adult  inquiline  behavior  and 
parasitism  (Trott,  1970).  In  addition,  the  tenuis  stage,  unknown 
in  Pyramodon  and  Snyderidia.  may  represent  an  advanced  state, 
namely  retaining  larval  characters  in  the  early  benthic  stage. 

Larvae  of  the  genus  Echiodon  display  a  wide  variety  of  mor- 
phology especially  in  gut  configuration,  vexillum  and  first  dorsal 
fin  ray  position  and  dorsal  pterygiophores  (Fig.  161;  Maul,  1976; 
Olney  and  Markle,  1979;  Markle  et  al.,  1983).  Williams  and 
Shipp  (1982)  consider  Echiodon  to  be  composed  of  two  species 
complexes  and  the  gross  morphology  of  larvae  seems  to  support 
this  contention.  In  addition,  the  peculiar  gut  configuration  of 
E.  rendahli  (Fig.  161B,  Robertson  1975b)  represents  another 
extreme  in  morphological  variability  which  suggests  the  genus 
(as  presently  known)  is  polyphyletic. 

Inter-ordinal  relationships.  — Based  upon  anatomical  similari- 
ties shared  with  the  cods,  the  ophidiiform  fishes  have  been 
treated  as  a  suborder  within  Gadiformes  (Greenwood  et  al., 
1966;  Rosen  and  Patterson,  1969).  These  similarities  include 
the  development  of  the  levator  maxillae  superioris  and  the  struc- 
ture of  the  caudal  skeleton.  Freihofer  (1963,  1970)  presented 
further  evidence  for  this  relationship  based  upon  the  pattern  of 
the  ramus  lateralis  accessorius  nerve.  Alternatively,  these  sim- 
ilarities may  be  the  result  of  convergence  due  to  the  require- 
ments of  bottom  feeding  behavior  (Gosline,  1968;  Fraser  1972b; 
Marshall  and  Cohen,  1973).  Similarities  to  the  perciform  fishes 
in  osteology  (Gosline,  1968;  Regan,  1912b)  and  biochemistry 
(Shaklee  and  Whitt,  1981)  point  to  a  perciform  origin  of  the 
group. 

The  structure  and  the  development  of  the  ophidiiform  caudal 
skeleton  support  the  hypothesis  of  a  closer  relationship  to  the 


gadiform  fishes  than  to  the  perciform  fishes.  In  Brotula.  as  in 
gadiforms,  two  separate  ural  centra  support  hypurals.  In  the 
Ophidiinae,  a  single  urostyle,  which  develops  from  a  single  car- 
tilaginous structure  in  the  larvae,  supports  two  hypurals.  This 
urostyle  is  apparently  homologous  to  the  two  ural  centra  of 
Brotula.  A  vestigial  neural  arch  develops  on  the  urostyle,  as  on 
the  first  ural  centrum  oi Brotula.  Also,  the  last  neural  and  haemal 
spines  in  both  Brotula  and  the  Ophidiinae  are  modified.  These 
spines  support  caudal  rays  in  Brotula  and  share  in  the  support 
of  the  last  dorsal  and  anal  rays  in  the  Ophidiinae.  In  the  gadiform 
caudal  skeleton,  similarly  modified  spines  on  the  first  preural 
centrum  support  caudal  rays.  In  both  gadiform  fishes  (Markle, 
1982)  and  ophidiine  fishes  these  spines  remain  cartilaginous  on 
the  distal  articular  surface. 

The  ophidiine  caudal  skeleton  differs  from  perciform  skele- 
tons in  the  development  of  the  hypural  elements  and  last  two 
haemal  arches.  In  ophidiine  larvae,  only  two  cartilaginous  hy- 
pural elements  form,  whereas  five  or  more  hypural  elements  are 
typically  present  in  the  skeleton  of  larval  perciforms.  The  last 
two  haemal  arches  in  perciform  fishes  remain  autogenous;  these 
arches  fuse  to  the  corresponding  centra  in  the  Ophidiinae. 

Ophidiiform  larvae  share  other  developmental  features  with 
gadiform  larvae.  Larvae  of  both  orders  develop  coiled  guts  (ex- 
cept for  aphyonid  larvae)  and  larvae  of  Carapidae  and  Ma- 
crouridae  have  high  vertebral  numbers  resulting  in  elongate 
larvae  with  reduced  or  absent  caudal  fins.  Another  similarity 
apparent  in  the  orders  is  the  presence,  in  larvae  of  some  species, 
of  modified  anterior  dorsal  rays.  In  Ophidiiformes,  this  char- 
acter IS  present  in  larval  Carapidae.  In  Gadiformes,  somewhat 
similar  structures  appear  in  larvae  of  Bregmaceros,  Enchelyopus 
and  Muraenolepis  although  comparative  studies  of  the  gross 
and  micro-structure  of  these  larval  specializations  are  lacking 
(Govoni  et  al.,  1984). 

Cohen  and  Nielsen  (1978)  consider  ophidiiform  fishes  to  be 
too  poorly  known  to  resolve  questions  of  phylogeny.  Our  as- 
sessment based  on  larval  data  is  similar.  Further  comparative 
studies  focusing  on  the  developmental  osteology  of  such  struc- 
tures as  the  caudal  fin,  anterior  vertebral  column  and  pectoral 
girdle,  as  well  as  the  development  of  the  gut,  will  allow  mean- 
ingful interpretation  of  the  significance  of  these  structures  to 
phyletic  studies. 

(D.J.G.)  RosENSTiEL  School  of  Marine  and  Atmospheric 
Science,  University  of  Miami,  4600  Rickenbacker 
Causeway,  Miami,  Florida  33149;  (D.F.M.)  The 
Huntsman  Marine  Laboratory,  St.  Andrews,  New 
Brunswick  EOG  2X0,  Canada;  (J.E.O.)  Virginia  Institute 
OF  Marine  Science,  College  of  William  and  Mary, 
Gloucester  Point,  Virginia  23062. 


Lophiiformes:  Development  and  Relationships 

T.  W.  PlETSCH 


THE  order  Lophiiformes  is  an  assemblage  of  18  families,  63 
genera,  and  approximately  282  living  species  of  marine 
teleosts,  the  monophyletic  origin  of  which  seems  certain  based 
on  the  following  synapomorphic  features:  (1)  Spinous  dorsal  fin 
primitively  of  six  spines,  the  anteriormost  three  of  which  are 
cephalic  in  position  and  modified  to  serve  as  a  luring  apparatus 
[involving  numerous  associated  specializations,  e.g.,  a  medial 
depression  of  the  anterior  portion  of  the  cranium,  loss  of  the 
nasal  bones  (nasal  of  Rosen  and  Patterson,  1969  =  lateral  eth- 
moid) and  supraoccipital  lateral-line  commissure,  and  modifi- 
cations of  associated  musculature  and  innervation];  (2)  Epiotics 
separated  from  parietals  and  meeting  on  the  midline  posterior 
to  the  supraoccipital;  (3)  Gill  opening  restricted  to  a  small,  elon- 
gate tubelike  opening  situated  immediately  dorsal  to,  posterior 
to,  or  ventral  to  (rarely  partly  anterior  to)  pectoral-fin  base;  (4) 
Second  ural  centrum  fused  with  the  first  ural  and  first  preural 
centra  to  form  a  single  hypural  plate  (sometimes  deeply  notched 
posteriorly)  that  emanates  from  a  single,  complex  half-centrum 
(Rosen  and  Patterson,  1969:441,  text  figs.  4E,  60);  (5)  Pectoral 
radials  narrow  and  elongate,  the  ventral-most  radial  consider- 
ably expanded  distally;  and  (6)  Eggs  spawned  in  a  double,  scroll- 
shaped  mucous  sheath  (Rasquin,  1958). 

Within  the  order  there  are  currently  recognized  three  subor- 
ders: the  Lophioidei,  containing  a  single  family  and  25  species 
of  relatively  shallow- water,  dorso-ventrally  flattened  forms  (Ca- 
ruso, 1981,  1983;  Caruso  and  Bullis,  1976);  the  Antennarioidei, 
with  six  families  and  approximately  121  species,  nearly  all  lat- 
erally-compressed, shallow-water,  benthic  forms  (Bradbury, 
1967;  Pietsch,  1981,  1984;  Pietsch  and  Grobecker,  in  press); 
and  the  Ceratioidei,  containing  about  136,  typically  globose, 
meso-  and  bathypelagic  species  (Bertelsen,  1951;  see  also  Ber- 
telsen,  this  volume). 

Development 

Little  is  known  about  the  early  life  stages  of  lophiiform  fishes, 
unequal  information  being  available  for  only  the  Lophiidae, 
Antennariidae  and  most  ceratioid  families.  Eggs  are  well-known 
in  lophiids  (Fulton,  1898;  Bowman,  1920;  Bigelow  and  Welsh, 
1925)  and  antennariids(Mosher,  1954;  Rasquin,  1958)  but  un- 
known in  all  other  lophiiforms.  Larvae  are  adequately  described 
in  lophiids  (Bowman,  1920;  Martin  and  Drewry,  1978),  anten- 
nariids(Mosher,  1954;  Rasquin,  1958)  and  most  ceratioids  (Ber- 
telsen, 1951),  but  remain  undescribed  in  chaunacids  and  ogco- 
cephalids  despite  some  available  material. 

Probably  the  most  striking  characteristic  of  early  ontogeny  in 
lophiiforms  is  the  fact  that  eggs  are  spawned  embedded  in  a 
continuous,  ribbon-like  sheath  of  gelatinous  mucous,  often  re- 
ferred to  as  an  "egg-raft"  or  "veil"  (with  one  known  exception, 
see  Pietsch  and  Grobecker,  1980).  Within  this  mucous  veil  are 
found  thousands  of  roughly-hexagonal,  liquid-filled  chambers 
arranged  in  one  to  several  irregular  layers,  each  chamber  con- 
taining from  one  to  three  eggs  (see  Rasquin,  1958  for  further 
details  and  figures  of  the  structure  of  egg  rafts).  Development 
is  fairly  direct,  with  the  larvae  in  all  known  groups  gradually 
acquiring  adult  characters  over  a  size  range  of  approximately  5 


mm  total  length  (TL)  in  antennariids  to  65  or  70  mm  TL  in 
lophiids.  Specialized  ontogenetic  stages  are  absent  except  for  the 
peculiar  "scutatus"  prejuvenile  present  in  the  ontogeny  o{  An- 
lennarius  radiosus  (see  below). 

Lophiidae 
Of  the  25  species  and  four  genera  of  the  Lophiidae  (Caruso, 
198 1),  early  life  stages  have  been  described  for  only  three  species, 
all  of  the  genus  Lophius:  L.  americamts  (Martin  and  Drewry, 
1978,  and  numerous  references  cited  therein),  L.  piscatohus 
(Tuning,  1923)  and  L. /'(/a'e^a5M(Padoa,  1 956e).  Of  these,  early 
ontogeny  is  best  documented  in  L.  americanus.  a  spring  or 
summer  spawner,  whose  egg  rafts  measure  0. 1 5-1 .5  m  wide  and 
6-12  m  long.  Living  eggs  are  slightly  oval,  their  major  axis 
measuring  1.61-1.94  mm.  The  outer  shell  appears  smooth  and 
transparent,  the  yolk  homogeneous  and  amber  in  coloration. 
The  perivitelline  space  is  narrow.  A  single,  copper,  orange  or 
pinkish-colored  oil  globule  is  present,  having  a  diameter  of  ap- 
proximately 0.40-0.45  mm.  Yolk-sac  larvae  measure  2.5-4.9 
mm  TL.  The  larvae,  ranging  in  size  from  6.5  to  approximately 
10.5  mm  TL,  are  prominently  pigmented,  with  early-forming 
dorsal  rays  and  pectoral  and  pelvic  fins  (Fig.  164A).  Relative  to 
antennariid  larvae,  the  head  is  small,  somewhat  less  than  30% 
of  standard  length.  The  gut  is  unusually  short.  The  dorsal  and 
pelvic  rays  are  unusually  elongate.  The  soft-dorsal  and  anal  fins 
are  last  to  form.  The  pectoral  fin  is  typically  large  and  fan- 
shaped.  Fin  ray  counts  are  complete  by  approximately  12  mm 
TL.  Transformation  to  the  prejuvenile  stage  takes  place  at  a  size 
somewhat  greater  than  approximately  12  mm  TL;  the  juvenile 
stage  is  not  reached  until  at  least  65  mm  TL.  In  well  preserved 
specimens  of  some  species  (i.e.,  Lophiodes  spilurus;  SIO  59-324, 
65.5  mm  TL)  the  epidermal  layer  of  the  head  and  body  is  greatly 
distended  by  transparent,  gelatinous  connective  tissue,  giving 
the  larvae  an  inflated  or  balloon-like  appearance  (as  described 
for  ceratioid  larvae  by  Bertelsen,  1951:12;  see  also  Bertelsen, 
this  volume).  (Largely  taken  from  Martin  and  Drewry,  1978: 
359-366,  where  the  reader  will  find  a  full  series  of  figures  and 
more  detailed  description  of  early  ontogeny.) 

Although  the  significance  of  variation  in  larval  pigmentation 
in  lophiids  is  largely  unknown,  larvae  of  the  American  species, 
Lophius  americanus  Valenciennes,  are  more  easily  distin- 
guished from  those  of  the  European  L.  piscatorius  Linnaeus  than 
are  the  adults,  using  characteristic  differences  in  pigmentation 
(Taning,  1923;  Martin  and  Drewry,  1978).  Tuning  (1923),  after 
studying  early  developmental  stages,  considered  the  two  species 
to  be  distinct  at  a  time  when  many  authors  regarded  them  as 
synonyms  (Martin  and  Drewry,  1978). 

Meristic  characters  that  typify  early  life  stages  of  lophiids  are 
compared  with  those  of  other  lophiiforms  in  Table  88. 

Antennariidae 

The  family  Antennariidae  consists  of  41  species  distributed 
among  1 2  genera  (a  modification  of  Schultz,  1957;  Pietsch,  1981, 
1 984;  Pietsch  and  Grobecker,  in  press).  Of  these,  early  life  stages 


320 


PIETSCH:  LOPHIIFORMES 


321 


»i>rVf-^--^' 


Fig.  164.  Representative  larvae  of  lophiiform  fishes:  (A)  Lophius  americanus.  12  mm  TL,  taken  from  Martin  and  Drewry,  1978:364,  fig. 
191B;  (B)  Hislrio  hislrio.  5.3  mm  TL,  taken  from  Adams,  1960:64,  fig.  IB;  (C)  Chaunax  sp.,  9.8  mm  TL,  ZMUC  P922155.  Gulf  of  Mexico, 
22°06'N,  84°58'W;  (D)  Ogcocephalus  sp..  10.4  mm  TL,  SHL  D-66-12,  P-4,  Western  North  Atlantic.  34<'17'N,  76°23.5'W;  (E)  Caulophrvnejordani. 
9.5  mm  TL,  ZMUC  P92198.  taken  from  Bertelsen,  1951:35,  fig.  UB. 


have  been  described  in  Antennarius  stria! us  (Phrynelox  scaber 
and  P.  nuttingi  oi  Kas(\\i\n,  1958,  and  P.  sca/)er  of  Martin  and 
Drewry.  1978)  and  Histrio  histrio  (Martin  and  Drewry,  1978 
and  numerous  references  cited  therein),  with  only  brief  descrip- 
tions of  the  "scutatus"  prejuvenile  stage  of  A.  radiosus  (see 
below).  Of  these,  H.  histrio  is  by  far  the  best  known.  Spawning 
occurs  year-round  except  in  February  and  March.  Freshly 
spawned  egg  rafts  measure  approximately  25-50  mm  wide  and 
90  mm  long.  Eggs  are  initially  oval  in  shape  (their  major  axis 
measuring  0.62-0.65  mm),  but  become  spherical  at  the  time  of 
the  second  cleavage.  Ova  are  extremely  transparent  (Mosher, 
1954)  and  colorless,  without  oil  globules.  As  development  pro- 
ceeds, the  raft  unrolls,  expanding  to  a  length  of  90  cm  (Smith, 
1907).  The  membranes  remain  firm  until  about  the  6-1 1  myo- 
mere stage,  but  then  begin  to  deteriorate,  the  raft  softening  and 


expanding  to  about  three  times  its  original  dimensions,  and 
finally  beginning  to  sink.  Yolk-sac  larvae  measure  0.88-1 .7  mm 
TL.  The  larvae,  most  strikingly  characterized  by  their  large  head 
(greater  than  45%  standard  length),  range  in  size  from  approx- 
imately 1.6-7.2  mm  TL  (Fig.  164B).  Pigmentation  is  conspic- 
uous about  the  head  and  midgut.  The  base  of  the  pelvic  fin 
elongates  at  about  12  mm  TL,  at  which  time  the  pigmentation 
of  the  midgut  begins  to  fade.  The  sequence  of  fin  formation  is 
as  follows:  caudal,  anal,  soft-dorsal,  pelvic,  pectoral,  the  dorsal 
spines  being  the  last  to  form  at  approximately  13  mm  TL. 
Prejuveniles  range  in  size  from  approximately  7,3-20  mm  TL. 
(Taken  from  Martin  and  Drewry,  1978:372-384,  where  the 
reader  will  find  a  full  series  of  figures,  and  a  more  detailed 
description  of  the  early  development  of  H.  histrio,  as  well  as 
that  o{  A.  striatus.) 


322 


ONTOGENY  AND  SYSTEMATICS  OF  HSHES-AHLSTROM  SYMPOSIUM 


Table  88.    Meristic  Values  for  Major  Taxa  of  the  Lophiiformes. 


Character 

Lophiidae 

Antenna  riidae 

Chaunacidae 

Ogcocephalidae 

Ceralioidei 

Dorsal  fin 

II-III  +  O-III  +  8- 

-12 

III  +  10-16 

III  +  10-12' 

II  +  1-6= 

II  +  3-22= 

Anal  fin 

6-10 

6-10 

6-7 

3-4 

3-19 

Caudal  fin 

8 

9 

8 

9 

8-10 

Pectoral  fin 

13-28 

6-14 

11-14 

10-15 

12-30 

Pelvic  fin 

I  +  5 

I  +  5 

I  +  4 

1  +  5 

3-4' 

Branchiostegal 

rays 

2  +  4 

2  +  4 

2  +  4 

2  +  4 

1  +  4,  2  +  4 

Vertebrae 

19-31 

18-23 

19 

18-21 

18-24 

'  Second  and  third  spine  embedded  beneath  skin  of  head- 

'  Second  spine  reduced  to  a  tiny  remnant  and  embedded  beneath  skin  of  head. 

*  Present  only  in  larvae  of  the  ceratioid  family  Cauiophrynidae. 


Meristic  characters  that  typify  early  life  stages  of  antennariids 
are  compared  with  those  of  other  lophiiforms  in  Table  88. 

The  so-called  "scutatus"  prejuvenile  form,  originally  de- 
scribed as  a  new  genus  and  species,  Kanazawaichthys  scutatus, 
by  Schultz  (1957),  but  later  found  by  Hubbs  (1958)  to  be  the 
prejuvenile  of  Antennarius  radiosus,  remains  unique  (Fig.  165). 
The  primary  morphological  features  that  characterize  these  early 
life  stages  are  so  drastic  (a  pair  of  shield-like,  bony  extensions 
of  the  cranium  that  reach  posteriorly  beyond  the  level  of  the 
opercular  bones,  and  an  expansion  of  the  anterior  margin  of  the 
bones  of  the  suspensorium;  see  Schultz  1957:63,  plate  14,  fig. 
A,  and  Hubbs,  1958)  that  their  appearance  in  other  antennariids 
of  similar  sizes,  particularly  among  closely  related  species  (such 
as  A.  ocellatus  and  A.  avalonis),  is  to  be  expected.  Yet,  despite 
the  fact  that  numerous  other  species  are  represented  by  small 
specimens,  no  comparable  morphological  adaptations  have  been 
discovered. 

Chaunacidae 

The  family  Chaunacidae  contains  a  single  genus  and  as  many 
as  12  species  (J.  H.  Caruso,  pers.  comm.,  8  June  1983).  Aside 
from  the  fact  that  larvae  and  "young  specimens"  are  often  caught 
bathypelagically  (Mead  et  al.,  1 964),  nothing  has  been  published 
on  their  early  life  stages,  despite  some  available  material.  The 
ovaries  of  members  of  this  family  are  scrolled  like  those  of  other 
lophiiforms,  suggesting  the  production  of  epipelagic  egg  rafts, 
although  neither  eggs  nor  rafts  have  been  reported. 

The  material  of  Chaunax  available  to  me,  32  specimens  in 
25  lots  (all  part  of  the  DANA  collections  of  the  Zoological 
Museum,  University  of  Copenhagen),  measured  4.3-10.6  mm 
TL.  Even  the  smallest  of  these  appear  to  have  reached  a  pre- 
juvenile stage,  with  all  fin  rays  formed  (including  the  illicium), 
and  the  skin  everywhere  covered  with  close-set  dermal  spinules 
(Fig.  I64C).  Pigmentation  appears  to  be  absent.  In  well  pre- 
served specimens,  the  epidermal  layer  of  the  head  and  body  is 
greatly  distended  by  transparent,  gelatinous  connective  tissue. 
As  in  antennariids,  the  head  is  large,  considerably  greater  than 
50%  of  SL  in  all  specimens  examined. 

Meristic  characters  that  typify  the  early  life  stages  of  chaun- 
acids  are  summarized  and  compared  to  those  of  other  lophii- 
forms in  Table  88. 

Ogcocephalidae 

The  Ogcocephalidae  contains  nine  genera  and  approximately 
60  species  (Bradbury,  1967).  Like  chaunacids,  practically  noth- 
ing is  known  about  their  early  life  stages.  The  only  published 
information,  aside  from  a  report  of  the  capture  of  a  single,  17.5 


mm  TL  specimen  of  Ogcocephalus  sp.  by  Clark  et  al.  (1969),  is 
a  comment  by  Mead  et  al.  (1964)  that  larval  and  postlarval 
specimens  have  been  caught  epipelagically.  The  scrolled  ovaries 
of  members  of  this  family  may  indicate  that  egg  rafts  are  pro- 
duced, but  neither  eggs  nor  rafts  have  been  reported. 

The  larval  ogcocephalid  material  available  to  me  (29  speci- 
mens in  nine  lots,  kindly  provided  by  Michael  P.  Fahay  of  the 
Northeast  Fisheries  Center,  Sandy  Hook  Laboratory,  and  ten- 
tatively identified  as  Ogcocephalus  sp.)  measured  3.1-18.4  mm 
TL.  In  the  smallest  of  these,  all  fins  are  fully  developed,  except 
for  the  illicium;  a  tiny  rudiment  of  this  structure  is  just  barely 
visible  in  a  4.9  mm  TL  specimen,  but  relatively  conspicuous  in 
a  5.1  mm  TL  specimen  (the  transition  to  a  prejuvenile  stage  is 
thus  taken  to  occur  at  approximately  5.0  mm  TL).  By  a  length 
of  8.6  mm  TL  the  pectoral  fins  are  large  and  fan-like,  the  base 
of  the  pelvic  fin  has  become  elongate,  and  small,  scattered  me- 
lanophores  are  present  on  top  of  the  head,  nape  of  the  neck,  on 
the  cheek  just  behind  the  eye,  the  pectoral  fin  base  and  blade, 
and  on  the  caudal  peduncle  (Fig.  164D).  At  8.9  mm  TL  the 
pigmentation  is  well  developed,  and  the  paired  fins  are  dispro- 
portionately large.  At  12.4  mm  TL  dermal  spinules  are  begin- 
ning to  form  in  the  skin;  a  lateral,  longitudinal  cluster  of  dermal 
spinules,  which  will  later  form  the  ridge-like,  outermost  margin 
of  the  adult,  is  just  beginning  to  develop.  By  18.4  mm  TL  the 
skin  is  everywhere  covered  with  broad-based  dermal  spinules, 
and  the  lateral  ridge  is  well-developed.  At  all  stages  of  devel- 
opment, but  particularly  the  prejuvenile  stage,  the  skin  is  highly 
inflated,  giving  the  larvae  an  almost  spherical  shape.  At  all  stages 
the  head  is  disproportionately  large,  considerably  greater  than 
50%  of  standard  length. 

Meristic  characters  that  typify  the  early  life  stages  of  ogco- 
cephalids  are  summarized  and  compared  to  those  of  other  lo- 
phiiforms in  Table  88. 

Ceratioid  Families 

The  Ceratioidei  contains  1 1  families,  34  genera  and  approx- 
imately 1 36  species.  Isolated  eggs  of  ceratioids  are  unknown; 
ovarian  eggs,  described  in  only  a  few  species,  are  slightly  oval, 
the  major  axis  of  the  largest  of  these  measuring  0.50-0.75  mm. 
The  larval  stages  of  all  of  the  families  and  most  of  the  genera 
have  been  described  (Bertelsen,  1951).  Generally  speaking,  cer- 
atioid larvae  are  typically  small.  According  to  Bertelsen  (1951), 
the  smallest  known  larvae  measure  2.0-3.0  mm  TL,  whereas 
the  largest  larvae  and  smallest  metamorphosis  stage  range  from 
12-25  mm  TL  for  females,  and  10-22  mm  TL  for  males.  As  in 
antennariid,  chaunacid  and  ogcocephalid  larvae,  the  head  is 
disproportionately  large;  as  in  some  lophiids,  chaunacids  and 
ogcocephalids,  the  head  and  body  are  enveloped  by  transparent. 


PIETSCH:  LOPHIIFORMES 


323 


Fig.  165.     "Scutatus"  prejuvenile  oi  Amennarius  radiosus,  21.2  mm  XL,  USNM  251937-F21.  North  Atlantic,  36°30'N,  74°30'W;  drawn  by 
B.  Washington. 


highly-inflated  skin  (Fig.  164E;  for  details  see  Bertelsen,  1951; 
Bertelsen,  this  volume). 

Meristic  characters  that  typify  the  early  life  stages  of  ceratioids 
are  summarized  for  all  eleven  families  and  compared  to  those 
of  other  lophiiforms  in  Table  88. 

Relationships 

Since  Regan  (1912a),  three  major  lophiiform  taxa  of  equal 
rank  have  been  recognized  by  nearly  all  authors.  These  taxa, 
together  with  their  currently  recognized  families  (the  eleven 
families  of  the  bathypelagic  Ceratioidei  excluded;  see  Bertelsen, 
this  volume),  are:  Suborder  Lophioidei  — Family  Lophiidae; 
Suborder  Antennarioidei  — Families  Antennariidae,  Tetrabra- 
chiidae,  Lophichthyidae,  Brachionichthyidae,  Chaunacidae,  and 
Ogcocephalidae;  Suborder  Ceratioidei.  , 

On  the  basis  of  adult  characters  alone,  Pietsch  (1981:416,  fig. 
41)  attempted  to  test  the  validity  of  Regan's  (1912a)  three 
major  lophiiform  taxa  using  cladistic  methodology,  but  ran  into 
serious  difficulty  in  attempting  to  establish  monophyly  for  the 
Antennarioidei;  while  a  number  of  synapomorphic  features  were 
found  to  support  a  sister-group  relationship  between  the  four 
families  Antennariidae  through  Brachionichthyidae,  and  be- 
tween the  families  Chaunacidae  and  Ogcocephalidae,  no  con- 
vincing synapomorphy  was  found  to  link  these  two  larger 
subgroups. 

To  date,  early  life  history  stages  have  not  been  used  in  for- 
mulating hypotheses  of  relationship  among  lophiiform  fishes; 
but,  several  egg  and  larval  characters  are  shown  here  to  be 
significant  in  resolving  a  number  of  problems  with  this  group. 
These  characters,  along  with  several  previously  unidentified  adult 
characters,  have  been  used  here  to  construct  a  revised  cladogram 
of  lophiiform  relationships  (Fig.  166). 

This  new  cladogram  differs  significantly  from  that  previously 
published  (Pietsch,  1981:  fig.  4 1 ).  The  suborder  Antennarioidei 


is  now  restricted  to  only  four  families:  The  Antennariidae,  rec- 
ognized as  the  primitive  sister-group  of  the  Tetrabrachiidae, 
these  two  families  together  forming  the  primitive  sister-group 
of  the  Lophichthyidae,  and  this  assemblage  of  three  families 
forming  the  primitive  sister-group  of  the  Brachionichthyidae. 
These  relationships  are  supported  by  a  total  of  seven  synapo- 
morphies  (most  of  which  were  previously  described  by  Pietsch, 
1981:413-414)  numbered  7  through  13  in  Fig.  166:  (7)  Pos- 
teromedial process  of  vomer  emerging  from  ventral  surface  as 
a  laterally-compressed,  keel-like  structure,  its  ventral  margin  (as 
seen  in  lateral  view)  strongly  convex  (Pietsch,  1981:397,  figs. 
4-6);  (8)  Postmaxillary  process  of  premaxilla  spatulate  (Pietsch, 
1981:398,  figs.  8,  20);  (9)  Opercle  similarly  reduced  in  size 
(Pietsch,  1981:401,  figs.  9,  21);  (10)  Ectoplerygoid  triradiate,  a 
dorsal  process  overlapping  the  medial  surface  of  the  metapter- 
ygoid  (Pietsch.  1981:400,  figs.  9,  21,  22);  (11)  Proximal  end  of 
hypobranchials  II  and  III  bifurcated  (Pietsch,  1981:407,  figs.  11, 
28.  29);  (12)  Interhyal  with  a  medial,  posterolaterally  directed 
process  that  comes  into  contact  with  the  respective  preopercle 
(Pietsch,  1981:400,  fig.  26);  and  (13)  Illicial  pterygiophore  and 
pterygiophore  of  the  third  dorsal  spine  with  highly  compressed, 
blade-like  dorsal  expansions  (Pietsch,  1981:410,  figs.  36,  37). 

The  present  interpretation  of  lophiiform  relationships  differs 
further  from  any  previously  proposed  hypothesis  in  considering 
the  Antennarioidei  (sensu  stricto)  to  form  the  primitive  sister- 
group  of  a  much  larger  group  that  includes  the  Chaunacioidei 
(new  suborder),  the  Ogcocephalioidei  (new  suborder)  and  the 
Ceratioidei.  The  Ogcocephalioidei  is,  in  turn,  recognized  as  the 
primitive  sister-group  of  the  Ceratioidei  (Fig.  166). 

Monophyly  for  a  group  containing  the  suborders  Antenna- 
rioidei, Chaunacioidei.  Ogcocephalioidei  and  Ceratioidei  is  sup- 
ported by  four,  previously  unidentified  synapomorphies  (num- 
bered as  they  appear  in  Fig.  1 66):  ( 1 4)  Eggs  and  larvae  small  (at 
all  stages  eggs  are  considerably  less  than  50%  the  diameter  of 


324 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  166.     Cladogram  showing  proposed  phylogenetic  relationships  of  the  major  subgroups  of  the  Lophiiformes.  Black  bars  and  numbers  refer 
to  synapomorphic  features  discussed  in  the  text. 


those  of  lophioids;  smallest  larvae  are  certainly  less  than  50%, 
and  probably  less  than  30%  the  size  of  those  of  lophioids;  size 
at  transformation  to  the  prejuvenile  stage  is  less  than  60%  that 
of  lophioids);  (15)  Head  of  larvae  proportionately  large  rel- 
ative to  body  (always  greater  than  45%  of  standard  length,  com- 
pared to  less  than  30%  in  lophioids);  (16)  Reduction  in  the 
number  of  dorsal  fin  spines  from  a  primitive  number  of  six  in 
lophioids  to  three  or  less  (Pietsch,  1981:409,  figs.  36-38);  and 
(17)  Loss  of  pharyngobranchial  IV  (present  and  well-toothed  in 
lophioids;  Pietsch,  1981:401,  figs.  1 1,  28-32). 

Monophyly  for  a  group  containing  the  suborders  Chauna- 
cioidei,  Ogcocephalioidei  and  Ceratioidei  is  supported  by  two 
synapomorphies:  (18)  Second  dorsal  spine  reduced  and  embed- 
ded beneath  skin  of  head  (Pietsch,  1981:410,  figs.  36-38);  and 
(19)  Gill  filaments  of  gill  arch  I  absent  (but  present  on  proximal 
end  of  ceratobranchial  I  of  some  ceratioids;  Bradbury,  1967: 
408;  Pietsch,  1981:415). 

That  the  Ogcocephalioidei  is  the  primitive  sister-group  of  all 
ceratioid  families  is  supported  by  three  synapomorphies:  (20) 
Second  dorsal  spine  reduced  to  a  small  remnant  (well  developed 
in  the  ceratioid  family  Diceratiidae.  and  in  all  other  lophiiforms; 
Bertelsen,  1951:17;  Pietsch,  1 98 1 :4 1 0,  fig.  38);  (2 1 )  Third  dorsal 
spine  and  pterygiophore  absent  (present  in  all  other  lophiiforms; 
Bertelsen,  1951:17;  Bradbury,  1967:401;  Pietsch,  1981:410,  fig. 
38);  and  (22)  Epibranchial  I  simple,  without  ligamentous  con- 
nection to  epibranchial  II  (in  batrachoidiforms  and  all  other 
lophiiforms  epibranchial  I  bears  a  medial  process  that  is  liga- 


mentously  attached  to  the  proximal  tip  of  epibranchial  II;  Pietsch, 
1981:401,  figs.  28-32). 

Of  the  possible  cladograms  that  could  be  constructed  on  the 
basis  of  the  data  provided  in  this  study,  the  one  shown  in  Fig. 
166  is  by  far  the  most  parsimonious.  But  at  the  same  time, 
acceptance  of  this  revised  hypothesis  of  relationships  of  lo- 
phiiform  fishes  requires  evolutionary  convergence  or  reversal 
in  three  derived  character  states  previously  used  by  me  (Pietsch, 
1981:415,  fig.  41)  to  support  a  hypothesis  of  sister-group  rela- 
tionship between  the  Chaunacidae  and  Ogcocephalidae:  ( 1 )  Pos- 
teriormost  branchiostegal  ray  exceptionally  large  (all  four  pos- 
teriormost  branchiostegal  rays  approximately  equal  in  size  in 
batrachoidiforms  and  all  other  lophiiforms;  Pietsch,  1981,  fig. 
27);  (2)  Gill  teeth  tiny,  arranged  in  a  tight  cluster  at  apex  of 
pedicel-like  tooth  plates  (in  all  other  lophiiforms  gill  teeth,  if 
present,  are  relatively  large,  and  either  single,  or  associated  with 
a  flat,  rounded  tooth  plate;  but  tiny,  and  at  apex  of  elongate 
pedicel-like  tooth  plates  in  at  least  some  batrachoidiforms,  e.g., 
Poriclithys;  Pietsch,  1981,  figs.  31,32)  and  (3)  Illicial  bone,  when 
retracted,  lying  within  an  illicial  cavity  (an  illicial  cavity  is  absent 
in  all  other  lophiiforms;  however,  the  illicium  and  esca  lie  within 
a  shallow  groove  on  the  dorsal  midline,  sometimes  enveloped 
by  folds  of  skin,  in  the  antennariid  genus  Histiophryne,  Pietsch, 
1981,  fig.  39;  Pietsch,  1984:40). 

The  cladistic  relationships  of  the  Lophiiformes  are  summa- 
rized in  the  following  revised  classification.  While  the  ranking 
of  taxa  is  not  dichotomous  (see  methods  in  Pietsch,  1981:388), 


PIETSCH:  LOPHIIFORMES 


325 


inter-nested  sets  of  vertical  lines  are  used  to  indicate  mono- 
phyletic  units. 

Order  Lophiiformes 

Suborder  Lophioidei 

Family  Lophiidae 
Suborder  Antennarioidei 

Family  Antennariidae 

Family  Tetrabrachiidae 
.  Family  Lophichthyidae 

Family  Brachionichthyidae 
Suborder  Chaunacioidei  New 

Family  Chaunacidae 
Suborder  Ogcocephalioidei  New 

Family  Ogcocephalidae 
Suborder  Ceratioidei 


As  a  final  note,  the  Lophiiformes  has  traditionally  been  allied 
with  the  Batrachoidiformes,  based  primarily  on  osteological 
characters  ofthe  cranium  (Regan,  1 9 1 2a;  Gregory,  1933;  Rosen 
and  Patterson,  1969).  However,  this  sister-group  relationship 
has  yet  to  be  shown  conclusively,  and  1  have  not  been  able  to 
assess  the  significance  of  early  life  stages  in  supporting  or  refuting 
this  hypothesis. 

School  of  Fisheries  WH-  1 0,  College  of  Ocean  and  Fisheries 
Sciences,  University  of  Washington,  Seattle, 
Washington  98195. 


Ceratioidei:  Development  and  Relationships 
E.  Bertelsen 


THE  Ceratioidei  differ  most  distinctly  from  all  other  mem- 
bers ofthe  order  Lophiiformes  in  being  meso-  and  bathy- 
pelagic,  lacking  pelvic  fins  (except  in  larval  Caulophrynidae) 
and  in  having  extreme  sexual  dimorphism.  Males  are  dwarfed 
and  differ  from  females  in  lacking  an  external  illicium  and  hav- 
ing denticular  teeth  on  the  tips  of  the  jaws  and  well-developed 
eyes  and/or  olfactory  organs.  Furthermore,  Ceratioidei  differ 
from  other  Lophiiformes,  except  the  family  Ogcocephalidae,  in 
lacking  a  third  cephalic  ray  and  its  pterygiophore,  and  except 
in  the  families  Caulophrynidae,  Neoceratiidae  and  the  gigan- 
tactinid  genus  Rhynchactis.  females  ofthe  suborder  differ  from 
all  other  Lophiiformes  in  having  a  bulbous  swelling  of  the  tip 
of  the  illicium  (escal  bulb)  containing  a  large  globular  photo- 
phore. 

The  suborder  contains  approximately  1 34  species  placed  in 
34  genera  and  1 1  families  (Table  89).  The  taxonomy  is  based 
mainly  on  studies  of  the  females.  Except  for  the  larval  stages 
and  the  basic  meristic  and  osteological  characters  shared  by  the 
two  sexes,  descriptions  require  separate  treatment  of  females 
and  males. 

The  separation  into  families  is  based  mainly  on  osteological 
characters,  of  which  some  ofthe  more  important  are  compared 
in  Table  89.  Most  ofthe  families  form  well-defined  and  mutually 
very  distinct  taxa  in  which  the  females  (especially)  possess  unique 
morphological  features  which  separate  them  from  members  of 
all  other  families. 

The  separation  into  genera  is  based  mainly  on  characters 
present  only  in  females.  Somewhat  varying  between  families, 
some  ofthe  most  important  of  these  characters  are  differences 
in:  (1)  shape  of  skull  and  other  bones  of  the  head  including 
development  of  its  spines;  (2)  jaw  mechanisms,  including  den- 
tition; (3)  illicial  apparatus,  including  basic  patterns  of  escal 
appendages;  and  (4)  pigmentation  of  skin  and  development  of 
dermal  spines.  Some  of  the  distinguishing  osteological  charac- 
ters, especially  in  shape  of  opercular  bones,  are  shared  with  the 


males,  like  the  fin  ray  numbers  which  in  some  families  show 
distinct  intergeneric  differences.  The  special  male  structures, 
such  as  denticular  teeth,  show  distinct  intrageneric  differences 
in  full  agreement  with  the  separations  based  on  characters  of 
the  females. 

The  species  of  Linophryne  have  been  grouped  into  subgenera 
and  those  of  Himantolophus  (in  ms.),  Oneirodes.  and  Gigan- 
tactis  into  "species-groups'"  based  on  shared  minor  differences 
in  one  or  more  of  the  characters  mentioned  above.  All  intra- 
generic separations  of  the  females  into  species  are  based  on 
differences  in  pattern  and  shape  of  escal  appendages,  often  com- 
bined with  differences  in  illicial  lengths.  In  a  majority  of  the 
recognized  species,  no  other  separating  characters  have  been 
shown.  In  others,  differences  in  meristic  characters  (numbers  of 
fin  rays  and  teeth)  and  minor  osteological  characters  (shape  of 
opercular  bones,  development  and  dentition  of  branchial  arches, 
etc.)  have  been  observed  supporting  the  separation  into  species. 
A  special  opportunity  to  check  the  validity  of  the  separations 
based  on  escal  characters  is  found  in  the  genus  Linophryne.  in 
which  females  have  hyoid  barbels  which  in  pattern  of  branching 
show  very  distinct  differences  between  species  and  subgenera 
(Bertelsen,  1982). 

In  most  cases  it  has  not  been  possible  to  separate  males  into 
taxa  below  the  level  of  genera  and  subgenera.  A  few  males 
differing  from  their  supposed  congeners  in  special  male  char- 
acters (especially  denticular  teeth)  have  been  tentatively  de- 
scribed as  representatives  of  separate  species.  Studies  of  males 
attached  to  identified  females  have  not  revealed  characters  of 
specific  order. 

Development 

No  spawned,  fertilized  eggs  of  Ceratioidei  have  been  described 
(re-examination  by  Bertelsen,  1980:66,  of  an  egg  referred  to 
Linophryne  arborifera  by  Beebe,  1932:93,  indicated  that  it  rep- 
resents a  diodontid). 


326 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  89.    Characteristics  of  Ceratioidei. 


X:  presumed  denved  characters 

O:  presumed  pnmilivc  characters 

Caul- 

Nco- 

Thauma- 

Centro- 

Lin- 

ophryni- 

cerati- 

Melano- 

Himantol- 

Dicerali- 

Oneir- 

hchthyi- 

phryni- 

Cerali- 

Gigantac- 

ophryn- 

Families  (11): 

dae 

idae 

cetidae 

ophidae 

idae 

odidae 

dae 

dae 

idae 

tinidae 

idae 

No.  of 

genera  (34): 

2 

1 

I 

1 

2 

15 

2 

1 

2 

2 

5 

No.  of  species  (ca.  134): 

4 

1 

4 

ca.  15 

4 

52 

6 

1 

3 

19 

25 

No. 

Characters  shared  by  sexes: 

1 

Dorsal  finrays 

6-22 

11-13 

12-17 

5-6 

5-7 

4-8 

5-7 

6-7 

4 

3-9 

3 

2 

Anal  finrays 

5-19 

10-13 

4 

4 

4 

4-7 

4-5 

5-6 

4 

3-7 

3 

3 

Caudal  finrays 

8 

9-10 

9 

9 

9 

9 

9 

9 

8-8  Vj 

8'/2 

8 1/2 

4 

Branchiostegal  rays 

5-6 

5-6 

6 

6 

6 

6 

6 

6 

6 

6 

5 

5 

Pectoral  radials 

2 

3 

4 

3 

3 

3 

3 

4 

4 

5 

3 

6 

Shape  of  pelvic  bones 

XI 

XI 

O 

O 

O 

O-Xl 

O 

XI 

XI 

Xl-2 

X2 

7 

Head  of  hyomandibular 

o 

X 

O 

O 

O 

O-X 

O 

O 

O 

X 

X 

8 

Parietals 

o 

O 

o 

X 

o 

O 

O 

O 

o 

o-x 

O 

9 

Pterosphenoid 

X 

O 

o 

o 

o 

O-X 

o 

O 

X 

X 

X 

10 

Epural 
Female  characters: 

o 

X 

X 

X 

X 

X 

X 

X 

X 

X 

O-X 

11 

Escal  photophore 

o 

O 

X 

X 

X 

X 

X 

X 

X 

O-X 

X 

12 

Photophore  on  2'ccphalic  ray 

o 

O 

o 

o 

X 

O 

o 

O 

X 

o 

O 

13 

Separation  of  frontals 

o 

O 

X 

X 

X 

X 

X 

X 

o 

X 

O-X 

14 

Shape  of  frontals 

o 

O 

o 

o 

o 

X 

X 

X 

X 

XI 

X 

15 

Maxillaries 

XI 

XI 

o 

o 

o 

o 

o 

O 

o 

X2 

Xl-2 

16 

Maxillomandibular  ligament 

X 

X 

o 

o 

o 

o 

X 

o 

o 

X 

X 

17 

Branchial  teeth 

X 

X 

X 

o 

X 

o-x 

X 

o 

X 

X 

O-X 

18 

Quadrate  and  Articular  spine 

o 

O 

o 

o 

o 

O-X 

X 

o 

o 

o 

o 

19 

Dermal  spines 

X 

X 

o 

o 

o 

O-X 

O-X 

o 

o 

o 

X 

20 

Shape  of  body 
Male  characters: 

o 

X 

o 

o 

o 

o 

X 

X 

X 

X 

o 

21 

Eyes 

o 

XI 

o 

o 

o 

o 

o 

XI 

X3 

XI 

X2 

22 

Olfactory  organs 

X 

o 

X 

X 

X 

X 

X 

X 

o 

X 

X 

23 

Anterior  nostrils 

X 

o 

o 

o 

o 

X 

X 

X 

o 

X 

X 

24 

Upper  denticular  teeth 

o 

o 

X 

X 

X 

X 

X 

X 

X 

X 

X 

25 

Dermal  spines 

X 

X 

o 

o 

o 

X 

o 

X 

o 

O-X 

X 

26 

Parasitic  males  observed 
Larval  characters: 

X 

X 

o 

o 

o 

X 

o 

o 

X 

o 

X 

27 

lllicial  rudiment 

o 

o 

X 

X 

X 

X 

X 

X 

X 

X 

X 

28 

Pelvic  fins 

o 

X 

X 

X 

X 

X 

X 

X 

X 

X 

X 

29 

Pectoral  fins 

o 

X 

X 

X 

X 

X 

X 

X 

X 

o 

X 

30 

Shape  of  body 

o 

X2 

o 

o 

o 

O-Xl 

o 

o 

X3 

o 

XI 

(1-2)  High  numbers  of  dorsal  and  anal  fin  rays  possibly  a  pnmitive  character  state;  (3)  Nine  rays  presumed  pnmitive;  8'/::  9th  ray  reduced  to  less  than  half  length  of  the  8th.  (4)  Six  rays 
presumed  pnmitive;  (5)  Three  radials.  shared  with  antennanoids,  here  presumed  to  be  pnmitive;  however,  the  trend  to  reduce  the  number  from  4  to  3  in  older  specimens  of  centrophrymds  and 
melanocetids  might  indicate  that  within  ceratioids  four  radials  are  pnmitive  and  three  a  result  of  secondary  reduction;  (6)  Distally  expanded  pelvic  bones  (triradiate  in  Himanfolophus  and  some 
specimens  of  the  oneirodid  genus  Chaenophryne).  presumed  pnmitive;  XI:  a  simple  rod;  X2:  absent;  (7)  Double  head  of  hyomandibular  presumed  pnmitive;  X:  single  head  (in  oneirodids  in 
only  one  of  the  15  genera);  (8)  Panetals  absent  in  all  himantolophids.  lost  in  metamorphosed  females  of  the  gigantactinid  genus  Rhynchactis.  (9)  Presence  of  pterosphenoid  presumed  pnmitive 
(absent  in  one  of  the  oneirodid  genera);  (10)  A  single  epural  observed  in  Caulophn'iie  and  in  the  linophrynid  genus  Phnlocon'nus.  (11)  The  absence  of  escal  photophore  presumed  pnmitive  in 
Caulophrynidae.  and  possibly  in  Neoceraltas  while  the  absence  in  the  gigantactinid  genus  Rhynchaclis  is  presumed  to  be  due  to  a  secondary  specialization  (cf  text);  (12)  While  the  presence  of  a 
photophore  on  2nd  cephalic  ray  is  a  denved  character  stale,  the  presence  of  this  ray  may  be  regarded  as  primitive;  (13)  Frontals  meeting  in  the  midline  presumed  pnmitive  (in  linophrynids 
present  in  Pholocorynus);  (26)  Parasitic  males  observed  in  Neoceraltas,  both  genera  of  Ceratiidae,  one  of  the  two  genera  of  Caulophrynidae.  four  of  the  five  genera  of  Linophrynidae.  and  one  of 
the  15  genera  of  Oneirodidae;  (27)  Presence  of  an  external  illicial  rudiment  in  larval  males  presumed  primitive;  (28)  Presence  of  pelvic  fins  presumed  pnmitive,  (29)  Enlarged  pectoral  fins  here 
presumed  pnmitive;  and  (30)  Short,  more  or  less  sphencal  body  presumed  pnmitive;  XI:  moderately  elongate;  X2:  slender;  X3:  hump-backed. 


A  cluster  of  eggs  embedded  in  a  mucoid  substance  hanging 
out  of  the  greatly  expanded  genital  opening  of  a  sexually  par- 
asitized female  of  Ltnophryne  arborifera  was  observed  by  Ber- 
telsen  ( 1 980:66).  This  indicates  that  Ceratioidei  expel  their  eggs 
in  free-floating  mucoid  egg  "rafts"  or  "veils,"  as  described  in 
species  of  the  other  suborders  of  Lophiiformes  (Rasquin,  1 958). 
It  is  possible  that  the  release  of  the  egg  veil  of  the  specimen  was 
caused  by  the  catch,  and  that  the  eggs  were  not  completely 
mature.  They  were  slightly  oval,  0.6-0.8  mm  in  diameter,  with 
smooth,  very  soft  outer  membranes  that  were  folded  or  shrunk- 
en in  several.  The  yolk,  which  contained  numerous  small  oil 
globules,  was  opaque  and  partially  surrounded  by  an  irregular 
perivitelline  space. 

The  observed  number  of  ceratioid  females  with  apparently 
nearly  mature  ovaries  is  relatively  small.  In  these,  the  largest 
eggs  have  diameters  of  0.5-0.75  mm. 


Larvae  and/or  metamorphic  stages  representing  all  1 1  fami- 
lies and  26  of  the  34  recognized  genera  have  been  described, 
the  majority  by  Bertelsen  (1951).  Identification  to  species  is 
restricted,  however,  to  those  genera,  subgenera  or  "species- 
groups"  in  which  only  a  single  species  is  recognized. 

No  specialized  ontogenetic  stage  between  larvae  and  juveniles 
occurs.  In  most  genera,  metamorphosis  begins  at  a  size  of  8-10 
mm  SL,  while  in  some  (Himantolophidae,  Thaumatichthys,  Gi- 
gantactinidae,  and  Linophrynidae),  the  larvae  may  reach  lengths 
of  15-25  mm.  During  metamorphosis,  covering  a  size  range 
somewhat  varying  between  genera,  adult  characters  are  gradu- 
ally acquired.  In  both  sexes,  the  skin  is  gradually  covered  with 
pigment  (except  in  Haplophryne),  and  in  certain  genera,  skin 
spines  are  developed.  In  females  the  illicial  and  escal  characters 
develop,  the  head  and  especially  the  jaws  increase  in  relative 
size,  the  larval  teeth  are  replaced,  and  the  growth  of  eyes  and 


BERTELSEN:  CERATIOIDEI 


327 


olfactory  organs  is  retarded.  In  males  the  body  elongates,  larval 
teeth  are  lost,  the  denticular  teeth  develop,  and  eyes  and/or 
olfactory  organs  increase  in  relative  size. 

The  larvae  have  been  referred  to  genera,  subgenera  or  species- 
groups  on  basis  of  (1)  meristic  characters  (especially  number  of 
dorsal  and  anal  fin  rays);  (2)  osteological  characters  (especially 
number  of  branchiostegal  rays  and  pectoral  radials  and  shape 
of  head  of  hymandibular,  pelvic  bones  and  opercular  bones); 
and  (3)  pattern  of  subdermal  pigmentation.  The  pattern  is  re- 
tained under  the  pigmented  skin  of  post-metamorphic  juveniles 
which  have  acquired  adult  characters  (Bertelsen,  1951).  In  most 
genera  the  smallest  larvae  observed  are  2.5-3.5  mm.  At  these 
stages,  in  which  no  distinguishing  characters  other  than  pig- 
mentation may  be  developed,  identification  is  based  on  com- 
parison with  developmental  series  of  older  larvae. 

Meristic  characters.— The  2-3  mm  smallest  known  larvae  have 
an  almost  straight  notochord  and  almost  undifferentiated  fins. 
The  fin  rays  of  the  unpaired  fins  are  laid  down  early  and  the 
full  number  is  usually  present  in  larvae  of  3-4  mm  SL  of  the 
numerous  species  where  the  number  of  dorsal  rays  does  not 
exceed  8.  The  pectoral  fin  rays  are  laid  down  somewhat  later 
than  those  of  the  unpaired  fins,  and  the  lowermost  rays  are  rarely 
discernible  in  specimens  of  less  than  about  5-6  mm.  Caulo- 
phrynidae  and  the  ceratiid  genus  Cryptopsaras  have  8  caudal 
fin  rays,  all  others  have  9  ( 10  in  some  specimens  of  Neoceratias). 
The  9th  (lowermost)  ray  is  rudimentary  or  short  (less  than  half 
the  length  of  the  8th  ray)  in  Linophrynidae,  Gigantactinidae, 
and  Ceratias. 

Except  in  the  three  genera  in  which  the  number  of  rays  in  the 
anal  and/or  dorsal  fin  exceeds  10  (Caulophryne.  Neoceratias. 
and  Melanocelus,  cf  Table  89),  the  intraspecific  variation  of  the 
number  of  fin  rays  in  these  fins  is  small,  rarely  more  than  ±  1 . 
Significant  differences  in  numbers  of  dorsal  and  anal  fin  rays 
have  been  found  between  species  within  the  genera  Caulo- 
phryne. Gigantactis,  and  Melanocetus  and  between  genera  in 
the  families  Caulophrynidae,  Gigantactinidae,  and  Oneirodi- 
dae. 

Pectoral  fin  rays  number  12-23  in  all  ceratioids  (except  Cten- 
ochirichthys  with  28-30).  As  an  intraspecific  range  of  variation 
of  5  to  7  fin  rays  has  been  observed,  this  character  may  aid 
identification  only  in  exceptional  cases. 

All  reported  vertebral  counts  of  Ceratioidei  fall  within  the 
range  of  19  to  24,  the  highest  number  in  Neoceratiidae,  the 
lowest  in  Linophrynidae.  The  limited  number  of  observations 
does  not  permit  an  evaluation  of  the  diagnostic  value  of  differ- 
ences within  this  range. 

Morphology.— The  head  and  body  of  larval  Ceratioidei  are  sur- 
rounded by  inflated  transparent  skin.  Due  to  this  balloon-like 
envelope,  their  shape  varies  from  nearly  spherical,  with  greatest 
width  and  depth  of  body  reaching  80-90%  SL,  to  elongated  or 
pear-shaped,  with  body  depth  of  40-60%  SL. 

The  inflation  of  the  skin  varies  with  preservation,  but  gen- 
erally its  greatest  development  is  found  in  Caulophrynidae  (Fig. 
167 A),  Gigantactinidae  (Fig.  168A,  B),  and  Himantolophidae 
(Fig.  169 A,  B);  less  pronounced  in  Neoceratiidae  (Fig.  167B), 
Ceratiidae  (Fig.  168C-E)  and  Oneirodidae  (Fig.  170).  No  dis- 
tinct change  has  been  observed  in  the  relative  development  of 
the  inflation  during  larval  life.  In  larvae  of  most  genera,  the 
relative  length  of  head,  measured  to  base  of  pectoral  peduncle, 
is  about  50%  SL.  In  Oneirodidae  and  Linophrynidae  it  is  some- 
what less  (generally  about  45%)  and  is  shortest  in  Neoceratias 


(35-40%).  In  the  late  larval  stages  of  males,  the  elongation  of 
the  body  may  start  before  other  metamorphic  characters  have 
appeared. 

In  larval  Ceratiidae,  the  vertebral  column  is  more  strongly  S- 
shaped  than  in  other  families,  resulting  in  a  characteristic  hump- 
backed appearance  of  these  larvae  (Fig.  168C-E).  The  larvae  of 
Caulophrynidae  (Fig.  167 A)  and  Gigantactinidae  (Fig.  168A,  B) 
differ  distinctly  from  those  of  other  Ceratioidei  in  the  size  of 
the  pectoral  fins,  which  have  lengths  of  40%  to  nearly  60%  SL, 
measured  from  base  of  pectoral  peduncle.  In  the  other  families, 
this  length  is  generally  20-25%  SL  and  does  not  exceed  30% 
SL. 

Pigment.— The  subdermal  pigment  occurring  in  larval  Cera- 
tioidei is  usually  separated  into  four  more  or  less  well-defined 
main  groups:  (I)  peritoneal;  (2)  opercular;  (3)  dorsal;  and  (4) 
caudal-peduncular. 

In  Neoceratias  (Fig.  167B)  and  some  linophrynids  (Haplo- 
phryne.  Fig.  167D),  the  subgenera  of  Linophryne:  Stephano- 
phryne  {Fig.  167F).  and  Rhizophryne (Fig.  167C),  the  subdermal 
pigment  forms  dorso-  and  ventrolateral  bands  along  the  body. 

In  all  species  in  which  one  of  these  main  groups  occur,  they 
are  generally  laid  down  in  the  youngest  larvae  as  a  few  small 
and  scattered  melanophores  which  during  larval  development 
gradually  increase  in  size,  number,  and  in  area  covered. 

Additional  groups  of  melanophores  occur  in  some  taxa  (for 
instance  on  base  of  pectoral  peduncle  in  Rhynchactis,  Fig.  I68B; 
internally  in  fin  rays  of  Pentherichthys,  Fig.  1 70E;  on  the  pos- 
terior angle  of  lower  jaw  in  Stephanophryne,  Fig.  I67F;  and  on 
a  swelling  of  the  outer  transparent  skin  in  front  of  the  dorsal  fin 
in  some  Himantolophus.  Fig.  I69B). 

Complete  lack  of  pigment  is  found  in  Ceratias  (Fig.  168E), 
some  Gigantactis,  and  the  linophrynid  genus  Borophryne  (Fig. 
167G).  Besides  in  these,  peritoneal  pigment  is  absent  only  in 
Neoceratias.  In  all  others,  this  group  is  laid  down  on  the  dorsal 
side  of  the  peritoneum  of  the  youngest  larvae  and  with  growth, 
gradually  spreads  to  its  lateral  and  posterior  sides. 

Pigmentation  of  the  opercular  region  varies  greatly  between 
taxa.  It  is  absent  or  weakly  developed  in  most  genera,  dense 
and  in  characteristically  different  patterns  in  genera  of  Onei- 
rodidae (for  instance  Oneirodes,  Fig.  1 70A;  Dolopichthys.  Fig. 
170B;  Microlophichthys.  Fig.  170F;  Thaumatichthys,  Fig.  169F; 
and  Cryptopsaras,  Fig.  168C,  D). 

Besides  in  the  completely  unpigmented  larvae  mentioned 
above,  dorsal  pigment  is  absent  in  Caulophryne,  Neoceratias, 
and  all  Linophrynidae.  In  all  others  it  is  laid  down  on  the  antero- 
dorsal  part  of  the  body.  Varying  between  genera  in  density  and 
coverage,  it  spreads  from  there  and  may  laterally  reach  and 
overlap  the  dorsal  part  of  the  peritoneal  pigmentation  and  pos- 
teriorly the  bases  of  the  dorsal  and  anal  fins,  in  some  becoming 
confluent  with  the  pigment  group  of  the  caudal  peduncle. 

In  occurrence,  position  and  development  in  relation  to  larval 
length,  pigmentation  on  the  caudal  peduncle  shows  very  distinct 
differences  between  genera  (cf  for  instance  Fig.  1 70)  or  between 
subgenera  or  species-groups  (cf  for  instance  Figs.  167E,  F;  169 A, 
B;  170C,  D). 

Other  larval  structures.  — Fe\\\c  fins.  — In  contrast  to  all  other 
Ceratioidei,  the  larvae  oi  Caulophryne  have  pelvic  fins  with  well- 
developed  fin  rays  (Fig.  167A).  The  longest  of  the  3-4  pelvic 
fin  rays  increase  in  relative  length  from  about  45%  SL  in  the 
smallest  larvae  (3  mm)  to  about  60%  SL  in  the  largest  (7.5  mm). 
In  the  only  known  free-living  stage  of  a  metamorphosed  male 


328 


ONTOGENfY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


t 1 I L_ 


Fig.  167.  Ceratioid  larvae.  (A)  Caulophrynidae,  Caulophryme  sp.,  sex  ?,  6.6  mm;  (B)  Neoceratiidae.  Neoceratias  spinifer.  sex  ?,  6.3  mm;  (C- 
G)  Linophrynidae;  (C)  Linophnme  subgen.  Rhizophryme  sp.,  female,  17.5  mm;  (D)  Haplophryne  mollis,  metamorphic  male,  13.2  mm;  (E) 
Linophryne  subgen.  Linophryne  sp.,  male,  3.8  mm;  (F)  Linophryne  subgen.  Stephanophryne  indica,  female  8.6  mm;  (G)  Borophryne  apogon, 
male,  4.3  mm  lateral  and  dorsal  views.  (All  from  Bertelsen,  1951.) 


(7.5  mm)  this  length  is  reduced  to  28%  SL,  and  pelvic  fin  rays 
are  absent  in  the  two  known  parasitic  males  (12-16  mm)  as  well 
as  in  all  metamorphosed  females  (10-109  mm  SL). 

Illicium,  2nd  cephalic  ray  and  caruncles.  — In  larvae  of  all  fam- 
ilies except  Caulophrynidae  and  Neoceratiidae,  sexual  dimor- 
phism in  the  development  of  the  illicium  is  present.  In  females, 
the  illicium  rudiment  is  club-shaped  and  protrudes  from  the 
head  or  from  the  bottom  of  a  groove  in  its  enveloping  skin  (Fig. 
168A,  B);  in  males  it  is  represented  only  by  the  tiny  subdermal 
rudiment  of  the  illicial  bone.  Similarly,  the  external  rudiment 
of  the  second  cephalic  rays  of  Diceratiidae  and  Ceratiidae  as 


well  as  the  caruncles  of  the  latter  family  are  present  in  the  female 
larvae  and  absent  in  the  males  (Figs.  168C,  D;  169E). 

Among  the  16  known  Caulophryne  and  the  1 1  known  Neo- 
ceratias larvae,  no  sexual  dimorphism  has  been  observed  (Ber- 
telsen, 1951).  In  Caulophryne,  in  which  metamorphosed  fe- 
males lack  an  escal  bulb  with  photophore  but  have  a  well- 
developed  illicium,  the  rudiment  protrudes  on  the  dorsal  side 
of  the  head  in  the  same  position  as  in  other  ceratioid  larvae 
(Fig.  167A).  In  Neoceratias.  in  which  the  illicium  is  completely 
absent  in  the  metamorphosed  females,  all  larvae  have  an  elon- 
gated cylindrical  illicium  rudiment  (pigmented  in  larger  larvae 
(Fig.   I67B)  slightly  protruding,  in  a  position  unique  among 


BERTELSEN:  CERATIOIDEI 


329 


Fig.  168.  Ceratioid  larvae.  (A-B)  Gigantactinidae.  (A)  Gigantactis  sp.,  female,  8.5  mm;  (B)  Rhynchactis  leplonema.  female,  7.2  mm;  (C-E) 
Ceratiidae;  (C)  Cryptopsaras  couesi.  female,  5.0  mm;  (D)  Cryplopsaras  couesi.  male,  5.0  mm;  (E)  Ceratias  holboetli.  7.6  mm;  (F)  Centrophrynidae, 
Centrophryne  spinulosa.  male,  7.2  mm.  (All  from  Bertelsen.  1951.) 


ceratioid  larvae,  on  the  tip  of  the  snout,  just  above  the  upper 
jaw. 

Barbels.  — In  Linophryne.  the  only  ceratioid  genus  in  which  the 
metamorphosed  females  have  a  hyoid  barbel,  a  rudiment  of  this 
is  present  as  an  opaque,  wart-like  thickenmg  of  the  skin  in  female 
larvae  of  more  than  about  10  mm  SL  (Fig.  167C). 

The  larvae  of  both  sexes  of  the  single  known  species  of  the 
family  Centrophrynidae  differ  from  all  other  ceratioid  larvae  in 
having  a  digitiform,  hyoid  barbel  (Fig.  168F).  The  barbel  re- 
mains digitiform  in  the  metamorphic  males,  but  after  meta- 
morphosis it  is  in  both  sexes  reduced  to  a  low  papilla  which 
gradually  disappears  in  females  larger  than  about  50  mm. 


Spines.  —  Both  male  and  female  larvae  of  the  Linophryne  sub- 
genus Linophryne  and  the  linophrynid  genus  Borophryne  differ 
from  all  other  ceratioid  larvae  in  having  well-developed,  pointed 
sphenotic  spines  (Fig.  167E,  G).  None  of  the  other  spines  (pre- 
opercular,  quadrate,  articular,  etc.)  of  the  head  skeleton  char- 
acteristic of  females  of  different  genera  is  developed  before 
larval  metamorphosis. 

Relationships 

Current  principal  hypotheses.— Thai  the  Ceratioidei  represent  a 
monophyletic  line  appears  most  clearly  from  the  fact  that  they 
all  differ  from  other  Lophiiformes  in  having  developed  an  ex- 
treme and  unique  sexual  dimorphism  in  which  the  males  are 


330 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  169.  Ceratioid  larvae.  (A-B)  Himantolophidae.  (A)  Himantolophus  groenlandicus  gr.,  female,  6.0  mm;  (B)  Himanlotophus  alhinares  gr., 
male,  7.1  mm;  (C-D)  Melanocetidae;  (C)  Melanocelus  Ijohnsoni.  female,  12.0  mm;  (D)  Metanocetus  murrayi.  male,  6.0  mm;  (E)  Diceratiidae, 
Diceralias  sp.,  metamorphic  female,  10.5  mm;  (F)  Thaumatichthyidae,  Thauinalichlhys  sp.,  female,  6.4  mm.  (All  from  Bertelsen,  1951.) 


dwarfed,  lack  an  external  illicium,  and  are  furnished  with  den- 
ticular teeth  adapted  to  attach  to  the  female. 

We  may  assume  an  ogcocephalid  or  chaunacid-like  ancestral 
ceratioid  which,  from  the  benthic  and  littoral  environment  of 
its  ancestors,  has  invaded  the  bathypelagic  zone  of  the  ocean. 
Probably  this  evolution  has  passed  through  forms  in  which  the 
adults  were  benthic,  while  the  juveniles  after  metamorphosis 
continued  the  pelagic  life  of  the  larvae  during  adolescence  as 
for  instance  found  in  the  family  Chaunacidae  and  as  retained 
or  reestablished  in  the  ceratioid  genus  Thaumatichthys.  This 
move  to  a  new  adaptive  zone  has  led  to  a  dimorphism  which 
separates  the  tasks  of  the  two  sexes,  the  females  obtaining  ad- 


aptations to  the  bathypelagic  conditions  of  the  lophiiform  feed- 
ing strategy  by  passive  luring,  the  males  being  adapted  solely  to 
active  search  for  a  sexual  partner.  In  both  sexes  the  change  from 
benthic  to  pelagic  life  has  induced  a  number  of  changes  of  which 
the  most  important  are:  loss  of  the  pelvic  fins;  a  change  of  the 
position  and  development  of  their  limb-like  pectoral  fins,  now 
used  for  counteracting  gravity  during  swimming  rather  than  for 
support  and  movement  on  the  bottom;  and  a  general  trend  to 
reduce  their  density  by  reduction  of  bony  structures  and  by 
retaining  the  thick  subdermal  layer  of  gelatinous  tissue  present 
in  the  larvae.  In  the  latter  character  and  in  the  position  and 
shape  of  the  pectoral  fins,  they  may  be  regarded  as  neotenic  as 


BERTELSEN:  CERATIOIDEI 


331 


_1 L_ 


Fig.  1 70.  Ceratioid  larvae.  Oneirodidae.  (A)  Oneirodes  sp.,  female,  8  mm;  (B)  Dolopichthys  sp.,  male,  5.4  mm;  (C)  Chaenophryne  draco  gr., 
female,  4.0  mm;  (D)  Chaenophryne  longiceps.  female,  5.5  mm;  (E)  Pentherkhlhys  sp.,  female,  10.6  mm;  (F)  Microlophkhthys  Tmkrolophus, 
female,  9.0  mm.  (All  from  Bertelsen,  1951.) 


proposed  by  Richard  Rosenblatt  (quoted  by  Moser,  1981).  In 
females  the  changed  conditions  have  led  to  extreme  specializa- 
tions of  the  luring  and  feeding  mechanisms  at  the  expense  of 
their  swimming  ability,  while  in  the  males  this  has  induced 
different  specializations  in  their  attachment  mechanisms  and 
sense  organs  and  a  development  into  more  streamlined  and 
efficient  swimmers. 

The  present  division  of  the  Ceratioidei  into  families  is  based 
mainly  on  revisions  by  Regan  (1912a.  1926)  and  Regan  and 
Trewavas  (1932).  Some  changes  have  been  introduced  by  Ber- 
telsen (1951)  and  Pietsch  (1972)  resulting  in  the  present  rec- 
ognition of  the  1 1  families  listed  in  Table  89. 

The  phylogenetic  relationships  between  the  families  of  the 


Ceratioidei  are  still  uncertain.  The  main  reason  for  this  is  that 
most  of  the  derived  osteological  characters  shared  by  two  or 
more  families  are  reduction  states  or  loss  of  parts  following  the 
general  trend  mentioned  above  and  similarities  in  such  char- 
acters may  in  many  cases  represent  convergent  developments. 
At  the  same  time  most  of  the  diagnostic  family  characters  which 
represent  new  structures  or  specialization  of  organs  are  auta- 
pomorphic  (and  for  this  reason  not  included  in  Table  89).  The 
more  conspicuous  of  these  are:  an  extreme  prolongation  of  dor- 
sal and  anal  rays  of  Caulophrynidae;  a  dense  cover  of  large 
papillae  on  snout  and  chin  of  female  Himantolophidae;  a  hyoid 
barbel  in  larvae  and  juveniles  of  both  sexes  of  Centrophrynidae; 
photophore-bearing.  modified  dorsal  fin  rays  (caruncles)  in  fe- 


•8 

o 


c 
u 


o 


O 

O 

a 


2 

3 


so 


BERTELSEN:  CERATIOIDEI 


333 


male  Ceratiidae;  and  very  different  high  specialization  of  the 
illicial  and  jaw  mechanisms  of  female  Neoceratiidae,  Thau- 
matichthyidae  and  Gigantactinidae. 

The  interrelationships  of  the  ceratioid  families  have  been  dis- 
cussed by  Regan  (1912a,  1926),  Regan  and  Trewavas  (1932), 
Bertelsen  (1951),  and  Pietsch  (1972,1 979),  the  latter  illustrating 
with  branching  diagrams,  alternative  proposals  for  phylogenetic 
relationships  of  the  families.  However,  no  detailed  analysis  or 
full  discussion  of  the  basis  for  these  proposals  has  been  pre- 
sented. For  this  reason  the  dendrogram  shown  in  Fig.  1  7 1  should 
be  regarded  only  as  a  very  schematic  compilation  of  the  ex- 
pressed views,  following  most  closely  Pietsch  ( 1 979:  fig.  26)  with 
some  modifications  discussed  below. 

In  accordance  with  Bertelsen  (1951)  and  Pietsch  (1979)  it  is 
assumed  that  sexual  parasitism  has  been  established  indepen- 
dently in  different  phylogenetic  lineages.  The  observation  of  a 
parasitic  male  (character  26)  in  a  representative  of  one  of  the 
1 5  oneirodid  genera  (Pietsch,  1 976)  makes  it  extremely  unlikely 
that  the  five  families  in  which  such  males  have  been  observed 
represent  a  monophyletic  line.  (Furthermore,  this  observation 
underlines  the  possibility  that  sexual  parasitism  might  be  found 
in  other  families  as  well.)  It  seems  that  the  evolutionary  step 
from  the  temporary  attachment  of  the  male  to  the  female,  by 
means  of  the  denticular  teeth  present  in  all  ceratioid  males  (and 
resulting  in  a  close  and  protracted  contact  between  the  dermis 
of  the  pair),  to  a  fusion  of  their  tissues  is  a  less  drastic  event 
than  it  might  be  supposed  and  has  been  established  indepen- 
dently in  different  taxa  and  possibly  even  facultative  in  some, 
as  proposed  by  Pietsch  (1976). 

Presence  of  an  escal  photophore  (no.  1 1)  is  presumed  to  be  a 
synapomorphy  separating  the  other  families  from  Caulophryn- 
idae  (and  ?Neoceratiidae),  a  primitive  sister-group.  This  implies 
that  the  similarity  of  some  derived  character  states  (nos.  4.  6, 
9,  15)  to  those  of  one  or  more  of  the  families  Linophrynidae. 
Gigantactinidae,  and  Ceratiidae  is  due  to  convergence  in  these 
bone  reductions.  The  alternative,  proposed  by  Pietsch  (1979), 
that  these  families  were  derived  from  a  caulophrynid-like  ances- 
tor, would  imply  that  the  escal  photophore  has  been  evolved 
independently  in  two  separate  lineages.  Morphologic  and  his- 
tologic studies  of  these  organs  in  different  families  show  simi- 
larities in  such  details  that  this  seems  extremely  unlikely  [cf  for 
instance  Brauer.  1908  (Gigantactis);  O'Day,  1974  (Oneirodes); 
Hansen  and  Herring,  1977  (Linophryne);  and  Munk  and  Ber- 
telsen, 1980  (Chaenophryne)]. 

Based  on  a  number  of  shared  derived  character  states  (nos. 
6,  7,  15,  20,  and  presence  of  teeth  externally  on  the  jaws)  it  has 
been  assumed  that  Neoceratiidae  are  closely  related  to  Gigan- 
tactinidae. However,  they  differ  considerably  in  other  characters 
(nos.  5,  9,  13,  14,  21,  and  24)  and  especially  in  the  illicial  and 
jaw  mechanisms  of  the  females.  While  the  complete  loss  of 
illicium  in  neoceratiids  undoubtedly  is  a  derived  character  state 
it  remains  uncertain  whether  this  family  is  derived  from  ances- 
tors with  or  without  escal  photophores.  As  discussed  in  the 
following  section,  some  larval  characters  might  indicate  the  lat- 
ter possibility.  In  reference  to  this  the  numerous  characters 
shared  by  the  two  genera  of  gigantactinids  leaves  no  doubt  that 
the  lack  of  photophore  in  Rhynchactis  is  due  to  secondary  re- 
duction (Bertelsen  et  al.,  1981).  While  none  of  the  highly  spe- 
cialized families  Linophrynidae,  Gigantactinidae,  and  Cerati- 
idae appear  closely  related,  their  shared  derived  character  states 
may  indicate  a  common  descendence  as  shown  in  Figure  171. 
As  pointed  out  by  Pietsch  (1972)  Centrophrynidae  has  retained 


a  number  of  primitive  character  states  but  may  be  more  closely 
related  to  Ceratiidae  than  to  any  other  family,  and  Thaumati- 
chthyidae  are  most  probably  derived  from  an  oneirodid-like 
ancestor  (Bertelsen  and  Struhsaker,  1977).  The  remaining  four 
families  Melanocetidae,  Himantolophidae,  Diceratiidae,  and 
Oneirodidae  appear  more  similar  to  each  other  than  the  more 
specialized  families  mentioned  above,  but  as  their  shared  char- 
acter states  are  nearly  all  primitive  their  interrelationships  are 
uncertain.  The  position  of  Melanocetidae  in  the  dendrogram 
(Fig.  171)  is  based  on  the  presumption  that  a  reduction  of  the 
number  of  dorsal  fin  rays  to  less  than  10  is  synapomorphic 
within  the  following  series  of  families. 

Except  for  the  significance  of  observed  sexual  parasitism 
the  characters  of  the  males  have  not  been  considered  in  previous 
discussions  of  the  interrelationships  of  the  ceratioid  families. 
The  presence  of  denticular  teeth  shared  by  all  families  is  a  de- 
rived character  state  in  relation  to  all  other  Lophiiformes.  The 
absence  of  such  denticles  on  the  snout  observed  in  caulophryn- 
ids  and  neoceratiids  may  represent  a  primitive  state  within  the 
suborder.  In  accordance  with  the  classification  based  on  the 
characters  of  the  females  or  shared  by  the  sexes,  the  males  are 
highly  but  very  differently  specialized  in  the  families  Ceratiidae, 
Gigantactinidae,  and  Linophrynidae  while  the  least  number  of 
presumed  derived  character  states  are  found  in  Melanocetidae, 
Himantolophidae,  and  Diceratiidae. 

Within  the  families  the  inter-generic  relationships  appear  close 
and  relatively  simple  in  the  four  families  divided  into  two  gen- 
era. In  each  of  these  one  of  the  genera  shows  more  derived 
character  states  in  reductions  and  specializations  than  the  other. 
(Rohia  in  Caulophrynidae;  Phrynichthys  in  Diceratiidae;  Tfiau- 
matichthys  in  Thaumatichthyidae;  Cryptopsaras  in  Ceratiidae 
and  Rhynchactis  in  Gigantactinidae).  Among  the  five  genera  of 
the  well-defined  family  Linophrynidae,  L/«op/;n'««' appears  the 
most  derived  (females  with  photophore  carrying  barbels).  Bor- 
ophryne  and  Acenlrophryne  seem  closely  related  to  this  genus 
(very  similar  osteology  and  dentition)  while  each  of  the  genera 
Edriolychmts  and  Photocorynus  appear  more  isolated;  the  latter 
has  retained  a  number  of  primitive  or  less  derived  character 
states  (nos.  10,  13,  15,  17). 

In  contrast  to  the  other  ceratioid  families  no  conspicuous 
distinctive  characters  have  been  found  which  are  common  to 
the  large  assemblage  of  genera  united  in  the  family  Oneirodidae. 
However,  the  presence  of  quadrate  and  articular  spines  in  most 
of  the  genera  and  shared  only  with  the  closely  related  thau- 
matichthyids  might  be  significant  and  their  absence  in  some 
genera  could  be  due  to  secondary  reduction.  On  the  basis  of 
osteological  characters  the  evolutionary  relationships  of  9  of 
the  1 5  genera  were  studied  by  Pietsch  ( 1 974)  and  notes  on  some 
of  the  others  have  been  added  by  Bertelsen  and  Pietsch  (1975) 
and  Pietsch  (1975).  According  to  these  studies  Spiniphryne  ap- 
pears the  most  primitive  of  these  genera,  having  retained  well- 
developed  dermal  spines,  among  a  number  of  other  primitive 
character  states.  Among  the  most  specialized  genera  are  Lo- 
phodolos  (reduction  or  loss  of  some  elements  of  the  skeleton 
and  enlargements  of  others)  and  Chaenophryne  (lack  of  sphen- 
otic,  quadrate  and  articular  spines,  shape  of  opercular  bones  and 
a  unique  structure  of  ossifications;  Pietsch,  1975). 

Contribution  of  early  life  history  stages.— Apart  from  meristic 
and  osteological  characters  shared  with  adults,  the  larvae  of 
ceratioid  taxa  differ  from  each  other  only  in  pigmentation  and 
to  some  extent  in  morphology.  As  the  pigment  patterns  vary 


334 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


greatly  within  families  only  the  latter  characters  may  be  relevant 
to  the  discussion  of  the  relationships  of  the  families. 

The  assumption  by  Bertelsen  (1951)  that  the  Caulophrynidae 
are  isolated  from  all  other  ceratioids  was  based  mainly  on  three 
larval  characters:  ( 1 )  presence  of  pelvic  fins;  (2)  absence  of  sexual 
dimorphism  in  the  illicial  rudiments;  and  (3)  lack  of  a  distal 
swelling  of  these  rudiments  representing  the  rudiment  of  an  escal 
bulb.  The  two  latter  character  states  indicate  that  the  absence 
of  an  escal  photophore  in  caulophrynids  is  not  due  to  a  sec- 
ondary reduction.  As  expressed  by  Pietsch  (1979)  who  found  a 
number  of  additional  resemblances  between  Caulophryne  and 
less  derived  Lophiiformes  (lophiids  and  antennarioids):  "That 
these  primitive  character  states  suddenly  reappeared  in  a  lineage 
that  arose  from  an  ancestorderivedinall,  is  highly  improbable." 

The  possibility,  mentioned  above,  that  the  neoceratiids  may 
represent  a  similar  isolated  lineage  derived  from  an  ancestor 
without  escal  photophore,  is  based  on  the  same  larval  characters: 
the  absence  of  sexual  dimorphism  in  their  illicial  rudiments 
which  lack  distal  swellings.  However,  the  fact  that  neoceratiids 
and  caulophrynids  share  these  primitive  character  states  fur- 
nishes no  information  on  their  relationship.  In  other  larval  char- 
acters, especially  body  shape  and  size  of  pectoral  fins,  the  two 
families  are  extremely  different. 

The  assumption  that  the  absence  of  escal  photophore  in  the 
gigantactinid  genus  Rhynchactis  is  due  to  a  secondary  reduction 
is  confirmed  by  the  presence  of  a  club-shaped  illicial  rudiment 
in  the  female  larvae. 

Little  information  on  the  relationships  between  the  ceratioid 
families  can  be  obtained  from  other  observed  differences  in 
larval  morphology.  The  greatly  enlarged  pectoral  fins  pres- 
ent only  in  gigantactinids  and  caulophrynids  may  as  assumed 
by  Pietsch  (1979)  represent  a  primitive  character  state  which 
has  been  retained  separately  in  the  two  lineages.  The  most  con- 
spicuous derived  character  states  of  the  larvae  are  the  extreme 
elongation  of  the  body  in  neoceratiids,  the  hump-backed  shape 
of  ceratiids,  and  the  barbels  of  centrophrynids.  Being  each  re- 
stricted to  a  single  family  they  only  confirm  the  distinct  sepa- 
ration of  these  lineages. 

Within  the  families,  inter-generic  comparisons  of  larvae  are 
possible  only  in  Gigantactinidae,  Linophrynidae,  Ceratiidae  and 


Oneirodidae.  Of  the  remaining  seven  families,  four  are  mono- 
typic  and  in  each  of  three,  which  are  divided  into  two  genera, 
larvae  of  only  one  is  known.  In  each  of  these  families  very 
distinct  inter-generic  differences  in  larval  pigment  patterns  have 
been  found. 

This  larval  character,  retained  in  juveniles  of  both  sexes,  has 
been  one  of  the  main  keys  to  the  identification  of  the  free-living 
metamorphosed  males  and  thus  has  contributed  considerably 
to  the  concept  of  the  relationships  within  the  ceratioid  families. 

The  fact  that  separation  of  larvae  (and  males)  below  generic 
level  has  been  possible  only  in  those  exceptional  cases  where 
intra-generic  differences  above  species  rank  (subgenera,  species- 
groups,  etc.)  have  been  observed,  underlines  that  within  this 
suborder  the  term  "genus"  indicates  a  remarkably  well-defined 
and  natural  group. 

However,  little  information  on  phylogenetic  relationships 
within  the  families  has  been  obtained  from  the  study  of  the 
larvae.  The  difficulties  in  interpreting  their  character  states  is 
well  illustrated  in  the  Linophrynidae.  Two  apparently  typical 
derived  larval  character  states  occur  in  this  family:  (1)  well 
developed  sphenotic  spines  (within  larval  Lophiiformes  found 
only  in  the  linophrynid  genus  Borophryne  and  in  one  of  the 
three  subgenera  of  Linophryne),  and  (2)  a  characteristic  sub- 
dermal  pigment  pattern  (found  only  in  the  linophrynid  genus 
Haplophryne  and  in  the  two  subgenera  of  Linophryne  lacking 
larval  sphenotic  spines).  If  it  is  assumed  very  unlikely  that  these 
specializations  have  evolved  independently  in  different  genera 
of  the  same  family,  the  only  alternative  is  that  apparently  prim- 
itive character  states  are  in  fact  due  to  three  secondary  reduc- 
tions: (1)  lack  of  sphenotic  spines  in  two  L/>!o/)/!n'«e  subgenera; 
(2)  lack  of  barbels  in  female  Borophryne,  making  this  one  more 
subgenus  oi  Linophryne,  and  (3)  lack  of  subdermal  pigment  in 
one  of  the  subgenera  of  Linophryne  and  in  Borophryne. 

Ceratioids  are  still  very  incompletely  known  and  future  stud- 
ies on  additional  characters  and  as  yet  unknown  forms  may 
bring  answers  to  at  least  some  of  the  many  questions  about  their 
phylogenetic  relationships. 

Zoological     Museum,     University     of    Copenhagen, 
Copenhagen  2100  0,  Denmark. 


Atherinomorpha:  Introduction 

B.  B.  COLLETTE 


THE  superorder  Atherinomorpha  (Greenwood  et  al.,  1966) 
includes  the  atherinoids  (silversides  and  phallostethids), 
cyprinodontoids  (killifishes),  and  beloniforms  (halfbeaks  and 
their  relatives),  first  grouped  together  by  Rosen  (1964)  as  the 
order  Atheriniformes.  The  series  Atherinomorpha  was  redefined 
by  Rosen  and  Parenti  (1981)  as  including  the  Atherinoidei  (of 
uncertain  rank),  Cyprinodontiformes  and  Beloniformes. 

Utilizing  17  apomorph  characters,  Rosen  and  Parenti  (1981) 
found  1 0  synapomorphies  uniting  the  atherinoids,  Cyprinodon- 
tiformes, and  Beloniformes.  Two  of  these  involve  early  life  his- 
tory characters:  complete  separation  of  embryonic  afferent  and 
efferent  circulation  by  development  of  the  heart  in  front  of, 
rather  than  under,  the  head  and  the  presence  of  large  demersal 
eggs  with  long  adhesive  and  short  filaments  and  many  lipid 


globules  that  coalesce  at  the  vegetal  pole.  Four  additional  syn- 
apomorphies between  the  Cyprinodontiformes  and  Beloni- 
formes show  the  atherinoids  to  be  the  plesiomorphic  sister  group 
of  these  two  orders. 

Rosen  and  Parenti  (1981)  were  unable  to  find  derived  char- 
acters to  unite  the  atherinoids  as  a  monophyletic  group  but 
White  et  al.  (this  volume)  have  discovered  two  early  life  history 
characters  which  define  the  Atheriniformes  as  the  plesiomor- 
phous  sister  group  of  the  Cyprinodontiformes  plus  Beloni- 
formes. 

National  Marine  Fisheries  Service  Systematics 
Laboratory  ,  National  Museum  of  Natural  History, 
Washington,  District  of  Columbia  20560. 


Beloniformes:  Development  and  Relationships 
B.  B.  CoLLETTE,  G.  E.  McGowEN,  N.  V.  Parin  and  S.  Mito 


THE  Beloniformes  (or  Synentognathi)  is  an  order  of  atherino- 
morph  fishes  containing  five  families.  37  genera,  and  about 
180  species.  Species  of  the  Adrianichthyidae  inhabit  fresh  and/ 
or  brackish  waters.  Most  species  of  the  other  four  families  are 
epipelagic  marine  fishes  but  several  genera  of  Belonidae  and 
Hemiramphidae  are  restricted  to  fresh  waters  and  a  few  other 
genera  contain  estuarine  and  freshwater  as  well  as  marine  species. 
Two  groups  have  been  recognized  under  various  names  by  a 
series  of  authors  starting  with  Schlesinger  ( 1 909)  and  continuing 
through  Regan  (1911b),  Nichols  and  Breder  (1928),  Rosen  ( 1 964), 
and  Collette  (1966).  Each  of  these  groups  contains  two  families, 
the  Scomberesocidae  and  Belonidae  in  the  first,  the  Hemiram- 
phidae and  Exocoetidae  in  the  second.  Recently,  Rosen  and 
Parenti  (1981)  expanded  the  Beloniformes  by  adding  the  Ad- 
rianichthyidae to  the  order  as  a  separate  suborder  Adrianichthy- 
oidei,  the  sister  group  of  the  Exocoetoidei  (containing  two  su- 
perfamilies  Scomberesocoidea  and  Exocoetoidea). 

Development 

Eggs 

Most  beloniform  fishes  produce  large  spherical  eggs  with  at- 
taching filaments,  characters  they  share  with  other  atherino- 
morph  fishes  (Rosen  and  Parenti,  1981).  Adrianichthyid  eggs 
are  the  smallest  (1.0-1.5  mm  in  diameter);  followed  by  exo- 
coetids  (generally  1.5-2  mm);  Hemiramphidae  (typically  1.5- 
2.5  mm);  Scomberesocidae  (slightly  elliptical,  1 .5-2.5  mm);  and 
belonid  eggs  which  are  generally  the  largest  (most  3-4  mm) 
(Table  90).  These  eggs  typically  have  a  homogeneous  yolk  and 
a  relatively  small  perivitelline  space.  According  to  Kovalev- 
skaya  (1982),  eggs  with  long  filaments,  distributed  over  the  en- 


tire sphere  of  the  egg  (one  filament  may  be  thicker  and  longer 
than  the  others)  should  be  considered  primitive.  Such  eggs  are 
found  in  the  Belonidae,  some  Hemiramphidae,  primitive  flying- 
fishes  of  the  genera  Fodialor  and  Parexocoetus.  and  also  in  many 
of  the  highly  specialized  species  of  the  subfamily  Cypselurinae. 

Eggs  of  the  Adrianichthyoidei  contain  numerous  small  oil 
globules  which  coalesce,  at  least  to  some  extent,  during  devel- 
opment (Matsui.  1949),  as  in  the  Atheriniformes  and  Cyprino- 
dontiformes  (Rosen  and  Parenti,  1981).  Exocoetoid  eggs  either 
contain  minute,  scattered  oil  globules  (Fig.  1 76C)  or  lack  oil 
globules  (Table  90). 

Adrianichthyid  eggs  have  filaments  distributed  over  the  entire 
chorion,  a  condition  we  refer  to  as  uniformly  spaced.  Most  of 
these  filaments  are  short,  0.21-0.35  mm  in  Horaichthys  setnai 
(Kulkami,  1940),  however,  on  one  portion  of  the  chorion  they 
are  as  long  as  or  longer  than  the  egg  diameter  (Fig.  1 72).  Pietri 
(1983)  described  these  two  topographically  distinct  types  of  fil- 
aments from  the  chorionic  surface  of  Oryzias  latipes  as  non- 
attaching  and  attaching.  Non-attaching  filaments  showed  a  reg- 
ular distribution  over  the  chorion  with  an  interfilament  distance 
of  about  65-70  /im,  and  functioned  to  maintain  the  integrity  of 
the  egg  cluster.  Attaching  filaments  were  located  at  one  pole  of 
the  egg  forming  a  clump  of  about  25  filaments  that  united  with 
those  of  neighboring  eggs  to  anchor  the  egg  cluster  to  the  gon- 
oduct  of  the  female.  In  Oryzias  melastigma,  the  attaching  fil- 
aments also  anchor  the  eggs  to  the  female  (Job,  1940)  or  to 
filamentous  algae. 

The  eggs  of  most  scomberesocids  {Scomberesox,  Namchthys 
and  Elassichthys)  are  pelagic,  without  long  filaments.  Eggs  of 
Scomberesox  (Fig.  1 73A),  however,  have  short  bristles  that  ap- 
parently represent  remnants  of  chorionic  filaments  (see  Boehlert, 


Table  90.    Eggs  of  Beloniformes  Fishes.  Much  of  this  information  is  based  entirely  on  illustrations  from  the  cited  references. 


Filaments 

Diameter 
(mm) 

Oil 

globule 

Taxon 

Arrangement 

Number 

Length  (mm) 

Remarks  and  sources 

Adrianichthyidae 

Horaichthys  setnai 

1 

Numerous  co- 

2 types: 

Demersal,  Kulkar- 

alesce  into  1 5- 

uniform 

Type  1.  most  nu- 

0.21-0.35 

ni,  1940 

30 

localized 

merous 
Type  2,  many 

>egg  dia. 

Oryzias  javanicus 

Numerous  co- 
alesce into  1 

2  types 

Ahlstrom  notes 

O.  latipes 

1.27-1.3 

Numerous  co- 

2 types: 

Ahlstrom  notes; 

alesce  into  1 

uniform 
localized 

Type  1 ,  nonat- 
taching.  most 
numerous 

Type  2,  attach- 
mg,  -25 

Matsui,  1949; 
Pietri,  1983; 
Hart  et  al., 
1984 

O.  luzonensis 

-1.5 

Numerous  co- 
alesce into  1 

2  types 

Ahlstrom  notes 

O  melastigma 

1.0-1.2 

30-40  coalesce 

2  types: 

Job,  1940 

into  1 

uniform 
localized 

Type  1,  most  nu- 
merous 
Type  2,  many 

<egg  dia. 
>egg  dia. 

335 


336 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  90.    Continued. 


Filaments 

Diameter 

Oil 

Taxon 

(mm) 

globule 

Arrangement 

Number 

Length  (mm) 

Remarks  and  sources 

Scomberesocidae 

Cototabis  saira 

1.7  X  1.9  (off 
round) 

No 

Polar  cluster 
lateral 

12-15 

1 

Demersal.  SWFC, 
unpublished  and 
original 

C.  adocetus 

-2.5 

None 

Orton,  1964 

Scomberesox 

Off  round. 

No 

Uniform  or 

Numerous 

Short,  rigid 

Pelagic,  Hardy, 

saurus 

range  of 
mean  diame- 
ters = 

2.32  X  2.52 

groups 

1978a 

S.  simulans 

None 

Hubbs  and  Wis- 
ner,  1980 

Belonidae 

Ablennes  hians 

3.0-3.5;  mean 

No 

Uniformly- 

1-6  per  tuft;  to- 

>egg dia. 

Hardy,  1978  and 

=  3.16  ripe 

spaced  tufts 

tal  37-59  (N  = 

original 

ovarian 

6) 

Belone  belone 

3.0-3.5 

No 

Uniform 

60-80 

4-18 

Demersal,  Russell, 

1976 

Pseudolylosurus 

1.2-1.4;  mean 

CoUette,  1974a 

microps 

=  1.23 

and  original 

Slrongylura  exilis 

2.3-2.8;  mean 
=  2.5  (run- 
ning ripe) 

No 

Uniform 

18-30 

Longer  than  egg 
dia. 

Original 

S.  Jluvialilis 

2.9-3.2;  mean 
=  3.1  (ovari- 
an) 

Original 

S.  hubbsi 

2.50-3.14; 
mean  =  2.75 
(ovarian) 

Collette,  1974c 

S.  incisa 

3.5-4.6;  mean 
=  3.9  (ovari- 
an) 

Original 

S.  krefftii 

2.7-4.0;  mean 
=  3.4  (ovari- 
an) 

Original 

S.  marina 

3.5-3.6 

No 

Uniform 

Numerous 

Variable,  but 
generally  <egg 
diameter 

Demersal,  Hardy, 
1978a 

S.  notata 

3.67-4.18; 
mean  =  3.95 

Uniform 

Demersal.  Breder, 

1959 

S.  slrongylura 

2.5 

No,  but  one  or 

Uniform 

Numerous  57  in 

All  long,  but  1-2 

Demersal,  Job  and 

more  clear 

illus. 

areas  longest 

Jones,  1938 

vesicles  project 

into  yolk 

Tylosurus  acus 

3.22-4.0 

Uniformly- 
spaced  tufts 

2-3  per  tuft 

Longer  than  egg 
dia. 

Demersal,  Hardy, 
1978a 

T.  a.  melanotus 

3.2-3.5 

No 

Uniformly- 
spaced  tufts 

2-3  per  tuft;  to- 
tal @  100 

2-3  X  egg  dia. 

Mito,  1958 

T.  crocodilus 

4.0-4. 1 

Minute 

Numerous 

Long 

Demersal,  Masu- 
rekar,  1967 

T.  punctulalus 

3.5-4.3;  mean 
=  3.9 

Original 

Xenenlodon  canci- 

2.9-3.2 

No 

Foster,  1973  and 

la 

original 

Hemiramphidae 

Dermogenvs  pusil- 

Viviparous 

No 

Mohr,  1936a; 

lus 

Brembach,  1976 

Euleptorhamphus 

1.1  (ovarian) 

"Disorderly" 

6.0-6.5 

Demersal;  Parin 

viridis 

and  Gorbunova, 
1964 

Hemira  mph  us  fa  r 

2.8-3.1 

Present 

Breder  and  Ro- 
sen, 1966 

He.  marginatus  = 

Mean  =  2.59 

No 

8-12  tufts 

4-6/tuft 

Both  >  &  <  egg 

Talwar,  1968 

He.  lutkei 

diameter 

COLLETTE  ET  AL.:  BELONIFORMES 


337 


Table  90.    Continued. 


nio  mAtdr 

Oil 

Filamenls 

Taxon 

(mm) 

globule 

Arrangement 

Number 

Length  (mm) 

Remarks  and  sources 

Hemirhamphodon 

Viviparous 

Mohr,  1936c; 
Brembach,  1976 

Hyporhamphus 

2.2-2.6 

14-16  tufts 

3-4/tuft 

Sudarsan.  1968b 

guoyi 

Hy.  capensis 

-1.6 

Present 

Smith,  1933b 

Hy.  intermedius 

1.3-1.4 

Minute 

Bipolar 

3^/tuft, 
1  polar 

7 

>25 

Uchida  et  al., 
1958 

Hy.  sajori 

2.1-2.3 

Minute 

Bipolar 

4-6/tuft. 
1  polar 

20-60 

Uchida  et  al., 
1958 

Hy.  unifascialus 

-2.0 

Several 

>egg  diameter 

Semibuoyant; 
Hardy,  1978a 

Hy.  xanthoplerus 

1.12-1.44; 
mean  =  1.23 
(ovarian) 

Collette,  1982b 

Melapedalion 

breve 

Nomorhamphus 

Viviparous 

Mohr,  1936b; 
Brembach,  1976 

Oxyporhamphus 

1.9-2.3 

Uniform 

Many 

0.05-0.06 

Pelagic;  Kovalev- 

conve.xus 

skaya,  1965 

O.  micropterus 

1.8-2.1 

No 

Uniform 

74-120 

0.08-0.12 

Pelagic;  Imai, 

Rhychorhamphus 

georgii 

R.  malabaricus 

-1.5 

Zenarchopterus 

Not  viviparous 

Z.  roberlsi 

3.0-3.5;  mean 

=  3.25  (ovari- 

an) 

Exocoetidae 

Cheilopogon  (A.) 

1.4-1.6 

agoo 

Ch.  (Ch.)pinnali- 

1.57-1.70; 

barbalus  califor- 

mean  =  1.64 

nicus 

Ch.  (Ch.)pinnali- 

1.9-2.1 

barbalus  japoni- 

cus 

Ch.  (Proc.)  cyanop- 

1.2  (ovarian. 

Icrus 

prob.  not  ful- 

ly ripe) 

Ch.  (Proc.)  katop- 

tron 

Ch.  (Proc.)  ni- 

2.0-2.2 

gricans 

1.8  (maximum 

ovarian) 

Ch.  (Proc.)  spilop- 

1.79-2.17;                No 

terus 

mean  =  2.02 

Ch.  (Pi.)  heterurus 

1.86                          No 

Ch.  (Pt.)h.  doeder- 

2.0-2.2 

leini  (type  1 ) 

Ch.  (Pl.)h.  doeder- 

1.8-2.2 

leini  (type  2) 

Ch.  (Pi.)  hiikeni 

-1.6 

Ch.  (Pi.)  melanu- 

1.8-1.9 

rus 

Bipolar  ? 


0.07-0.08 


1959 

Kovalevskaya, 
1965 

Demersal;  Koval- 
evskaya, 1965 

Demersal;  Deva- 
nesen,  1937 

Mohr,  1926; 
Brembach,  1976 

Collette,  1982c 


Uniform 

34-66 

Demersal,  Imai, 

1960 

Uniform 

-60 

Hubbs  and  Kam- 
pa,  1946 

Uniform 

56-78 

-10 

Demersal,  Imai, 

1959 

Uniform 

Gibbs  and  Staiger, 

1970 

Uniform 

Numerous 

0.5-0.55 

Pelagic,  Kovalev- 
skaya, 1965 

Uniform 

Numerous 

<egg  diameter 

Pelagic,  Pann  and 
Gorbunova, 
1964 

Gibbs  and  Staiger, 
1970 

Uniform 

12-19 

Long 

Demersal,  Vijay- 

(usually  13) 

araghavan,  1975 

Uniform 

Numerous 

Long 

Demersal,  Hardv, 
1978a 

Bipolar 

30-86 

10 

Demersal,  Imai, 
1959 

Uniform 

30-48 

8-11 

Demersal,  Imai, 
1959 

45-1- 

Hubbs  and  Kam- 
pa,  1946 

Uniform 

Gibbs  and  Staiger, 

1970 

338 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  90.    Continued. 


Diameter 
(mm) 


Oil 
globule 


Arrangemeni 


Length  (mm) 


Remarks  and  sources 


Ch.  (Pt.)  unicolor         1.5-1.6 


Cypselurus  (Cy) 

1.8-2.1 

naresii 

Cy.  (Cy.)  opisthops 

1.5-1.6 

hiraii 

Cy.  (Poec.)  cy- 

1.2  (ovarian. 

anopterus 

prob.  not  ful- 

ly ripe) 

Cy.  (Poec.)  starksi 

1.6 

Exocoetus  mono- 

2.8-3.3  (ovari- 

cirrhus 

an) 

E.  obtusirostris 

2.8-2.9  (ovari- 

an) 

E.  volitans 

1.7-2.0 

2.7-3.0 

Fodiator  acutus 

1.4-1.7;  mean 

=  1.53  (ovari- 

an) 

Hirundichthvs  (D.) 

1.4-1.5 

rondeleti 

H.  (//.)  affinis 

1.6  +  0.1  (ripe 

ovarian) 

H.  (//.)  coroman- 

Mean=  1.87 

delensis 

H.  (//.)  oxycepha- 

1.5-1.7 

lus 

H.  (//.)  speculiger 

1.53-1.87; 

mean  =  1.65 

2.05-2.15 

(ovarian;  after 

swelling) 

Parexocoetus  bra- 

1.2-1.5  (ovari- 

chypterus bra- 

an) 

chyplerus 

P.  memo  memo 

1.7-1.8 

Prognichthys  brevi- 

pinms 

No 


Minute 


Minute 


Bipolar 

Uniform 
Uniform 
Uniform 

Uniform 

None 

None 

None 

None 

Uniform 

Bipolar 
Bipolar 

Bipolar 

Bipolar 
Bipolar 
Bipolar 

Bipolar 

Uniform 

Uniform 


10-12 
18-20 


40-52 
55-88 

40-52 


9  (from  illustra- 
tion) 

90-100 

1,  many 

? 

8-14 

mean  =  1 1 

1,3-5 
5-12 

9-18 

10-25 

18-20 

30 

7-10 

1 

6-10 
10-18 

20 


Short 
Long 


15-21 


>egg  diameter 


Short 

40-1- 

102.5,  4.6 
1.1 

4 

0.15 

5 

0.17-0.18 

<  egg  diameter 

35-40 

1.5-2.0 
1.5-2.0 

3.5-8.0 


Short,  less  than 
egg  diameter 


Demersal,  Gorbu- 
nova  and  Parin, 
1963;  Pann  and 
Gorbunova, 
1964 

Demersal,  Imai, 
1959 

Demersal,  Imai, 
1959 

Gibbs  and  Staiger, 
1970 

Demersal,  Imai, 

1959 
Pelagic,  Kovalev- 

skaya,  1964 
Pelagic.  Breder, 

1938 
Pelagic,  Bruun, 

1935 
Kovalevskaya, 

1964 
Demersal,  Breder, 

1938 

Demersal,  D'An- 

cona.  1929 
Kovalevskaya, 

1972 
Demersal,  Evans, 

1962 

Demersal, 

Vijayaraghavan, 
1973 

Demersal,  Imai, 
1960 

Demersal,  Munro, 
1954 

Parin  and  Gorbu- 
nova, 1964 

Demersal,  Imai, 
1959 

Demersal,  Tsuka- 
hara  and  Shiok- 
awa,  1957 

Kovalevskaya, 
1982 


this  volume).  Cololabis  is  the  only  scomberesocid  with  fila- 
ments, a  polar  cluster  of  relatively  long  filaments  plus  a  single 
long  lateral  filament  (Fig.  173B).  Cololabis  eggs  typically  are 
attached  to  floating  objects  such  as  kelp. 

Belonid  eggs  have  filaments  (Table  90).  typically  long,  nu- 
merous and  uniformly  spaced  over  the  chorion  (Fig.  1 74A).  In 
at  least  one  species,  Strongylura  strongylura.  some  filaments  are 
markedly  longer  than  others  (Fig.  1 74B),  as  in  the  adrianich- 
Ihyids.  The  filaments  on  eggs  of  Tylosaurus  acus  are  arranged 
in  uniformly  distributed  tufts  containing  2-3  filaments  each  (Fig. 
174C). 


Hemiramphids  have  eggs  with  attaching  filaments  (Rhyn- 
chorhamphus.  Hyporhamphus,  and  Hemiramphus.  Fig.  175A), 
pelagic  eggs  with  very  short  uniformly-spaced  filaments  (O.xy- 
porhamphus.  Fig.  175B),  or  are  viviparous  (Hemirhampliodon, 
Dermogenys,  Nomorhamphus).  Filaments  were  not  reported  on 
ovarian  eggs  of  Dermogenys  by  Flegler  (1977)  and  we  did  not 
note  their  presence  in  Hcmirhamphodon  or  Zenarchopterus  but 
this  needs  to  be  checked  more  thoroughly.  Kovalevskaya  (1965) 
reported  that  filaments  on  Rhynchorhamphus  georgii  have  a 
bipolar  arrangement;  however,  this  is  not  clear  in  her  illustration 
(Fig.  175 A).  Talwar  (1968)  and  Sudarsan  (1968b)  reported  the 


COLLETTE  ET  AL.:  BELONIFORMES 


339 


Fig.   172.     Adrianichthyidae  egg.  Horaichthys  selnai.  (From:  Kul- 
kami,  1940.) 


filaments  of  what  they  called  Hemiramphus  marginatus  (Fig. 
175C)  and  Hyporhamphus  qitoyi,  respectively,  to  be  grouped  in 
tufts.  The  filaments  in  a  tuft  may  be  of  different  lengths  (e.g.. 
He.  marginatus). 

Most  exocoetids  have  eggs  that  are  attached  with  thin  thread- 
like filaments  to  objects  floating  in  the  water  column  or  to  sea- 
weed growing  near  shore.  The  size  and  structure  of  the  eggs  and 
the  size,  nature,  and  location  of  the  filaments  vary  among  species. 
The  eggs  of  Fodiator  and  Prognichthys  have  uniformly-spaced 
filaments  (Fig.  176B).  Filaments  on  Hirundichthys  eggs  have  a 
bipolar  arrangement.  One  species,  H.  coromandelensis,  has  three 
types  of  filaments  (Vijayaraghavan,  1973),  but  they  are  grouped 
in  a  bipolar  distribution.  This  type  of  egg  has  a  single  long  (103 
mm),  stout  filament  arising  from  the  basal  pole,  which  is  sur- 
rounded by  3-5  medium  length  (.v  =  4.6  mm)  filaments.  Five 
to  1 2  short  (.V  =1.1  mm)  filaments  are  located  at  the  distal  pole 
(Fig.  1 76C).  Chorionic  filaments  in  Parexocoetus.  Cheilopogon 
and  Cypselurus  vary.  Some  species  have  uniformly-spaced  fil- 
aments, whereas  others  have  a  bipolar  arrangement  with  the 
filaments  usually  longer  at  one  pole  than  at  the  other.  Unlike 
all  other  flying  fishes,  species  of  Exocoetus  have  eggs  with  a 
smooth  membrane,  devoid  of  filaments  (Fig.  176A). 

Larvae 

A  relatively  long  incubation  period  is  typical  of  the  Beloni- 
formes  (Kovalevskaya,  1982).  The  eggs  develop  for  one  to  two 
weeks,  and  the  larvae  are  well  formed  and  capable  of  actively 
capturing  food  at  hatching.  Time  of  development  is  comparable 
in  pelagic  and  attaching  eggs.  Following  a  pattern  similar  to  that 


Fig.  173.    Scomberesocidae  eggs.  (A)  Scomberesox  saurns.  SWFC 
Cr.  Est  1-4  Sta.  Surf.  I;  (B)  Cololabis saira.  SWFC  CalCOH.  (Original.) 


reported  for  egg  size,  belonids  hatch  at  the  largest  sizes  (6.8- 
14.4  mm)  followed  by  hemiramphids  (4.8-11  mm),  scomber- 
esocids  (at  least  as  small  as  6.0-8.5  mm),  exocoetids  (3.5-6.1 
mm),  and  adrianichthyids  (3.5-4.5  mm). 

Gut  length  differs  between  the  two  suborders.  Adrianichthy- 
oid  larvae  have  a  short  gut,  as  in  Atheriniformes  and  Cyprin- 
odontiformes,  40-50%  of  standard  length  (Fig.  177A).  Exocoe- 
toid  larvae  are  generally  elongate  and  have  a  straight  gut  extending 
approximately  two-thirds  the  standard  length  (Fig.  I77B-E,  G, 
and  H). 

Presence  of  a  preanal  finfold  appears  to  be  plesiomorphic. 
Job's  (1940)  illustration  of  a  yolk-sac  adrianichthyoid,  Oryzias 
melastigma.  shows  a  preanal  finfold  (Fig.  1 77A),  but  Kulkami's 
( 1 940)  illustration  of  a  yolk-sac  Horaichthys  setnai  does  not.  A 
preanal  finfold  is  present  until  after  formation  of  all  fins  in  the 
belonids,  hemiramphids  and  scomberesocids  (Fig.  I77B-E).  The 
situation  in  the  Exocoetidae  is  not  clear.  Most  published  illus- 
trations of  exocoetids  do  not  show  a  preanal  finfold.  Ones  that 


B 


Fig.  174.  Belonidae  eggs.  (A)  Slrongylura  exilis.  LACM  43475-1; 
(B)  Slrongylura  slrongylura;  (C)  Tylosurus  acus.  (From:  A.  Original.  B. 
Job  and  Jones,  1938.  C.  Mito.  1958.) 


Fig.  175.  Hemiramphidae  eggs.  (A)  Rhynchorhamphus  georgii;  (B) 
Oxyporhamphus  muroplerus  micropterus;  (C)  Hemiramphus  margin- 
alus.  (From:  A.  Kovalevskaya,  1965.  B.  Imai.  1959.  C.  Taiwan  1968.) 


COLLETTE  ET  AL.:  BELONIFORMES 


341 


do  (Evans,  1962— Hirundichthys  affinis;  Vijayarghavan,  1973  — 
H.  coromandelensis;  Vijayaraghavan,  1975  — Cypselurus  spilop- 
terus  and  Kovalevskaya,  1965  — Cheilopogon  katoptron)  are  of 
embryos  or  newly  hatched  larvae  and,  except  for  Kovalevskaya 
(1965),  were  hatched  in  the  laboratory  (Fig.  177H).  In  these 
examples  the  preanal  finfold  was  small  and  soon  lost.  We  have 
examined  field-collected  yolk-sac  Cheilopogon  (pTes\imab\y  Ch. 
pinnatibarbatus  californicus)  without  finding  a  preanal  finfold 
(Fig.  177G).  Perhaps  some  exocoetids  have  a  preanal  finfold, 
but  lose  it  soon  after  hatching.  If  so,  most  field-collected  spec- 
imens may  have  already  lost  the  preanal  finfold  by  the  minimum 
sizes  typically  illustrated. 

Fin  formation  generally  begins  during  the  embryonic  stages 
or  soon  after  hatching.  In  fact,  flexion  of  the  caudal  fin  precedes 
hatching  in  flyingfishes  (Ahlstrom  and  Moser,  1980).  In  the 
scomberesocids,  belonids  and  hemiramphids,  caudal,  dorsal  and 
anal  fins  generally  form  first  followed  by  the  pectorals  and  lastly 
the  pelvics.  Pectoral  and  pelvic  buds  as  well  as  dorsal  and  anal 
anlagen  are  typically  present  at  hatching  in  exocoetids.  Pectoral 
fins  form  last  in  exocoetids,  after  the  pelvic  fins. 

Belonids,  scomberesocids  and  exocoetids  generally  hatch  with 
heavy,  uniform  pigmentation  formed  or  forming  over  essentially 
the  entire  body  (Fig.  1 778,  C,  and  G).  Exceptions  are  the  fresh- 
water needlefish  Xenentodon  cancila.  which  has  9-10  saddle- 
shaped  dorsal  aggregations  plus  a  ventrolateral  stripe  (Fig.  1  77D) 
and  some  exocoetids  of  the  genera  Parexocoetus  and  Cheilo- 
pogon. which  have  patterns  somewhat  reminiscent  of  the  hem- 
iramphids (compare  Fig.  177E  and  H).  This  pattern  consists  of 
three  rows  of  melanophores  on  each  side  of  the  body,  one  dorsal, 
one  lateral  and  one  ventral.  Two  hemiramphids  reported  to  be 
exceptions  to  this  are  Hyporhamphus  quoyi  and  Hemiramphus 
marginatus.  These  species  hatch  with  pigment  over  the  entire 
body;  a  pattern  reminiscent  of  most  other  beloniforms.  The 
pigment  pattern  in  adrianichthyids  resembles  that  in  hemiram- 
phids except  dorsally  where  the  adrianichthyids  have  a  single 
middorsal  row  of  melanophores  (Fig.  181  A),  similar  to  the  con- 
dition observed  in  Atheriniformes  (see  White  et  al.,  this  volume) 
rather  than  the  double  row  typical  of  most  hemiramphids  (Fig. 
177F). 

Specialized  Ontogenetic  Stages 
During  post-embryonic  development,  beloniform  fishes 
undergo  a  number  of  complex  changes.Their  larvae  differ  fairly 
strongly  from  juveniles  and  the  juveniles  are  frequently  unlike 
adults.  Juveniles  of  related  species  frequently  differ  more  from 
each  other  than  do  larvae  or  adults.  In  this  section,  notable 
ontogenetic  changes  are  described  for  several  character  suites 
in  the  four  families  of  the  Exocoetoidei.  Adrianichthyoids  lack 
specialized  ontogenetic  stages. 

Jaws,  beaks,  and  barbels 

Scomberesocidae.— Juveniles  (20-40  mm  SL)  have  slightly 
elongate  upper  and  lower  jaws  but  no  prominent  beaks  (Fig. 
18 IB;  Hubbs  and  Wisner,  1980:  fig.  a).  At  about  60  mm 
SL,  both  upper  and  lower  jaws,  but  especially  the  lower  jaw, 
elongate  in  Scomberesox  and  Namchthys.  Elongation  continues 
in  both  taxa  to  100-120  mm  SL.  Both  jaws  elongate  almost 
equally  in  Scomberesox;  the  lower  jaw  exceeds  the  upper  in 


Fig.  176.  Exocoetidae  eggs.  (A)  Exocoetus  volilansAM)  t'odiator  acu- 
lus  pacificus:  (C)  Hirundichthys  coromandelensis.  (From:  A.  Parin  and 
Gorbunova,  1964.  B.  Breder,  1938.  C.  Vijayaraghavan,  1973.) 


342 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES- AHLSTROM  SYMPOSIUM 


Fig.  177.  Beloniform  larvae.  (A)  Adrianichthyidae:  Oryzias  melastigma.  4.3  mm;  (B)  Scomberesocidae:  Cololabis  saira.  SWFC  5009-50. 1 10, 
5.1  mm  SL;  (C-D)  Belonidae;  (C)  Strongylura  exilis.  LACM  42756-5,  8.6  mm  SL;  (D)  Xenemodon  cancila.  9.6  mm  SL  ANSP  124230:  (E-F) 
Hemiramphidae:  (E-F)  Hyporhamphus  rosae.  5.7  mm  SL,  LACM  42870-5;  (G-H)  Exocoetidae:  (G)  Cheilopogon  pinnatibarbatus  califonucus. 
LACM  IP77-3,  3.7  mm  SL;  (H)  Cheilopogon  katoptron.  3.2  mm  SL.  From:  (A)  Job,  1940;  (B-G)  Original:  (H)  Kovalevskaya,  1965. 


Nanichthys.  A  slight  beak  develops  in  Cololabis;  Elassichthys 
does  not  develop  a  distinct  beak. 

Belonidae.  — Most  species  of  Belonidae  pass  through  a  "half- 
beak"  stage  in  which  the  lower  jaw,  but  not  the  upper  jaw,  is 
greatly  elongate.  Juveniles  of  Belone  belone  remain  in  the  half- 
beak  stage  for  a  longer  time  than  other  needlefishes.  This  has 
led  directly  to  four  synonyms  of  Belone  belone  described  as 
halfbeaks  (CoUette  and  Farin,  1 970: 1 6-17).  Plotting  the  relative 
length  of  the  lower  jaw  extension,  as  a  percentage  of  head  length 
against  body  length  (Fig.  1 78),  shows  that  lower  jaw  extension 
in  B.  belone  may  be  nearly  150%  of  head  length  at  25  mm  BL 
(body  length)  and  decreases  to  less  than  10%  by  175  mm  BL. 
Petalichthys  and  Platybelone  (Fig.  179E)  also  remain  in  the 
halfbeak  stage  for  a  long  time.  The  duration  of  the  halfbeak 
stage  varies  among  species  of  Strongylura  (Fig.  1 79C  and  F). 


Comparative  development  of  Platybelone  {as  Strongylura  long- 
leyi),  Strongylura  marina,  S.  nolata,  and  two  species  of  Tyto- 
surus,  (T.  acus  and  T.  crocodilus.  Fig.  179G  and  J)  was  illus- 
trated by  Breder  ( 1 932:  figs.  7  and  1 0,  plates  1  and  2).  Tylosurus 
crocodilus  (Fig.  1 79J)  completely  lacks  a  halfbeak  stage,  upper 
and  lower  jaws  growing  at  the  same  rate  from  larval  to  adult 
stages  of  development  (Breder,  1932:  plate  2,  fig.  2,  as  T.  ra- 
phidoma).  The  South  American  freshwater  genus  Belonion 
(maximum  size  42  mm  body  length)  is  characterized  by  ma- 
turing while  still  in  the  halfbeak  stage  (Fig.  1  79A-B)  and  was 
considered  paedomorphic  by  Collette  (1966). 

Hemiramphidae.— Adults  of  four  genera  of  halfbeaks  lack  the 
elongate  lower  jaw  that  characterizes  most  members  of  the  fam- 
ily. The  lower  jaw  extends  only  1.5-1 1.0  mm  beyond  the  upper 
jaw  throughout  the  size  range  in  Arrhamphus  (Collette,  1974b). 


COLLETTE  ET  AL.:  BELONIFORMES 


343 


g 

CO 

z 

UJ 

t- 
X 
UJ 

< 

— ) 
a: 


I 
»- 
O 

z 

UJ 

_j 

Q 
< 

LU 

I 


150 


100 


50 


-..^^^^ 


T~^^^s#^-^^ 


•  i  .•%  • 


•• 


25 


100 


200 


300 


400 


500 


BODY  LENGTH  (mm) 


Fig.  178.     Relative  growth  of  upper  jaw  in  Belone  belone.  Lower  jaw  extension  as  a  percent  of  head  length  plotted  against  body  length.  Inset 
is  of  a  43.7  mm  BL  B.  belone  from  Ireland  in  the  "halfbeak"  stage.  (From:  Collette  and  Parin.  1970.) 


The  lower  jaw  is  even  shorter  in  Melapedalion  and  virtually 
absent  in  adult  Chriodorus  and  Oxyporhamphus.  Chriodorus 
looks  superficially  more  like  an  atherinid  than  a  halfbeak,  hence 
its  specific  name  of  at herinoides.  Adult  Oxyporhamphus  resem- 
ble flyingfishes  because  of  the  enlarged  pectoral  fins.  Juveniles 
of  all  four  genera  have  a  distinct  beak.  Arrhamphus,  Melape- 
dalion. and  Chriodorus  have  always  been  considered  halfbeaks. 
Oxyporhamphus  has  usually  been  considered  an  exocoetid  or 
placed  in  a  separate  family.'  Even  with  its  short  beak,  Arrham- 
phus varies  geographically  in  beak  length:  Arrhamphus  s.  scler- 
olepis  of  northern  Australia  has  a  proportionately  shorter  lower 
jaw  (up  to  20  times  in  head  length)  than  does  A.  sclerolepis 
krefflii  of  southern  Queensland  and  New  South  Wales  (up  to  1 1 
times  in  head  length,  see  Collette,  1974b:  fig.  4). 

Exocoetidae.— The  two  most  primitive  genera  of  flyingfishes, 
Fodiator  and  Parexocoetus,  have  an  elongate  lower  jaw  (Parin, 


'  Parin  (1961),  although  still  recognizing  the  Oxyporhamphidae  as 
valid,  clearly  showed  that  Oxyporhamphus  is  a  halfbeak,  even  though 
it  has  a  straight  margin  to  the  upper  jaw  instead  of  triangular  as  in  other 
halfbeaks.  Two  developmental  characters  support  placement  of  O.vr- 
porhamphus  in  the  Hemiramphidae:  a  preanal  fin  fold  is  present  in 
larvae  (absent  or  lost  soon  after  hatching  in  Exocoetidae)  and  the  pelvic 
fins  form  last  (pectoral  fins  form  last  in  Exocoetidae). 


1961;  Kovalevskaya,  1982).  This  clearly  is  a  beak  m  juvenile 
(15-55  mm  SL)  Fodiator,  which  like  several  genera  of  halfbeaks, 
lose  their  beaks  as  they  grow  larger  (Fig.  1 8 1 C  and  Breder,  1938: 
figs.  5  and  6E).  A  beak  is  present  in  Parexocoetus  mento  (Imai, 
1959).  Small  ( 1 9-20  mm)  P.  hrachypterus  have  a  pair  of  barbels 
that  are  attached  to  the  ventral  surface  of  the  beak  and  obscure 
it  (Fig.  182).  Thus,  a  beak  which  is  absent  in  advanced  flying- 
fishes, is  present  in  both  primitive  genera. 

Juvenile  stages  of  many  exocoetids  develop  barbels  on  the 
lower  jaw  (Table  91).  Barbels  range  from  relatively  short  to 
longer  than  body  length  (Fig.  181D-I).  Parexocoetus  mento 
does  not  develop  a  barbel  nor  do  species  of  Prognichthys  and 
Hirundichthys  (Kovalevskaya,  1982).  Paired  barbels  develop  in 
Parexocoetus  hrachypterus  and  in  all  species  of  Cheilopogon 
(Fig.  181D,  G-I;  Kovalevskaya,  1982).  In  species  of  Cheilo- 
pogon (subgenus  Procypselurus,  Ch.  nigricans  group),  the  bar- 
bels consist  of  a  thick  strand  with  a  leathery  fold  branching  oflT 
it  in  the  form  of  a  lobe  (Parin,  1961;  Kovalevskaya,  1982).  In 
small  specimens  of  Ch.  cyanopterus  the  barbel  may  be  complex 
and  have  2-3  flaps.  Members  of  Cheilopogon  (subgenus  Mac- 
ulocoetus)  have  flattened  barbels,  joined  together  at  the  base. 
These  may  be  large.  The  barbels  of  C/!«7opo^o«  (subgenus  Pten- 
ichthys)  range  from  short  (in  Ch.  heterurus  doederleini)  to  long 
(in  Ch.  unicolor).  The  barbels  in  Cheilopogon  pinnatibarbatus 


344 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


T^#^"^ 


.m^ 

._^  ^  I  1,  II  I  ijii,.,i.iiijjui:(xc.  ■" 


-'^-^-■:'^^^^^ 


<.  .  ^S%l 


COLLETTE  ET  AL.:  BELONIFORMES 


345 


(subgenus  Cheilopogon  s.  str.)  are  flaplike  and  fringed  (Fig.  1811). 
Kovalevskaya  (1982)  considered  this  to  result  from  the  fusion 
of  paired  barbels  and  our  examination  of  Ch.  pmnatibarbatus 
calijorniciis  supports  this.  The  barbel  is  single  in  Cypselurus 
(subgenus  Cypselurus  s.  str.)  and  Exocoetus  monocirrhus  (Fig. 
1 8 1 F;  Parin,  1961;  Kovalevskaya,  1982).  Some  species  of  Cyp- 
selurus (subgenus  Poecilocypselurus,  Cy.  poecilopterus  and  Cy. 
starksi)  do  not  develop  a  barbel,  nor  do  the  remaining  two 
species  of  Exocoetus,  E.  obtusirostris  and  E.  volilans. 

Melanistic  dorsal  fin  lobe 

Pelagic  members  of  three  families  (all  except  Scomberesoci- 
dae)  develop  prominent  melanistic  lobes  in  the  dorsal  fin.  The 
lobe  is  in  the  posterior  part  of  the  dorsal  fin  in  the  Belonidae 
and  Hemiramphidae  but  in  the  middle  of  the  fin  in  the  Exo- 
coetidae  so  presence  of  the  lobe  is  not  necessarily  homologous. 

Beionidae.— A blennes  and  Tylosurus  are  characterized  by  hav- 
ing a  prominent  enlarged  melanistic  lobe  in  the  posterior  part 
of  the  dorsal  fin  (Fig.  179D,  G-J).  Other  genera  of  needlefishes 
(Fig.  1 79)  lack  any  trace  of  this  posterior  dorsal  lobe.  Breder 
(1932:  plates  3-5)  illustrated  the  development  of  this  posterior 
lobe  in  T.  acus  and  T.  crocodilus  and  its  absence  in  Strongylura 
and  Platybelone.  Parin  (1967)  left  an  Australian  species  difficult 
to  place  in  either  Tylosurus  or  Strongylura  in  a  monotypic  genus 
described  by  Whitley,  Lhotskia  gavialoides.  A  juvenile  with  a 
well-developed  posterior  dorsal  lobe,  captured  by  Collette.  con- 
vinces us  that  it  is  a  species  of  Tylosurus  (Fig.  1 79H).  The  lobe 
is  apparently  sloughed  off  in  Tylosurus  crocodilus  (Breder  and 
Rasquin,  1952),  resorbedin  T.  aoM  (Breder  and  Rasquin.  1954), 
and  retained  in  adult  Ablennes. 

Hemiramphidae.— Juveniles  of  Hemiramphus  and  Oxypo- 
rhamphus  develop  a  darkened  posterior  lobe  on  the  dorsal  fin 
(Fig.  180)  similar  to  that  present  in  two  genera  of  needlefishes, 
Ablennes  and  Tylosurus. 

Exocoetidae.  — In  juveniles  of  many  species  oi  Cheilopogon.  the 
middle  portion  of  the  dorsal  fin  develops  a  melanistic  lobe  (Fig. 
1 8 1 H).  This  is  reminiscent  of  the  adult  stage  of  Parexocoetus 
and  Eodiator  acutus. 

Body  bars 

Juveniles  of  some  species  in  three  exocoetoid  families  (all 
except  Scomberesocidae)  have  vertical  bars  on  their  body. 

Belonidae.— Juveniles  of  two  species  of  Tylosurus,  T.  gavi- 
aloides (Fig.  179H)  and  T.  acus  (see  Collette  and  Parin.  1970: 
fig.  12)  and  .Ablennes  hians  have  bars.  These  bars  are  retamed 
in  adult  .-iblennes  as  is  the  posterior  dorsal  fin  lobe. 


Hemiramphidae.  — The  10  species  of  the  genus  Hemiramphus 
all  have  a  series  of  broad  vertical  bars  on  the  body  (Fig.  180A- 
E)  at  some  stage  of  their  development.  Body  bars  are  retained 
for  different  periods  of  time  during  development:  all  body  bars 
are  lost  before  105  mm  SL  in  He.  lutkei  and  He.  depauperatus 
(Parin  et  al.,  1980:  fig.  32),  before  120  mm  SL  in  He.  bermu- 
densis  and  He.  brasiliensis  (Collette,  1962:  fig.  1),  but  are  re- 
tained past  175  mm  SL  in  He.  balao;  one  blotch  is  retained 
throughout  life  in  He.  robustus,  and  all  are  retained  in  He.  far. 

Pelvic  fin  pigment 

All  10  species  of  Hemiramphus  also  have  pigmented  pelvic 
fins  as  juveniles  (Fig.  183).  The  patterns  of  pelvic  fin  pigmen- 
tation divide  the  genus  into  two  species  groups,  one  with  pig- 
mentation concentrated  proximally  on  the  fin  (balao  group.  Fig. 
183,  top  two  rows),  the  other  with  pigment  absent  basally  and 
concentrated  distally  (J'ar-brasiliensis  group.  Fig.  183.  bottom 
row).  Body  bars  and  pelvic  fin  pigmentation  are  absent  in  Hy- 
porhamphus. 

Exocoetidae.  — In  late  larval  and  juvenile  stages  of  many  flying- 
fishes,  Exocoetus,  Cheilopogon  (at  least  some  species  in  all  sub- 
genera except  possibly  Paracypselurus,  for  which  we  lack  data), 
Cypselurus  (subgenus  Poecilocypselurus— see  Imai,  1959),  and 
Hirundichthys  oxycephalus  (Imai,  1960)  transverse  stripes  de- 
velop on  the  abdomen  and  sides  of  the  body  which  disappear 
(sometimes  leaving  traces)  in  adults.  The  coloration  of  the  larvae 
and  particularly  of  the  juveniles  of  flyingfishes  is  diverse,  and, 
as  a  rule,  differs  greatly  from  the  coloration  of  adults.  A  partic- 
ularly bright  variegated  coloration  is  characteristic  of  young  of 
neritic  species  living  among  algae  (Parin,  1961;  Kovalevskaya, 
1982). 

Relationships 

Beloniformes 

The  Beloniformes  were  defined  by  7  characters  by  Rosen  and 
Parenti  (1981:16).  Meristic  characters  for  the  beloniform  genera 
are  summarized  in  Table  92.  A  cladogram  for  the  families  and 
higher  taxa  of  the  Beloniformes  is  presented  as  Fig.  184. 

Adrianichthyoidei 

Rosen  and  Parenti  (1981)  defined  the  adrianichthyoids  by  5 
characters.  Larval  adrianichthyids  also  differ  from  exocoetoids 
in  having  a  shorter  preanal  distance,  40-50%  of  standard  length. 
Rosen  and  Parenti  (1981)  included  the  Horaichthyidae  and  Ory- 
ziidae  in  the  Adrianichthyidae.  By  this  definition  the  Adrianich- 
thyidae  includes  four  genera,  .Adrianichthys,  Horaichthys,  Ory- 
:ias  and  Xenopoecihis  with  a  total  of  1 1  species  (Nelson,  1976). 
These  fishes  inhabit  fresh  and/or  brackish  waters  from  India 
and  Japan  to  the  Indo-Australian  Archipelago. 


Fig.  179.     Halfbeak  stages  of  Belonidae,  arranged  by  relative  length  of  upper  Jaw.  (A)  Belonion  apodion  Collelle:  USNM  199540;  Brazil,  Borba; 

29.4  mm  BL;  (B)  Belonion  dihranchodon  Collette;  USNM  199463;  Venezuela,  Rio  Atabapo;  38.2  mm  BL;  (C)  Strongylura  marina  (Walbaum), 
USNM  189006;  Nicaragua;  23.5  mm  BL;  (D)  Ablennes  hians  (Valenciennes);  USNM  188843;  Gulf  of  Honduras;  36.1  mm  BL;  (E)  Platybelone 
argatus  argalus  (Le  Sueur)  USNM  198102;  39°28'N,  69°30'W;  96  mm  BL;  (F)  Strongylura  p.v;fa  (Girard);  SIO  H47-158-23A;  Calif,  La  Jolla; 

72.5  mm  BL;  (G)  Tylosurus  acus  acus  (Ucepede);  USNM  1 98402;  38°00'N,  65''25'W;  1 30  mm  BL;  (H)  Tylosurus  gavialoides  (Castelnau);  USNM 
226666;  Australia,  New  South  Wales;  72.5  mm  BL;  (1)  T.  choram  (Forsskal);  USNM  147438;  Red  Sea;  95.0  mm  BL;  (J)  T.  c.  crocodilus  (Peron 
and  Le  Sueur);  USNM  198407;  37°08'N,  66°14'W;  96.3  mm  BL.  A-G,  1-J  drawn  by  Mildred  H.  Carrington;  H  by  Keiko  Hiratsuka  Moore;  A- 
C  from  Collette  (1966:  tig.  1). 


346 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  180.  Juvenile  banded  stages  of  five  species  of  Hemiramphus  and  Oxyporhamphus  micropterus.  (A)  Hemiramphus  balao  (Le  Sueur); 
USNM  200592;  off  Cape  Hatteras,  North  Carolina;  53.7  mm  SL;  (B)  He  saltator  (GWben  and  Starks);  SIO  55-247;  Gulf  of  Panama;  49.5  mm 
SL;  (C)  He.  depauperatus  Lay  and  Bennett;  Hawaii  Inst.  Mar.  Biol.;  Hawaiian  Is.;  41.4  mm  SL;  (D)  He.  far  (ForsskSl);  USNM  148020;  Persian 
Gulf;  47.0  mm  SL;  (E)  He.  hrasiliensis  (Linneaus);  USNM  188748;  off  North  Carolina;  50.0  mm  SL;  (F)  Oxyporhamphus  micropterus  similis 
Bruun;  USNM  159032;  Gulf  of  Mexico;  41.2  mm  SL.  A-B,  D-F  drawn  by  Mildred  H.  Camngton;  C  by  Keiko  Hiratsuka  Moore. 


COLLETTE  ET  AL.:  BELONIFORMES 


347 


Table  91.    Larvae  of  Beloniform  Fishes.  Much  of  this  information  is  based  entirely  on  illustrations  from  the  cited  references. 


Hatching 
length  (mm) 


Adrianichthyidae 
Horaichthys  selnai 
Oryzias  melasligma 

Scomberesocidae 

Cololabis  saira 

C.  adocetus 
Scomheresox  saurus 
S.  simulans 

Belonidae 

Abennes  hians 
Belone  belone 
Platybelone  argalus 
Strongylura  manna 
S.  strongylura 

Tylosurus  acus 
T.  crocodilus 
Xenentodon  cancila 

Hemiramphidae 
Dermogenys  pusillus 
Hemiramphus  hrasiliensis 

He.  marginatus 

Hemirhamphodon  pogonognathus 
Hyporhamphus  guoyi 
Hy.  limbatus 

Hy.  intermedius  lutkei? 
Hy.  sajori 

Oxyporhamphus  convexus 
O.  microplerus  microplerus 


Exocoetidae 
Cheilopogon  (A.)  agoo 

Ch.  (Ch.)  pmnalibarbalus 

californicus 
Ch.  (Ch.)  pinnatibarbatus 

japonicus 
Ch.  (M.)  spilonotopleriLs 


None 


3.5-4.0 

4.0-4.5 

Slight  or  none 

None 

6.0-8.5 

Present  by  15-17  mm 

Present  by  40  mm 

Present  bv  13.9  mm 

9.0 

Present  bv  1 8  mm 

Present  by  47  mm 

9.2-14.4 

Present  by  14.3  mm 

6.75 

Present  by  fifth 

day 

10.16 

Present  by  14.1  mm 

10.7-12.0 

Present  by  15.2  mm 

10.5 

Present 

-11 

5-7  SL 

Present  by  13.0  mm 

5.85 

-11 

-6.3 

Present  at  hatching 

Present  by  1 2.0  mm 

4.8 

Present  by  10.7  mm 

Present  by  12.3  mm 

Present  by  14.7  mm 

7.7 

Present  by  6-8  mm 

Ch.  (A/.)  spiloplcrus 
Ch.  (M.)  sulloni 

Ch.  (Proc.)  cyanoplenis 


Ch.  (Proc.)  exsiliens 

Ch.  (Proc.)  katoptron 
Ch.  (Proc.)  nigricans 
Ch.  (Pt.)furcalus 


Ch.  (Pt.)  helcrunis 
Ch.  (Pt.)  h.  doederleini 

Ch.  (Pt.)  umcolor 

Cypselurus  (Cy.)  comatus 
Cy.  (Cy.)  naresii 


4.5-5.3 

None 

4.1-4.8 

None 

:an  =  4.45 

4.6-5.8 

None 

None 

4.52 

None 

None 

4.5? 


5.2-6.1 


None 


None 

None 
None 
None 


None 
None 

None 

None 
None 


None 


None 


None 


None 


Pair,  short,  present  by 

19.0  mm 

Fan-like,  complex  with  14 

fimbriae 
Fan-like  with  flaps,  present 

by  20.1  mm 
Pair,  very  long,  present 

by  10.4  mm 

Pair,  flattened  and 

joined  at  the  base 
Pair,  complex  on  smaller 

individuals,  then  long 

filaments 
Pair,  medium,  present  by 

14.5  mm 
Short,  present  by  10.2  mm 
Pair,  long,  complex 
Pair,  medium  length,  develop 

on  individuals  between 

7.7  and  18  mm 

Pair,  present  by  19  mm  TL 
Pair,  short,  present  by 

18.1  mm 

Pair,  long,  present  by 

5.8  mm 

Single,  medium  length 
Single  tape-like,  very  long 

with  appendages  at  the 

base 


Kulkami,  1940 
Job,  1940 


Hubbs  and  Wisner,  1980  and 

original 
Hubbs  and  Wisner,  1980 
Hardy,  1978a;  Fahay,  1983 
Hubbs  and  Wisner,  1980 


Mito,  1966 

Russell,  1976 

Original 

Hardy,  1978a 

Job  and  Jones,  1938 

Hardy,  1978a 
Masurekar,  1968 
Foster,  1973 

Soong,  1968 

Hardy,  1978a;  Berkeley 

and  Houde,  1978 
Talwar,  1968 
Soong,  1968 
Sudarsan,  1968b 
Nair,  1952b;  Job  and 

Jones,  1938 
Uchidaet  al.,  1958 
Uchidaet  al.,  1958 
Kovalevskaya,  1965 
Chrapkova-Kovalevskaya, 

1963;  Kovalevskaya, 

1965c 

Imai,  1960 

Hubbs  and  Kampa,  1946 

Imai,  1959 

Imai,  1959 

Vijayaraghavan,  1975 
Kovalevskaya,  1982 

Breder,  1938;  original 

Imai,  1959 

Kovalevskaya,  1965 
Kovalevskaya,  1982 
Hildebrand  and  Cable, 
1930 

Hardy,  1978a 
Imai,  1959 

Gorbunova  and  Parin, 

1963 
Breder,  1938 
Imai,  1959 


348 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  91.    Continued. 


Halching 
length  (mm) 


Cy.  (Cy.)  opisthops  hiraii 

Cy.  (Poec.)  poecilopterus 
Cy.  {Poec.)  slarksi 
Exocoetus  monocirrhus 


E.  ohlusirostris 
E.  volitans 
Fodiator  aculus 
Hirundichthys  (D.) 

albtmaculatus 
H.  (D.)  marginatus 
H.  (D)  rondeteti 
H.  (//.)  affinis 
H.  (H.)  coromandelensis 


H.  (//.)  o.xycephalus 
H.  (H.)  speculiger 
Parexocoelus  brachyplenis 

brachyplerus 
P.  mcnto  memo 

Prognichlhys  gibbifrons 
P.  sealei 


4.5-5.1 

None 

Smgle,  petalous 
by  17.5  mm 

present 

Imai,  1959 

None 

Absent 

Imai,  1959 

5.2 

None 

Absent 

Imai,  1959 

None 

Single,  long 

dev 

elops  on 

Kovalevskaya, 

1964;  Imai. 

individuals  between  16.0 

1959 

and  18.6 

•nm 

None 

Absent 

Kovalevskaya, 

1964 

None 

Absent 

Kovalevskaya, 

1964 

Present 

by 

14.6 

mm 

Absent 

Original 

None 

Absent 

Kovalevskaya, 

1972 

None 

Absent 

Kovalevskaya, 

1972 

None 

Absent 

Imai,  1960 

None 

Absent 

Evans,  1962;  Breder,  1938 

3.47-4.23 

None 

Absent 

Vijayaraghavan,  1973; 
Kovalevskaya,  1972 

None 

Absent 

Imai,  1960 

None 

Absent 

Imai,  1960 

Present  by 

18.1 

mm 

Pair,  short 

Imai,  1959 

4.5-5.2 

Present  by 

23.8 

mm 

Absent 

Tsukahara  and  Shiokawa, 

1957;  Imai, 

1959 

None 

Absent 

Origmal 

None 

Absent 

Imai,  1960 

Exocoetoidei 

Defined  by  six  characters  by  Rosen  and  Parenti  (1981:  16). 
We  here  add  two  developmental  characters:  oil  droplets  in  egg 
minute  or  absent  and  preanal  distance  of  larvae  increased  to 
about  66%  of  standard  length. 

The  Exocoetoidei  is  undoubtedly  a  monophyletic  group. 
However,  vainous  opinions  have  been  expressed  as  to  which 
group  within  the  suborder  is  the  most  primitive.  A  number  of 
authors  have  considered  the  Hemiramphidae  to  be  the  most 
generalized  family  based  largely  on  the  fact  that  needlefishes 
and  primitive  flying  fishes  {Fodiator)  pass  through  an  ontoge- 
netic halfbeak  stage  during  development.  Parin  (1961)  and  Ro- 
sen ( 1 964)  supported  this  viewpoint.  On  the  other  hand,  Nichols 
and  Breder  (1928)  and  Breder  (1932)  considered  the  family 
Belonidae  the  most  primitive.  To  resolve  the  directionality  of 
the  "halfbeak"  stage  (suite  four),  three  additional  character  suites, 
each  suite  consisting  of  several  correlated  transformation  series, 
were  considered.  Apomorphic  character  states  are  numbered 
higher  than  plesiomorphic  states  on  Fig.  184. 

The  first  suite  involves  pharyngeal  tooth  plate  fusion,  trans- 
formation series  A-B.  State  Al  is  close  opposition  of  left  and 
right  fifth  ceratobranchial  tooth  plates  characteristic  of  more 
primitive  Atherinomorpha  and  the  Adrianichthyoidei.  State  A2 
is  the  fusion  of  left  and  right  lower  pharyngeal  bones  into  a  tooth 


plate  in  the  Exocoetoidei.  Series  B  state  1  is  when  the  third 
upper  pharyngeals  are  separated  by  a  gap.  State  B2  is  when  they 
are  joined  but  not  fused  in  the  Exocoetoidea.  State  B3  is  the 
complete  fusion  of  the  third  upper  pharyngeals  into  a  tooth  plate 
in  the  Hemiramphidae. 

The  second  suite  involves  loss  of  gill  arch  skeleton  bones, 
transformation  series  C-D.  State  CI  is  presence  of  the  fourth 
epibranchial,  C2  its  loss  in  the  Beloniformes.  State  Dl  is  the 
presence  of  the  fourth  upper  pharyngeal  tooth  plates,  D2  their 
loss  in  the  superfamily  Exocoetoidea. 

The  third  suite  involves  reduction  in  the  cephalic  lateralis 
system,  transformation  series  E-F  (data  from  Parin  and  Astak- 
hov,  1982).  The  cephalic  system  is  more  complete  in  the  Scom- 
beresocoidea  than  in  the  Exocoetoidea,  including  the  presence 
of  a  premaxillary  canal  (EI),  an  autapomorphy  unique  among 
teleosts.  The  pre-,  supra-,  and  post-orbital  system  is  continuous 
across  the  top  of  the  head  in  state  1 .  There  are  short  interruptions 
in  the  system  in  state  2  in  the  Belonidae.  The  postorbital  section 
is  lost  in  state  3  and  secondary  bony  canals  are  lost  in  state  4, 
both  charactenstic  of  the  superfamily  Exocoetoidea. 

We  now  return  to  the  fourth  suite  of  transformation  series 
and  resolve  the  directionality  of  the  "halfbeak"  stage.  The  fourth 
suite  includes  elongation  of  upper  and  lower  jaws  and  presence 
of  barbels  in  juveniles,  transformation  series  G-I.  State  Gl  is 


Fig.  181.  Late  larval  and  early  juvenile  stages  of  beloniform  fishes.  (A)  Adrianichthyidae:  Oryzias  melastigma,  12  mm;  (B)  Scomberesocidae: 
Cololabis  adocetus.  SWFC  7205  J-20.145,  25  mm  SL;  (C-I)  Exocoetidae:  (C)  Fodiator  acutus  pacificus.  SWFC  FB-62-242,  15.4  mm  SL;  (D) 
Cheitopogon  unicolor.  18.5  mm  SL;  (E)  Cypselurus  comatus.  25  mm  SL;  (F)  Exocoetus  monocirrhus.  SWFC  FB-62-203,  27  mm  SL;  (G)  Pare.xocoetus 
brachyplerus,  43.5  mm  SL;  (H)  Cheitopogon  cyanopterus.  54  mm  SL;  (I)  Cheitopogon  pinnatibarbatus japomcus.  80  mm  SL.  From:  (A)  Job,  1940; 
(B,  C,  and  F)  Original;  (D)  Gorbunova  and  Parin,  1963;  (E,  G  and  H)  Breder,  1938;  (F)  Kovalevskaya,  1964;  and  (I)  Abe,  1954. 


COLLETTE  ET  AL.:  BELONIFORMES 


349 


350 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


■aeur' 


/iecir- 


Fig.  182.     Exocoetidae,  Parexocoetusbrachypterusbrachypterus.  (A) 
18.1  mm;  (B)  19.7  mm. 


lower  jaw  elongate  in  juveniles  and  adults,  G2  elongate  only  in 
juveniles,  and  G3  never  elongate,  even  in  juveniles.  Presence 
of  an  elongate  lower  jaw  is  considered  a  synapomorphy  of  the 
suborder  Exocoetoidei  because  it  is  present  in  the  most  gener- 
alized members  of  each  of  the  four  families.  This  is  supported 
ontogenetically  by  its  presence  in  juveniles  and  loss  in  adults  of 
four  genera  of  Hemiramphidae  and  in  the  two  least  derived 
subfamilies  of  Exocoetidae. 

Transformation  series  H  involves  elongation  of  the  upper  jaw. 
It  is  most  parsimonious  to  hypothesize  the  elongation  of  the 
upper  jaw  as  a  synapomorphy  (H2)  of  the  superfamily  Scom- 
beresocoidea.  Thus,  the  absence  of  an  elongate  upper  jaw  is 
plesiomorphous  (HI)  in  the  Exocoetoidea. 

Transformation  series  I  is  the  development  of  barbels  in  ju- 
veniles of  advanced  flyingfishes.  State  1 1  is  the  absence  of  bar- 
bels. If  we  consider  barbels  in  flyingfishes  to  be  derived  from 
the  pair  of  cutaneous  lappets  on  the  lower  jaw  of  needlefishes, 
halfbeaks,  and  primitive  flyingfishes,  the  most  generalized  state 


of  this  character  is  the  presence  of  two  separate  barbels,  12  (Fig. 
185).  This  supposition  is  supported  ontogenetically  by  two  ju- 
venile Parexocoetus  brachypterus  brachypterus.  The  smaller  one 
(Fig.  182A,  18.1  mm)  has  a  short  beak  from  the  ventral  surface 
of  which  a  pair  of  small  barbels  develop  in  the  larger  one  (Fig. 
182B,  19.7  mm).  Fusion  into  a  single  barbel  (13)  and  secondary 
loss  of  the  barbels  (14)  are  more  derived  states.  Loss  of  the 
barbels  has  apparently  occurred  independently  in  the  three  most 
advanced  subfamilies  of  the  Exocoetidae. 


Scomberesocoidea 

The  superfamily  is  defined  by  two  derived  characters:  pres- 
ence of  a  premaxillary  canal,  unique  among  teleosts;  and  upper 
jaw  at  least  slightly  elongate.  Other  diagnostic  characters  in- 
clude: third  pair  of  upper  pharyngeal  bones  separate,  fourth 
upper  pharyngeal  bone  usually  present,  scales  on  body  small. 
The  Scomberesocoidea  differ  from  the  Exocoetoidea  in  four 
characters  of  the  acoustico-lateralis  system  (Parin  and  Astak- 
hov,  1982).  The  cephalic  system  is  more  complete  in  the  Scom- 
beresocoidea than  in  the  Exocoetoidea. 

Scomberesocidae 

Defined  by  one  derived  character:  dorsal  and  anal  fins  fol- 
lowed by  a  series  of  finlets.  Other  diagnostic  characters  include: 
upper  and  lower  jaws  only  slightly  elongate,  teeth  small;  pectoral 
branch  of  lateral  line  absent,  posttemporal  simple.  There  are 
four  more  differences  in  the  acoustico-lateralis  system  between 
the  Scomberesocidae  and  the  Belonidae  (Parin  and  Astakhov, 
1982). 

Four  monotypic  genera  were  recognized  by  Hubbs  and  Wisner 
(1980):  Scomhereso.x  and  its  dwarf  derivative  Nantchthys.  and 
Cololabis  and  its  dwarf  derivative  Elassichthys.  All  sauries  are 
marine  holoepipelagic  fishes.  Scomberesox  reaches  the  largest 
size,  450  mm  SL,  Nanichthys  reaches  126  mm;  Cololabis  reaches 
350-400  mm,  Elassichthys  only  68  mm.  The  two  dwarf  taxa 
differ  convergently  from  Scomberesox  and  Cololabis  in  losing 
one  ovary  and  the  swimbladder  and  in  having  fewer  vertebrae, 
branchiostegal  rays,  pectoral  fin  rays,  and  gill  rakers.  Rather 
than  recognizing  four  monotypic  genera,  we  recognize  two  evo- 
lutionary lines  in  the  family  by  considering  Nanichthys  as  a 
synonym  of  Scomberesox  and  Elassichthys  a  synonym  of  Col- 
olabis as  previously  suggested  by  Parin  (1968). 

Belonidae 

Defined  by  one  derived  reductive  character:  interruptions  in 
the  cephalic  lateralis  system.  Other  diagnostic  characters  in- 
clude: no  finlets  following  dorsal  and  anal  fins;  both  upper  and 
lower  jaws  usually  elongate  and  studded  with  relatively  large 
sharp  teeth;  pectoral  branch  of  lateral  line  present;  posttemporal 
forked. 

The  Belonidae  contain  10  genera  and  32  species  (Collette, 
1966,  1974a,  1982a).  Four  genera  are  monotypic:  the  southern 
African  Petalichthys.  the  worldwide  Ablennes  and  Platybelone. 
and  the  Asian  freshwater  Xenentodon.  Belone  contains  two  east- 
em  Atlantic  species.  Three  genera  are  restricted  to  freshwaters 
of  South  America:  Pseudotylosiirus  (two  species),  Potamorrha- 
phis  (three),  and  Belonion  (two).  Tylosurus  contains  five  species 
of  strictly  marine  species;  Slrongylura  14  species,  some  marine, 
some  estuarine,  and  three  strictly  freshwater. 

The  genera  Belone  and  Petalichthys  appear  to  be  most  gen- 


COLLETTE  ET  AL.:  BELONIFORMES 


351 


<^^ 


352 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  92. 


Number  of  Dorsal,  Anal,  and  Pectoral  Fin  Rays,  Vertebrae  and  Gill  Rakers  on  the  First  Gill  Arch  in  the  Genera  of 

Beloniformes. 


No-  ofspp. 

D 

A 

p, 

Venebrae 

Family  and  genus 

Precaud, 

Caud- 

Total 

gr 

Adrianichthyidae 

Adnanichlhys 

1 

17 

25 

16 

Horaichthys 

1 

6-7 

22-32 

Oryzias 

7 

6-9 

17-25 

11-15 

29-32 

13 

Xenopoecilus 

2 

11-13 

21-27 

11-13 

Belonidae 

Belone 

2 

16-20 

19-23 

11-14 

48-54 

25-30 

75-84 

27-52 

Petatichthys 

1 

16-19 

21-23 

10-12 

46-47 

26-27 

72-74 

27-35 

Platybelone 

1 

11-17 

15-21 

10-13 

39-48 

23-29 

62-76 

7-14 

Tylosurus 

5 

18-27 

17-25 

11-15 

41-65 

23-33 

67-96 

0 

Ablennes 

1 

22-26 

24-29 

11-15 

51-63 

30-37 

82-97 

0 

Strong\'liira 

14 

12-23 

12-27 

9-13 

34-57 

19-34 

53-90 

0 

Xenentodon 

1 

14-19 

14-19 

10-12 

35-40 

21-25 

57-62 

0 

Pseudotylosurus 

2 

13-16 

14-19 

8-11 

42-47 

25-28 

67-73 

0 

Potamorrhaphis 

3 

27-43 

24-39 

6-8 

35-42 

28-44 

64-85 

0 

Belonion 

2 

11-14 

12-15 

5-6 

32-37 

19-23 

52-59 

0 

Scomberesocidae 

Scomberesox 

1 

15-18 

17-21 

12-15 

39-43 

24-28 

64-70 

34-51 

Nanichthys 

1 

14-16 

17-20 

10-11 

35-38 

22-26 

58-62 

19-26 

Cololabis 

1 

14-18 

18-21 

12-15 

37-40 

24-29 

62-69 

32-42 

Elassichthys 

1 

14-18 

16-21 

8-11 

32-35 

21-24 

54-59 

15-21 

Hemiramphidae 

Arrhamphus 

1 

13-16 

14-17 

12-14 

28-32 

16-19 

18-25 

45-50 

Chriodonts 

1 

15-18 

15-17 

12-14 

31-33 

18-19 

49-51 

20-24 

Melapedalion 

1 

15-17 

14-16 

12-13 

33-35 

17-18 

51-52 

25-31 

Hemiraniphus 

10 

11-15 

9-14 

10-13 

30-41 

16-19 

50-59 

25-48 

Rhyruiwrhamphiis 

4 

13-17 

12-16 

10-12 

37-40 

16-19 

54-59 

47-78 

Hyporhamphus  (Hyporhamphus) 

23 

12-17 

13-17 

10-13 

28-37 

16-20 

45-55 

20-53 

Hyporhamphiis  (Rcporhamphus) 

II 

13-18 

13-19 

10-13 

31-42 

15-20 

49-61 

26-47 

Oxyporhamphus 

2 

12-15 

13-17 

11-13 

30-33 

17-19 

47-50 

26-36 

Euleplorhamphus 

2 

20-25 

20-25 

7-9 

44-46 

26-29 

70-75 

24-35 

Zenarchoplerus 

17 

10-16 

8-14 

7-11 

25-36 

11-18 

38-51 

11-18 

Dennogenys 

3 

8-12 

14-17 

9-13 

21-24 

16-17 

38-40 

11-14 

Hemirhamphodon 

3 

14-23 

8-9 

8-9 

24-26 

14-16 

38-41 

11-16 

Nomorhamphus 

2 

12-13 

13-17 

11-13 

21-24 

17-19 

40-42 

0 

Exocoetidae 

Fodiator 

1 

9-11 

10-12 

13-14 

25-26 

14-16 

39-41 

29-33 

Parexocoetus 

2 

9-13 

10-14 

12-14 

21-25 

14-16 

36-40 

23-32 

Exocoetus 

3 

13-15 

12-15 

14-17 

24-27 

16-20 

42-44 

23-35 

Cypselurus 

11 

10-14 

7-10 

13-17 

28-20 

13-15 

39-48 

17-24 

Cheilopogon 

18 

9-16 

7-12 

13-17 

25-30 

12-16 

41-51 

19-28 

Prognichlhys 

4 

10-13 

8-10 

14-19 

26-34 

12-17 

43-45 

21-28 

Hirundichlhys 

8 

9-13 

9-13 

15-20 

26-31 

14-19 

42-47 

23-32 

eralized  (CoUette  and  Berry,  1965;  Parin,  1967),  having  well- 
developed  gill  rakers,  large  scales,  comparatively  weak  canine 
teeth  and  other  generalized  characters.  Belone  also  is  charac- 
terized by  the  most  completely  developed  cranial  lateralis  sys- 
tem (Parin  and  Astakhov,  1982).  Of  the  other  genera,  the  great- 
est number  of  primitive  characters  are  found  in  Platybelone. 
characterized  at  the  same  time  by  several  specialized  features 
(in  particular,  the  well-developed  cutaneous  lateral  keel  on  the 
caudal  peduncle).  The  remaining  three  genera  of  marine  needle- 
fishes (Strongylura,  Ablennes.  and  Tylosurus)  are  more  ad- 
vanced but  their  relationships  have  been  interpreted  differently 
by  Collette  and  Berry  (1965),  Parin  (1967),  and  Astakhov  ( 1 980). 
The  freshwater  genera  of  needlefishes,  in  the  opinion  of  all  au- 


thors, have  been  derived  from  the  genus  Strongylura  or  its 
ancestor  and  are  secondary  freshwater  fishes. 

Exocoetoidea 

The  superfamily  is  defined  by  one  derived  character  and  three 
losses:  third  pair  of  upper  pharyngeal  bones  united  into  a  plate; 
fourth  upper  pharyngeal  bone  lost;  postorbital  section  and  sec- 
ondary bony  canals  of  cephalic  lateralis  system  lost.  Other  di- 
agnostic characters  include:  scales  on  body  large,  lower  jaw  fre- 
quently elongate  but  upper  jaw  never  significantly  elongate;  and 
premaxillary  canal  absent. 


COLLETTE  ET  AL.:  BELONIFORMES 


353 


BELONIFORMES 
C2 


Fig.  184.     Cladogram  of  the  Beloniformes.  See  text  for  explanation 
of  character  transformation  series  A-H. 


o 

(b 

0) 

.c 

<b 

o 

(3 

■^ 

(3 

.c 

.C 

(Li 
0 

o 
o 

•V 

.C 

^^ 

o 

o 

o 
o 
o 

(Z) 

■^ 

^° 

<^* 

'^ 

Exocoetidae 


Fig.  185.  Cladogram  of  the  Exocoetidae.  I.  Swimbladder  extends 
into  haemal  canal.  2.  Pectoral  fins  enlarged.  3.  Lower  jaw  not  elongate 
in  adults.  4.  Loss  of  preanal  finfold.  5.  Barbels  present  in  juveniles  (12). 
6.  Pectoral  lateral  line  branch  lost.  7.  Beak  lost  in  juvenile  (G3).  8. 
Pectoral  fins  greatly  enlarged.  9.  Swimbladder  extends  far  into  haemal 
canal.  10.  Pelvic  fins  enlarged.  1 1.  Egg  filaments  lost. 


Hemiramphidae 

Defined  by  one  derived  character:  third  pair  of  upper  pha- 
ryngeal bones  ankylosed  into  a  plate.  Other  diagnostic  characters 
include:  pectoral  fins  short  or  moderately  long;  premaxillae 
pointed  anteriorly,  forming  a  triangular  upper  jaw  (except  in 
Oxyporhamphiis);  lower  jaw  elongate  in  juveniles  of  all  genera, 
adults  of  most  genera;  parapophyses  forked;  swimbladder  not 
extending  into  haemal  canal. 

The  Hemiramphidae  contains  12  genera  and  at  least  80  species 
(Parin  et  al.,  1980).  Four  genera,  the  first  three  monotypic  (Ar- 
rhamphus,  Chriodorus.  Melapedalion,  and  Oxyporhamphus) 
have  very  short  or  no  beaks.  Euleplorhamphus  and  Oxypo- 
rhamphus contain  two  offshore  species  each.  Zenarchoptenis, 
Dermogenys,  Hemirhamphodon.  and  Nomorhamphus  contain 
about  25  sexually  dimorphic  Indo-West  Pacific  estuarine  or 
freshwater  species.  Three  of  these  genera  (Dermogenys.  Hemi- 
ramphodon.  and  Nomorhamphus)  are  viviparous  and  have  the 
anal  fin  of  the  male  modified  into  what  Brembach  (1976)  has 
termed  an  andropodium.  Hemiramphus  (with  10  species)  is  a 
world  wide  manne  genus.  Rhynchorhamphus  (with  4  species) 
has  fimbriate  nasal  papillae  and  is  confined  to  Indo-West  Pacific 


marine  waters.  Hyporhamphus.  the  most  speciose  genus,  in- 
cludes two  subgenera,  Hyporhamphus  with  23  species,  Repo- 
rhamphus  with  1 1  species.  Some  of  these  are  marine,  some  es- 
tuarine, and  some  freshwater.  All  genera  are  characterized  by 
particular  lateral  line  characters  (Parin  and  Astakhov,  1982). 

Exocoetidae 

Defined  by  one  derived  morphological  character  and  three 
derived  early  life  history  characters:  swimbladder  extending  into 
haemal  canal;  lower  jaw  of  adults  not  elongate;  preanal  finfold 
reduced  or  lost;  and  pectoral  fins  form  last.  Other  diagnostic 
characters  include:  third  pair  of  upper  pharyngeal  bones  coales- 
cent,  the  plate  readily  separating  into  its  left  and  right  compo- 
nents; pectoral  fins  long;  premaxillae  with  straight  anterior  mar- 
gin; parapophyses  simple,  not  forked. 

The  family  Exocoetidae  contains  7  genera  and  about  50-55 
species  (Parin,  1961)  which  have  been  placed  in  four  subfamilies 
(Bruun,  1935;  Parin,  1961;  Fig.  185):  Fodiatorinae  containing 


354 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES -AHLSTROM  SYMPOSIUM 


only  Fodiator  acutus  (with  two  subspecies;  reaching  195  mm 
SL);  Parexocoetinae  with  two  species  of  Parexocoetus  (reaching 
140  mm  SL);  Exocoetinae  with  three  species  of  Exocoetus 
(reaching  200  mm  SL);  and  Cypselurinae  with  four  genera— 
Prognichthys  (4  species;  reaching  1 90  mm  SL),  Cypselurus  sensu 
stricto  (11  species;  reaching  260  mm  SL),  Cheilopogon  (not 
differentiated  from  Cypselurus  by  some  authors;  1 8  species;  con- 
tains the  largest  species  of  flyingfishes,  some  reaching  380  mm 
SL),  and  Hinmdichthys  (8  species;  reaching  190  mm  SL;  in- 
cludes the  more  specialized  subgenus  Danichthys  which  was 
recognized  as  a  genus  by  Bruun  and  others).  All  are  strictly 
marine,  mostly  in  tropical  and  subtropical  waters. 

Similarities  in  the  skeletal  structure  (Parin,  1961)  and  lateralis 
system  (Parin  and  Astakhov,  1982)  between  Exocoetus  and  the 
Cypselurinae  (Cheilopogon,  Cypselurus,  Prognichthys,  and  Hi- 
rundichthyus)  indicate  that  differentiation  of  Exocoetus  from 
the  main  stem  took  place  significantly  later  than  separation  of 
the  primitive  short-winged  flyingfishes  (Fodiator  and  Parexo- 
coetus).  There  is  particular  interest  in  the  interrelationships  within 
the  subfamily  Cypselurinae.  One  problem  concerns  whether 
Cypselurus  should  be  accepted  in  the  wide  sense  (Bruun,  1935; 
Staiger,  1965;  Gibbs  and  Staiger,  1970)  or  divided  into  two 
genera,  Cypselurus  and  Cheilopogon  (Parin,  1961).  The  diag- 
nostic differences  between  these  two  genera  are  not  simple. 
Therefore,  Parin  herein  presents  the  following  definition:  "lower 
jaw  usually  a  little  shorter  than  the  upper;  at  least  some  jaw 
teeth  tricuspid;  juveniles  with  a  single  chin  barbel  or  without 
barbels"  in  Cypselurus,  and  "lower  jaw  a  little  longer  than  upper, 
teeth  mostly  unicuspid  or  with  smaller  supplementary  cusps 
laterally;  juveniles  with  two  barbels  which  may  be  fused  into  a 
napkin-like  appendage"  in  Cheilopogon.  Each  genus  contains 
groups  of  species,  several  of  which  were  distinguished  by  Bruun 
(1935)  or  Parin  (1961)  at  the  level  of  subgenera. 

The  similarities  and  differences  between  species  groups  are 
most  noticeable  in  the  juvenile  stages  and  form  the  basis  of  the 
systematics  of  the  Cypselurinae  worked  out  by  Parin  (1961).  If 
we  consider  barbels  in  flyingfishes  to  be  derived  from  the  pair 
of  cutaneous  lappets  on  the  lower  jaw  of  needlefishes,  halfbeaks, 
and  primitive  flyingfishes,  the  most  generalized  state  of  this 
character  is  the  presence  of  two  separate  barbels.  Their  deriv- 
atives are  fusion  into  a  single  appendage  or  complete  loss.  In 
the  speciose  genus  Cheilopogon,  according  to  the  classification 
of  Parin  (1961),  the  juvenile  stages  of  most  intrageneric  group- 
ings—the subgenera  Procypselurus  (composed  of  the  Ch.  ni- 
gricans and  Ch.  cyanoptenis  groups),  Maculocoetus,  and  Abe- 
ichthys— are  characterized  by  a  pair  of  barbels,  sometimes  joined 
at  their  bases,  and  presence  of  an  enlarged  melanistic  dorsal  fin 
("Parexocoetus  stage").  In  juveniles  of  the  subgenus  Cheilopo- 


gon. the  dorsal  fin  is  greatly  enlarged,  but  the  barbels  are  fused 
into  a  fringed  appendage.  In  the  subgenus  Ptemchthys,  paired 
barbels  remain  but  the  " Pare.xocoetus  stage"  is  lost  (present  only 
in  Ch.  longibarbits,  which,  apparently  should  be  removed  from 
this  subgenus).  The  subgenus  Paracypselurus  is  somewhat  in- 
termediate between  Cheilopogon  and  Cypselurus.  Juveniles  have 
paired  barbels  and  an  enlarged  dorsal  fin,  but  adults  are  closer 
to  Cypselurus  \n  structure  ofthejaw  and  other  characters  (except 
absence  of  tricuspid  teeth). 

Summary 

There  is  a  considerable  amount  of  information  available  on 
the  early  life  stages  of  beloniform  fishes.  Specialized  structures 
such  as  egg  filaments,  barbels,  beaks,  and  melanistic  dorsal  fin 
lobes  have  systematic  value.  It  is  pleisiomorphous  for  the  eggs 
of  beloniform  fishes  to  have  chorionic  filaments  (Rosen  and 
Parenti,  1981).  One  or  more  loss  events  presumably  gave  rise 
to  the  apomorphous  condition,  an  absence  of  chorionic  fila- 
ments, seen  in  the  dwarf  sauries  (Cololabis  adocetus  and  Scom- 
heresox  simidans)  and  in  the  flyingfishes  of  the  genus  Exocoetus. 
The  development  of  a  beak  during  some  life  stage  is  a  derived 
feature  that  occurs  in  all  belonids,  scomberesocids  (except  C 
adocetus)  and  hemiramphids,  and  the  two  most  primitive  ex- 
ocoetid  genera  (Fodiator  and  Pare.xocoetus).  It  is  never  found 
in  the  adrianichthyids.  Presence  of  a  beak  is  a  synapomorphy 
for  the  Exocoetoidei  and  supports  Rosen  and  Parenti's  (1981) 
division  of  the  Beloniformes  into  two  suborders,  the  Adrianich- 
thyoidei  (no  beak)  and  the  Exocoetoidei  (beak).  A  second  char- 
acter that  supports  this  is  relative  length  of  the  gut  at  hatching, 
40-50%  standard  length  in  Adrianichthyoidei  and  approxi- 
mately 66%  in  the  Exocoetoidei.  The  superfamily  Scombere- 
socoidea  differs  from  the  Exocoetoidea  in  having  a  premaxillary 
lateral  line  canal  and  in  having  the  upper  jaw  at  least  slightly 
elongate. 

(B.B.C.)  National  Marine  Fisheries  Service  Systematics 
Laboratory,  National  Mliseum  of  Natural  History, 
Washington,  District  of  Columbia  20560;  (G.E.M.) 
Section  of  Ichthyology.  Los  Angeles  County  Museiim 
OF  Natural  History,  900  E.xposition  Boulevard,  Los 
Angeles,  California  90007;  (N.V.P.)  P.P.  Shirshov 
Institute  of  Oceanology,  Academy  of  Sciences  of  the 
U.S.S.R.,  Krasikova  Street  22,  Moscow  1 1 72 1 8,  U.S.S.R.; 
(S.M.)  Research  Division,  Fisheries  Agency,  Ministry 
OF  Agriculture,  Forestry  and  Fisheries,  Government 
OF  Japan,  2-1,  1-Chome,  Kasumigasekj,  Chiyoda-Ku, 
Tokyo,  Japan. 


Atheriniformes:  Development  and  Relationships 
B.  N.  White,  R.  J.  Lavenberg  and  G.  E.  McGowen 


IN  the  latest  statement  on  the  evolutionary  relationships  of 
the  atherinomorph  fishes  (Rosen  and  Parenti,  1981),  mono- 
phyly  could  not  be  established  fijr  the  Atherinoidei.  No  derived 
characters  could  be  offered  to  unite  the  constituent  families 
(Atherinidae.  Bedotiidae,  Isonidae,  Melanotaeniidae,  Phallo- 
stethidae,  and  Telmatherinidae)  and  the  group  term  Atheri- 
noidei was  dropped  in  favor  of  a  listing  convention  placing  them 
in  Division  I  of  a  general  classification  of  the  series  Atherino- 
morpha.  In  this  report,  two  synapomorphic  character  states  are 
described  that  suggest  that  the  Division  I  fishes  are  indeed  a 
monophyletic  group  and  the  group  name  Atheriniformes  is  res- 
urrected for  this  assemblage.  This  new  order  is  defined  by  a 
derived  larval  pigmentation  pattern  and  a  reduction  in  preanal 
length  that  persists  from  hatching  through  early  flexion.  Except 
for  this  modification,  the  classification  and  familial  designations 
of  Rosen  and  Parenti  (1981)  are  accepted  here. 

Development 
Eggs 

Information  on  atheriniform  egg  morphology  is  assembled  in 
Table  93.  The  smallest  atheriniform  egg  known,  that  of  Atherion 
elymus.  measures  0.55-0.58  mm  in  diameter  (Nakamura,  1936). 
The  largest  eggs  average  approximately  2.3  mm  in  diameter  and 
are  found  in  the  genus  Aihehna  (Marion,  1 894a;  Kanidev,  1961). 
Numerous  oil  globules  are  found  in  the  yolk  of  most  species. 
Usually,  the  globules  aggregate  at  the  vegetal  pole  and  may 
coalesce  into  a  single  droplet  that  comes  to  lie  near  the  heart. 
In  Bedotia  geayi.  the  globules  form  an  equatorial  ring  two  hours 
after  fertilization  and  reach  the  vegetal  pole  by  the  blastula  stage 
(N.  R.  Foster,  Fish.  Wildl.  Serv.,  Michigan,  pers.  comm.).  At 
fertilization,  there  may  be  as  few  as  one  oil  globule,  in  Chiros- 
toma  bartoni  (de  Buen,  1 940),  or  as  many  as  115,  in  Leuresthes 
tenuis  (David.  1939). 

Although  absent  in  Leuresthes,  Atherion,  and  Bedotia.  cho- 
rionic filaments  are  found  on  the  eggs  of  most  species.  The  eggs 
can  be  bound  together  in  a  mass  by  these  filaments  or  attached 
singly  to  a  substratum.  There  is  only  one  filament  on  the  eggs 
of  Eurystole  eriarcha.  Menidia  extensa.  and  Telinathenna  la- 
digesi  but  most  species  have  more.  The  filaments  can  be  scat- 
tered over  the  surface  of  the  egg,  as  in  Atherinops  and  Atheri- 
nopsis.  or  gathered  together  in  a  tuft  as  in  .-itherina.  Membras, 
Odontesthes.  Melanotaenia.  Memdia  menidia  and  Afenidia  ber- 
yllina.  In  Menidia  beryllina,  one  filament  is  much  enlarged; 
being  longer  and  thicker  than  the  others  making  up  the  tuft 
(Hildebrand,  1922).  Until  more  information  is  available,  it  will 
be  difficult  to  assess  the  phylogenetic  significance  of  this  vari- 
ation in  the  size,  number  and  placement  of  the  chorionic  fila- 
ments. No  pattern  is  readily  apparent.  In  some  cases,  not  all  of 
the  species  assigned  to  a  genus  have  the  chorionic  filaments 
arranged  in  the  same  way.  In  both  Menidia  and  .Austroinenidia 
there  are  species  in  which  the  filaments  are  collected  in  a  tuft 
and  species  in  which  they  are  randomly  scattered.  Two  egg  types 
may  occur  in  .tt/urinops  affinis.  There  are  approximately  6  fil- 


aments attached  at  one  end  to  the  chorion  (Crabtree,  pers.  comm.) 
(Fig.  186A)  or  40-78  looped  filaments  attached  by  both  ends 
to  the  egg  surface  (Curless,  1979).  This  unusual  occurrence  of 
two  egg  types  in  Atherinops  may  support  the  contention  that 
there  is  more  than  one  species  in  the  genus  (Hubbs,  1918). 

The  remarkable  ovarian  egg  of  Eurystole  eriarcha  is  unlike 
that  known  for  any  other  atheriniform  species.  It  averages  1.7 
mm  in  diameter  and  is  pigmented,  with  a  brownish  band  swirl- 
ing over  its  surface  (Fig.  I86B).  Arising  from  the  pigmented 
portion  of  the  chorion  are  numerous  small  anchor-shaped  ped- 
icels. Each  egg  has  one  major  filament  arising  from  the  side  of 
one  of  these  unusually  shaped  pedicels  (Fig.  187  upper).  Some 
eggs  appear  to  have  a  small  number  of  finer  filaments  similarly 
attached  to  some  of  the  other  pedicels,  but  the  majority  of  these 
chorionic  projections  do  not  have  attached  filaments.  Each  fil- 
ament can  become  entangled  in  the  pedicels  of  its  own  and 
neighboring  eggs  (Fig.  1 86B).  The  pedicels  and  small  depressions 
that  serve  as  bases  of  attachment  are  unpigmented. 

The  vitelline  circulatory  system  of  all  atheriniform  species 
examined  is  simple,  unbranched  and  looping.  This  pattern  is 
common  within  the  Atherinomorpha.  However,  the  vitelline 
circulatory  system  of  the  cyprinodontoids  is  characterized  by  a 
complex  branching  pattern. 

Larvae 

Morphologically,  the  larvae  of  the  atheriniform  fishes  are  much 
less  variable  than  the  eggs.  Development  is  direct  and  the  known 
larvae  are  similar  in  appearance  (Fig.  188).  Pectoral  fin  buds 
appear  in  embryos.  Throughout  the  Atheriniformes  the  preanal 
finfold  regresses  as  the  origin  of  the  dorsal  finfold  comes  to  be 
more  posteriorly  placed.  After  hatching,  fin  rays  develop  in  the 
caudal  fin  ventral  to  the  upturned  tip  of  the  vertebral  column. 
Next,  the  pectoral,  anal  and  second  dorsal  fins  become  rayed 
and  then  the  pelvic  fin  buds  develop.  Finally,  spines  appear  in 
the  first  dorsal  and  anal  fins.  The  gut  is  short;  with  the  preanal 
length  averaging  one-third  the  body  length  (NL  or  SL)  from 
hatching  through  the  time  of  flexion.  In  all  atheriniform  larvae 
known,  except  Odontesthes  debueni.  preanal  length  is  less  than 
40%  of  body  length  at  flexion.  Preanal  length  in  Odontesthes 
debueni  is  45%  of  body  length  (Fig.  188 A).  All  known  ather- 
iniform larvae  are  similarly  pigmented.  Melanophores  occur  on 
the  top  of  the  head  and  dorsally  and  laterally  on  the  gut.  Typ- 
ically, a  single  row  of  melanophores  occurs  mid-laterally  along 
the  body,  as  well  as  on  the  dorsal  and  ventral  margins. 

Within  the  Atheriniformes,  the  total  number  of  vertebrae 
ranges  between  2 1  and  60.  with  the  typical  number  of  precaudal 
vertebrae  being  22-23  (Ahlstrom  notes;  Rosen  and  Parenti, 
1981).  Meristic  data  are  compiled  for  89  atheriniform  species 
and  subspecies  in  Table  94. 

Information  is  available  on  the  early  life  history  of  a  variety 
of  atheriniform  species.  The  larvae  ofAlherinomorus  insularum 
(Miller  etal.,  1979),  Iso  hawaiiensis (MxWcr  tlaX.,  1979),  Odon- 
testhes regia  (Fischer,  1 963)  and  Menidia  menidia  (Hildebrand, 
1922)  follow  the  normal  mode  of  atheriniform  development 


355 


356 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  93.    Egg  Characteristics  of  the  Atheriniformes. 


Diameter 

Vitelline 

Taxon 

(mm) 

Oil  globules 

circulation 

Filaments 

Sources 

Atherinidae 

Atherina  boyeri 

1.72-1.76 

Coalesce  and  move 
towards  heart 

Unbranched, 
looping 

— 

Sparta,  1942b 

Athenna  hepselus 

2.0-2.5 

Coalesce  and  move 

Unbranched, 

Tuft 

Marion,  1894a;  Breder 

towards  heart 

looping 

and  Rosen,  1966 

Alheruia  mochon  pontica 

2.3 

Coalesce  and  move 
towards  heart 

Unbranched. 
looping 

— 

Kanidev,  1961 

Atherina  presbyter 

1.85-1.95 

— 

— 

Numerous 

Miller,  1961 

Alhennops  affinis 

Channel  Islands  and  Mainland 

1.62 

40-78  attached 
at  both  ends 
to  the  chorion 

Curiess,  1979 

Mainland  California 

~6  attached  at 
one  end  to  the 
chorion 

Crabtree,  pers.  comm. 

Atherinopsis  californtensis 

2.0-2.5 

Numerous 

-12 

Clark,  1929 

Atherion  elymus 

0.55-0.58 

Several  large  (0.16  mm) 
and  small 

Absent 

Nakamura,  1936 

A  ustromenidia  incisa 

1.7-2.15 

Numerous,  small 

Present, 
scattered 

de  Ciechomski,  1972 

Austromenidia  regia 

1.8-2.15 

>  50.  coalesce  into  1 

Unbranched, 
looping 

Tuft  of  5-10 

Fischer,  1963 

Chirosloma  bartoni 

1.0-1.1 

1;  or  several 
coalesce  into  1 

Unbranched, 
looping 

Tuft 

de  Buen,  1 940 

Eurystole  enarcha 

1  attached  to  one 
of  numerous 
anchors 

This  study 

Leuresthes  tenuis 

1.5-1.6 

25-115  that  coalesce 

Unbranched, 
looping 

Absent 

David,  1939 

Membras  vagrans 

0.8-1.1 

1 2  that  coalesce 

Unbranched, 
looping 

Tuft 

Kuntz,  1916 

Menidia  beryllina 

0.75 

7-10  that  coalesce 

Unbranched, 
looping 

Tuft 

Hildebrand,  1922 

Menidia  extensa 

0.6 

— 

— 

1 

Davis  and  Louder,  1969 

Menidia  menidia 

1.25 

7-10  that  coalesce 

Unbranched, 
looping 

Tuft 

Hildebrand,  1922 

Menidia  notata 

2.1 

Several  of  differing 

— 

Tuft  of  4 

Ryder,  1883 

Odontesthes  bonariensis 

-1.0 

size 
Several  that  coalesce 



-10 

Minoprio,  1944 

Odontesthes  debueni 

1.65-1.86 

10-20  globules  that 
coalesce 

Unbranched, 
looping 

Tuft  of  6-9 

Fischer,  1963 

Telmatherinidae 

Telmatherina  ladigesi 

— 

— 

- 

1 

Breder  and  Rosen,  1966 

Bedotiidae 

Bedolia  geayi 

-2.0 

Numerous 

Unbranched, 

Absent 

Neal  R.  Foster, 

looping 


Phallostethidae 

Gulaphallus  mirabilis 

Gulaphallus  falcifer 


Melanotaeniidae 

Melanotaenia  maccuUochi 

Pseudomugil  signata 
Pseudomugil  signifer 


1.0-1.1 

1 

1.5 

1.6 


70-80  do  not  coalesce 

10-15  large;  numerous 
small,  do  not 
coalesce 


Numerous,  cluster  but 

do  not  coalesce 
Numerous 


8  scattered 
Many  in  a  tuft 

Tuft 


Unbranched, 
looping 


pers.  comm. 

Villadolid  and 

Manacop,  1934 
Manacop,  1936 


Neal  R.  Foster, 

pers.  comm. 
Neal  R.  Foster, 

pers.  comm. 
Breder  and  Rosen,  1 966 


WHITE  ET  AL.:  ATHERINIFORMES 


357 


with  some  minor  exceptions.  The  larva  of  /.  hawaiiensis  (Fig. 
188B)  has  a  deeper  body  than  any  other  known  atheriniform 
larva  and  A.  msidarum  (Fig.  188C)  lacks  the  ventral  melano- 
phore  series  typical  of  the  order.  The  ventral  melanophore  series 
is  also  absent  in  Odontesthes  regia  (Fischer,  1963)  which  has 
only  sparse  midlateral  pigment  at  hatching. 

In  Menidia.  dorsal  pigmentation  can  be  sparse  or  even  lacking 
(Hildebrand,  1 922).  In  M.  memdia.  it  has  been  reported  that  a 
double  row  of  dorsal  melanophores  occurs  in  older  larvae  (ca. 
11  mm)  (Lippson  and  Moran,  1974).  However,  in  a  smaller 
flexion  specimen  (8  mm),  we  found  a  double  row  of  melano- 
phores in  the  area  of  the  dorsal  fin,  but  only  a  single  row  anterior 
to  the  fin.  The  dorsal  melanophore  row  is  interrupted  by  the 
dorsal  fin  in  other  atheriniform  larvae  as  well.  This  pattern  also 
occurs  in  the  melanotaeniid  genus,  Psendomugtl  (Foster,  pers. 
comm.).  It  is  not  unusual  for  single  melanophores  to  be  divided 
by  a  developing  fin  in  other  fishes  and  it  is  assumed  here  that 
the  more  complex  distribution  of  dorsal  pigment  in  Menidia 
and  Pseudomugil  is  a  variation  on  the  simpler  pattern  seen  in 
Atherinomorus,  Iso,  Odontesthes  and  most  other  atheriniform 
larvae.  In  Melanotaenia,  a  single  dorsal  row  develops.  The  larval 
morphology  of  Melanotaenia  and  Pseudomugil  closely  resem- 
bles that  of  the  other  atheriniform  fishes  (Foster,  pers.  comm.). 

Larval  Dentatherina  mercen  differ  from  all  other  known  ath- 
eriniform larvae  but  resemble  larval  Oryzias  in  having  a  double 
row  of  melanophores  on  the  nape.  The  melanophores  on  the 
dorsal  surface  of  the  trunk  are  unpaired  except  where  they  are 
interrupted  by  the  developing  dorsal  fins.  The  larva  of  Bedolia 
geayi  (Fig.  I88D)  has  the  single  dorsal  melanophore  row  and 
short  gut  typical  of  the  Atheriniformes.  Interestingly,  the  ventral 
pigment  series  of  Bedotia  is  paired,  with  a  row  of  melanophores 
flanking  both  sides  of  the  anal  finfold  (Foster,  pers.  comm.). 

The  early  life  history  stages  of  phallostethid  fishes  follow  closely 
the  atheriniform  pattern.  In  both  Gulaphallus  mirahilts  (Villa- 
dolid  and  Manacop,  1934)  and  G.  /a/a/f'r  (Manacop,  1936)  the 
preanal  length  is  short  and  a  median  series  of  melanophores 
develops  middorsally.  The  exact  disposition  of  the  dorsal  me- 
lanophores has  not  been  described  nor  can  it  be  assessed  from 
published  illustrations. 

Relationships 

Two  ontogenetic  character  states  suggest  that  the  atheriniform 
fishes  are  a  monophyletic  group  compwising  an  order,  the  Ath- 
eriniformes, of  equal  standing  with  the  Beloniformes  and  Cy- 
prinodontiformes.  First,  the  preanal  length  of  all  known  ath- 
eriniform flexion  larvae,  except  Odontesthes  dehueni,  is  short; 
being  approximately  one-third  of  body  length.  Preanal  length  is 
variable  in  the  other  two  atherinomorph  orders  but  the  preanal 
lengths  of  few,  if  any,  beloniform  or  cyprinodontiform  species 
are  this  short  between  hatching  and  early  flexion.  The  Perco- 
morpha  is  thought  to  be  the  sister  group  of  the  Atherinomorpha 
(Rosen  and  Parenti,  1981).  In  almost  all  primitive  percomorphs, 
preanal  length  exceeds  that  of  the  Atheriniformes  through  flex- 
ion and  approaches  as  much  as  50-70%  of  body  length  (Ahl- 
strom  and  Moser,  1976).  The  same  can  be  said  of  the  paracan- 
thopterygian,  myctophiform  and  aulopiform  fishes  (sensu  Rosen; 
1973,  1982).  Preanal  length  is  reduced  in  gadid  fishes  (Dunn, 
this  volume),  but  the  short  gut  typical  of  the  cods  is  always 
looped  and  therefore  is  considered  here  to  be  nonhomologous 
with  the  condition  seen  in  the  atheriniforms.  Outgroup  com- 
parison thus  suggests  that  the  reduced  larval  preanal  length  can 


Fig.  1 86.  (A)  Atheriniform  eggs.  Mature  egg,  Atherinops  affinis.  San- 
ta Catalina  Island,  California.  LACM  field  no.  IP-77-43;  (B)  Ovarian 
egg.  Euryslole  enarcha.  LACM  31784-5;  and  (C)  Atherinopsis  califor- 
niensis.  egg.  LACM  43446-1. 


Fig.  187.     Ovarian  egg,  Eurystole  eriarcha.  LACM  31784-5.  (upper)  lOOx;  (lower)  l.OOOx. 


WHITE  ET  AL.:  ATHERINIFORMES 


359 


Fig.  188.     (A)  Atheriniform  larvae.  Odomesthes  debueni;  10.2  mm  SL,  from  Fischer,  (1963);  (B)  ho  hawaiwnsis:  6.2  mm  SL,  from  Miller  et 
al.,  (1979);  (C)  Athermomorus  insularum;  5.4  mm  SL,  from  Miller  et  al.,  (1979);  and  (D)  Bedotia  geayi;  5.3  mm  SL,  LACM  uncatalogued. 


be  used  as  a  synapomorphous  character  state  to  define  the  Ath- 
eriniformes. 

The  second  ontogenetic  character  stale  suggesting  that  the 
atheriniform  fishes  comprise  a  monophyletic  group  relates  to 
larval  pigmentation  and  may  contribute  to  their  cladistic  di- 
agnosis. In  all  atheriniform  larvae  a  single  row  of  melanophores 
develops  on  the  dorsal  margin  (Fig.  189A).  This  situation  con- 
trasts with  the  Beloniformes  and  Cyprinodontiformes,  where  no 
consistent  larval  pigmentation  pattern  is  evident  (Hardy,  1978a). 
What  is  known  of  larval  halfbeaks  suggests  that  when  a  dorsal 
pigment  series  occurs  it  is  always  composed  of  at  least  a  double 
row  of  melanophores  (Fig.  I89B). 

While  it  is  typical  for  cyprinodontiform  larvae  to  develop 
dorsal,  lateral  and  ventral  pigment  series  (Foster,  1967),  no 
consistent  pattern  is  evident.  In  Fundulus.  the  middorsal  me- 


lanophores are  arranged  in  a  paired  series  (Hardy,  1978a).  In 
Cyprinodon  vanegatus  obscure  blotches  of  pigment  occur  on  the 
body  (Hardy,  1978a).  Melanophores  are  evenly  distributed  over 
the  larva  of  Liuania  parva  (Hardy,  1978a).  The  larva  of  Epi- 
platys  sexfasciatus  has  melanophores  randomly  distributed  over 
its  dorsal  surface  (Scheel,  1968).  In  the  Atherinomorpha,  only 
the  adrianichthyoid  fishes  have  larvae  with  dorsal  melanophores 
arranged  in  a  single  row  (Kulkami,  1940;  Job,  1940).  This  re- 
semblance to  the  Athcriniformes  is  considered  to  be  convergent 
because,  given  the  mtemal  relationships  of  the  Atherinomorpha 
(Rosen  and  Parenti,  1981),  it  is  more  parsimonious  to  assume 
that  a  single  dorsal  melanophore  row  evolved  independently  in 
the  Athcriniformes  and  Adrianichthyoidei  because  only  two 
evolutionary  events  are  involved.  However,  if  this  pigment  pat- 
tern is  viewed  as  a  sympleisiomorphy,  it  is  necessary  to  invoke 


360 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  94.    Meristics  of  Selected  Atheriniform  Species.  Only  total  fin  elements  are  reported  because  of  confusion  in  the  literature  as  to  the 
proper  definition  of  spines  and  rays  and  the  inconsistencies  in  published  descriptions  that  resulted  from  this  confusion. 


Fin 

rays 

Gill 

Dl 

D2 

A 

Pect. 

Vert. 

rakers 

Reference 

Alherinidae 

Alia  net  la  mugi  hides 

5-7 

8-11 

9-12 

13-15 

34-37 

14-17 

Ivantsoff, 

1978 

Alherina  boyen 

6-9 

11-14 

12-16 

15-17 

42-46 

17-24 

Ivantsoff, 

1978 

Atherina  hepselus 

7-9 

10-13 

11-14 

17-19 

52-55 

25-28 

Ivantsoff, 

1978 

Alherina  presbyter 

7-9 

11-15 

15-19 

14-17 

45-52 

19-24 

Ivantsofl^, 

1978 

Alherinason  esox 

5-9 

10-15 

11-15 

12-14 

45-48 

18-24 

Ivantsoff", 

1978 

Atherinason  hepseloides 

5-10 

10-14 

12-15 

13-16 

43-48 

15-20 

Ivantsoff, 

1978 

Alherinomorus  capricornensis 

3-7 

10-11 

13-16 

15-18 

42-44 

20-24 

Ivantsoff^, 

1978 

A  therinomorus  endrachtensis 

4-6 

9-11 

12-15 

13-17 

35-39 

18-22 

Ivantsoff", 

1978 

Alherinomorus  ogilbyi 

4-7 

9-12 

12-17 

16-19 

37-42 

22-28 

Ivantsoff", 

1978 

Alherinomorus  pinquis 

4-7 

9-12 

13-16 

15-19 

38-43 

18-25 

Ivantsoflf, 

1978 

Atherinops  affinis 

3-7 

10-14 

9-14 

12-15 

43-49 

14-27 

White,  unpub. 

Atherinopsis  californiensis 

4-9 

10-15 

20-29 

14-17 

46-53 

18-44 

White,  unpub. 

Athennosoma  elongala 

4-8 

9-13 

9-15 

12-16 

37-43 

12-16 

Ivantsoff", 

1978 

Atherinosoma  microstoma 

5-9 

9-12 

9-14 

12-16 

37-42 

12-15 

IvantsoflT, 

1978 

Atherlnosoma  presbyteroides 

6-8 

10-14 

11-16 

12-14 

41-48 

15-20 

Ivantsoff", 

1978 

Athenon  elymus 

3-5 

9-14 

14-16 

12-15 

38-42 

10-14 

Ivantsoff", 

1978 

Atherion  macculocht 

3-5 

10-13 

15-19 

12-14 

43-46 

10-16 

Ivantsoff", 

1978 

Colpichthys  regis 

5-8 

10-13 

20-24 

13-16 

44-48 

16-19 

White,  un 

pub. 

Craterocephalus  cuneiceps 

4-6 

6-9 

7-9 

12-15 

31-34 

10-13 

Ivantsoff, 

1978 

Craterocephalus  dalhousiensis 

4-6 

6-8 

8-10 

13-15 

30-32 

8-9 

Ivantsoff", 

1978 

Craterocephalus  eyresll 

3-6 

7-9 

7-11 

12-14 

32-41 

10-13 

Ivantsoff, 

1978 

Craterocephalus  honoriae 

5-6 

8-9 

10-13 

13-15 

35-38 

12-15 

Ivantsoff", 

1978 

Craterocephalus  lacustris 

5-8 

8-10 

9-11 

13-17 

35-39 

10-13 

Ivantsoff, 

1978 

Craterocephalus  marjoriae 

5-7 

7-9 

7-10 

12-16 

29-34 

10-12 

Ivantsoff, 

1978 

Craterocephalus  nouhuysi 

6-8 

9-11 

10-12 

13-17 

37-38 

7-9 

Ivantsoff", 

1978 

Craterocephalus  pauciradiatus 

4-6 

7-9 

8-11 

12-14 

30-35 

10-13 

Ivantsoff". 

1978 

Craterocephalus  randl 

5-8 

8-10 

8-12 

13-15 

34-39 

7-11 

Ivantsoff", 

1978 

C  stercusmuscarum  fulvus 

4-8 

7-11 

8-11 

13-16 

31-36 

10-13 

Ivantsoff, 

1978 

C  stercumuscaruin  stercumuscarum 

5-8 

6-9 

8-10 

12-15 

35-38 

9-12 

Ivantsoff", 

1978 

Hypoatherina  barnesi 

5-7 

9-12 

13-15 

13-16 

41-42 

15-18 

Ivantsoff", 

1978 

Hypoatherlna  ovalaua 

4-7 

9-11 

10-13 

16-18 

38-40 

22-25 

Ivantsoff, 

1978 

Hypoatherina  temminckii 

5-7 

9-11 

12-15 

17-20 

38-44 

21-25 

Ivantsoff", 

1978 

Hypoatherina  Iropicalis 

5-8 

9-12 

12-15 

16-19 

40-47 

18-22 

Ivantsoff, 

1978 

Hypoatherina  valenciennsi 

4-7 

9-11 

13-14 

15-17 

39-42 

20-25 

Ivantsoff", 

1978 

Leuresthes  tenuis 

4-7 

9-13 

20-24 

13-16 

47-50 

20-29 

White,  un 

pub. 

Stenatherina  panalela 

6-7 

9-10 

10-13 

17-19 

21-24 

40-45 

Ivantsoff, 

1978 

Bedotiidae 

Bedotia  geayi 

4-5 

10-13 

15-18 

12 

Pellegrin, 

1907,  1914 

Bedotia  longianalis 

5 

14 

20 

12 

Pellegrin, 

1914 

Bedotia  madagascariensis 

5 

12 

19 

Pellegrin. 

1914 

Regan,  1903a 

Isonidae 

Iso  hawaiiensis 
Melanotaeniidae 
Cairnsichthys  rhombosomoides 
Chilatherina  campsi 
Chilatherina  crasslspinosa 
Chilatherina  lorentzi 
Chilatherina  sentaniensis 
Glossolepis  incisus 
Glossolepis  multisquamata 
Glossolepis  pseudoincisus 
Iriatherina  werneri 
Melanotaenia  affinis 
Melanotaenia  ajamaruensis 
Melanotaenia  hoesemani 
Melanotaenia  jluviatilis 
Melanotaenia  goldiei 
Melanotaenia  japenensis 
Melanotaenia  lacustris 
Melanotaenia  nigrans 
Melanotaenia  ogilbyi 


4-5 


17 


23-25 


12-13 


5-6 

14 

19-21 

4-8 

13-17 

21-25 

4-5 

9-13 

21-25 

4-7 

13-18 

24-31 

4-5 

10-15 

23-26 

5-6 

10-11 

21-24 

4-5 

9-12 

18-22 

5-6 

11-13 

19-23 

13-14 

6-9 

8 

11-13 

4-5 

15-19 

21-25 

4-6 

16-20 

22-28 

13-15 

4-6 

11-15 

18-24 

13-16 

5-7 

13-14 

19-21 

13-15 

5-6 

12-16 

21-25 

4-5 

16-18 

27-29 

13 

4-5 

12-14 

19-20 

4-7 

10-14 

18-22 

5-7 

10-12 

18-19 

35-38 

Miller  etal..  197 

9 

36-37 

10-12 

Allen,  1980 

13-14 

Munro,  1967 

14 

Munro,  1967 

15 

Munro,  1967 

13-15 

Munro,  1967 

32 

Munro,  1967 

16-19 

Munro,  1967 

26-30 

Allen  and  Cross, 

1980 

32-33 

11-13 

Allen.  1980 

13-17 

Munro.  1967 

14-15 

Allen  and  Cross. 

1980 

14-15 

Allen  and  Cross, 
Scott  etal.,  1980 

1980 

14-16 

Munro,  1967 

13-14 

Allen  and  Cross, 

1980 

14 

Munro,  1967 

13-15 

Munro,  1967 

12 

Munro,  1967 

WHITE  ET  AL.:  ATHERINIFORMES 


361 


Table  94.    Continued. 


Fin  rays 


Dl 

D2 

A 

Peel. 

Ven 

5-6 

19-20 

25-27 

14-15 

4-6 

11-15 

18-21 

5-7 

10-13 

17-18 

5-7 

11-12 

20-22 

13-16 

5-7 

10-14 

18-23 

4-6 

11-13 

19-22 

16-20 

4-7 

19-22 

24-27 

5-8 

10-12 

17-21 

32 

4-5 

6-7 

10-12 

10-11 

4-5 

6-7 

9-10 

12-13 

3-5 

11-15 

19-23 

35 

1-2 

5-6 

14-16 

11 

2 

7 

16-18 

9 

2 

7-8 

15-19 

9 

1-2 

7 

16-17 

9-10 

6-7 

14-16 

13 

1-2 

5-7 

13-18 

9 

1-2 

5 

14-15 

9 

1-2 

5-6 

11-14 

10 

1 

5-6 

15-16 

10 

1 

6-8 

14-15 

11 

2 

6 

16 

12 

2 

5 

14-16 

9 

1-2 

5-6 
8-10 

15-17 
26-28 

9 

6 

14 

10 

35 

5 

14 

9 

36 

2 

5 

14-16 

2 

5-7 

10-11 

8-10 

19-23 

12-13 

Gill 
rakers 


Melanotaenia  oktediensis 
Melanotaenia  praecox 
Melanotaenia  sexlmeata 
Melanotaenia  splendida  australis 
Melanotaenia  s.  rubrostnata 
Melanotaenia  trifasciata 
Melanotaenia  vanheurni 
Popondelta  furcatus 
Pseudomugil  gertrudae 
Pseudomugil  tenellus 
Rhadinocentnts  ornalus 

Phallostethidae 
Ceratostethus  bicornis 
Gulaphallus  eximius 
Gulaphallus  mirabilis 
Manacopus  falcifer 
Mirophallus  bikolanus 
Neostethus  amaricola 
Neostethus  borneensis 
Neostethus  coronensis 
Neostethus  lankestn 
Neostethus  panayensis 
Neostethus  siamensis 
Neostethus  villadolidi 
Neostethus  zamboanga 
Phallostethus  duncken 
Phenacostethus  smithi 
Phenacostethus  posthon 
Plectrostethus  palawanensis 
Solenophallus  ctenophonis 
Solenophallus  thessa 

Telmatherinidae 

Tehnathenna  celebensis 


6-7 


11-12 


13-15 


16 


32 


15 

Allen  and  Cross,  1980 

11-12 

Munro,  1967 

12 

Munro,  1967 

16 

Taylor.  1964 

14-16 

Munro,  1967 

Taylor,  1964 

14 

Munro,  1967 

8-10 

Allen,  1980 

Taylor,  1964 

9 

Taylor,  1964 

11-12 

Allen,  1980 

Herre,  1942 

Herre,  1942 

12-13 

Herre,  1942 

Herre,  1942 

12 

Herre,  1942 

Herre,  1942 

Herre,  1942 

10-11 

Herre,  1942 

Herre,  1942 

15 

Herre,  1942 

Herre,  1942 

15 

Herre,  1942 

12 

Herre,  1942 

Regan,  1913c 

Roberts,  1971 

Roberts,  1971 

Herre,  1942 

13 

Herre,  1942 

16 

Herre,  1942 

Boulenger,  1897 

Fig.  189.     Dorsal  pigment  series  in  athennomorph  larvae.  (A)  Atherinops  affinis;  6.7  mm  SL,  LACM  field  no.  42841-3;  and  (B)  Dermogenys 
pusdlus:  7.7  mm  SL,  LACM  43448-1. 


362 


ONTOGENY  AND  SYSTEM ATICS  OF  FISHES -AHLSTROM  SYMPOSIUM 


Table   95. 


Derived  Developmental  Character   States   in   the 
Atherinomorpha. 


Alhenni- 
formes 


Cypnno- 
Beloni-        donti- 
formes        formes 


Egg  large,  demersal  with  chorionic  fila- 
ments and  lipid  globules  coalescing  at 
vegetal  pole 

Separation  of  embryonic  afferent  and  effer- 
ent circulations  by  development  of  heart 
in  front  of  the  head 

Formation  of  spermatogonia  near  the  tuni- 
ca albuginea 

Gut  length  less  than  40%  of  flexion  length 

Single  row  of  melanophores  on  dorsal  sur- 
face 

Fin  rays  present  at  hatching 


X 
X 


the  development  of  a  single  dorsal  melanophore  row  in  the 
common  ancestor  of  the  Atherinomorpha  and  separate  loss 
events  in  the  Exocoetoidei  and  Cyprinodontiformes. 


'  After  this  paper  went  to  press,  a  report  on  the  relationships  of  the 
phallostethid  fishes  appeared  in  which  Parenti  (1984)  questions  our 
conclusions  on  atherinifomi  monophyly.  The  evidence  she  presents 
suggests  another  phylogenetic  interpretation  but  at  the  present  time 
neither  hypothesis  can  definitely  be  rejected. 


Aspects  of  the  variable  reproductive  behavior  of  the  ather- 
iniform  fishes  might  be  useful  in  determining  relationships  with- 
in the  order.  The  habit  shared  by  Leuresthes  tenuis  and  L.  sar- 
dina  of  spawning  on  the  beach  in  synchrony  with  the  lunar  cycle 
(Thompson  and  Thompson,  1919;  Clark,  1925;  Walker,  1952) 
is  a  synapomorphy  identifying  the  two  grunion  species  as  each 
other's  closest  relative.  Another  lunar  spawner,  Menidia  men- 
idia  deposits  its  eggs  in  detrital  mats  and  on  the  stems  and 
exposed  roots  of  the  cordgrass  plant  Spartina  alternijlora  (Moore, 
1980;  Middaugh  et  al.,  1981).  Telmatherina  ladigesi  deposit 
their  eggs  over  a  period  of  several  days,  attaching  them  singly 
and  in  a  widely  spaced  pattern  to  aquatic  vegetation.  In  the 
Phallostethidae,  fertilization  is  internal  and  the  eggs  are  attached 
to  a  substratum  by  their  adhesive  filaments.  There  is  much 
variation  in  the  reproductive  behavior  of  atheriniform  fishes 
and  investigation  of  their  breeding  habits  might  secure  infor- 
mation bearing  on  their  systematic  relationships. 

Table  95  summarizes  the  derived  ontogenetic  characters  that 
bear  on  atherinomorph  relationships.  There  is  still  much  that 
is  unknown  about  the  early  life  history  of  the  atheriniform  fishes 
and  it  is  reasonable  to  hope  that  future  investigation,  particularly 
of  their  reproductive  habits  and  egg  morphology,  will  contribute 
to  the  elucidation  of  their  evolutionary  relationships.' 

Section  of  Fishes,  Natural  History  Museum,  Los  Angeles, 
California  90007. 


Cyprinodontiformes:  Development 
K.  W.  Able 


THE  approximately  800  members  of  the  Cyprinodontiformes 
(killifishes)  are  small  to  medium-sized  fishes  (8-300  mm 
SL)  that  live  in  shallow  fresh  and  brackish  water.  They  are  nearly 
worldwide  in  their  distribution  in  temperate  and  tropical  areas 
(Parenti,  1981).  Cyprinodontiformes  is  considered  to  be  mono- 
phyletic  based  on  several  adult  osteological  characters  and  the 
long  embryonic  development  time  (Parenti,  1 98 1 ).  I  here  follow 
the  most  recent  and  extensive  revision  of  the  group  by  Parenti 
( 1 98 1 )  in  which  she  rearranges  them  into  two  suborders:  Aploch- 
eiloidei  with  2  families  (Aplocheilidae  and  Rivulidae)  and  Cy- 
prinodontoidei  with  7  families  (Profundulidae,  Fundulidae, 
Valenciidae,  Anablepidae,  Poeciliidae,  Goodeidae,  Cyprino- 
dontidae).  See  Nelson  (1976)  and  Parenti  (1981,  Table  3)  for 
prior  classification  schemes.  Comments  on  portions  of  Parenti's 
reclassification  can  be  found  in  Klee  (1982)  and  Foster  (1982). 
Reproduction  and  development  within  the  group  is  excep- 
tionally varied,  with  oviparity,  ovoviviparity,  viviparity  (in- 
cluding functional  states  of  each)  and  functional  hermaphro- 
ditism represented.  In  addition,  viviparity  may  have  evolved 
independently  at  least  four  times  within  the  order  (Parenti,  1981). 
Among  the  viviparous  forms  occur  a  vast  array  of  schedules 
and  morphological  modifications  for  internal  development  such 
as  the  trophotaeniae  of  the  goodeids  and  the  intra-  and  extra- 
follicular  gestation  and  superfetation  in  some  poeciliids.  De- 
velopment reportedly  is  long,  from  four  days  to  more  than  one 


year  (Scheel,  1962)  in  some  of  the  "annual"  killifishes.  The 
rivulid,  Rivulus  mannoratus,  is  unique  among  fishes,  and  ver- 
tebrates in  general,  in  that  it  is  a  functional  hermaphrodite  with 
internal  fertilization  (Harrington,  1961).  Published  early  life  his- 
tory descriptions  are  listed  in  Table  96. 

Eggs 

The  eggs  of  some  Cyprinodontiformes  are  among  the  smallest 
known  for  fishes.  Scrimshaw  (1946)  recorded  fertilized  eggs  of 
the  poeciliid  Hcterandria  fonnosa.  in  which  development  is 
internal,  to  average  0.30  mm  and  Roberts  (1970)  recorded  "ripe" 
eggs  of  another  poeciliid,  Fluviphylax,  as  0.1  mm  (not  substan- 
tiated). The  eggs  of  other  cyprinodontiforms  are  larger  (Table 
97)  with  the  largest  that  of  Fiindiilus  majalis  at  2.0-3.0  mm. 
Egg  size  varies  within  some  species  (i.e.,  the  cyprinodontid 
Aphanius  anatoliae.  Grimm,  1979a,  b)  and  judging  from  the 
data  in  Table  97,  may  vary  in  other  species  as  well.  Other  authors 
have  noted  differences  in  the  egg  size  of  cyprinodontids  of  the 
genus  Cyprinodon  and  considered  them  to  be  environmental 
(Soltz  and  Hirschfield,  1 98 1)  or  genetic  (Garrett,  1 982).  Fecun- 
dity is  correlated  with  egg  size  in  the  aplocheilid  Nothohranchius 
(Bailey,  1972)  and  with  female  size  in  poeciliids  (see  Thibault 
and  Schultz,  1978).  Fecundity  also  varies  between  females  and 
populations  in  the  oviparous  goodeid,  Crentchthys  baileyi  (Es- 
pinosa,  1968).  Superfetation  occurs  in  several  poeciliid  genera 


ABLE:  CYPRINODONTIFORMES 


363 


(Turner,  1937;  Scrimshaw,  1945;  Turner  1940a;  Thibault  and 
Schultz,  1978). 

The  eggs  of  all  cyprinodontiforms  contain  conspicuous  oil 
droplets  (Foster.  1967)  (Table  97)  including  the  viviparous  poe- 
ciliids  such  as  Gambusia  affinis  (Kuntz,  1914a).  Within  the 
Fundulidae  the  size  and  number  of  oil  droplets  is  extreme;  Lu- 
cania  parva  has  8-12  large  droplets  (Fig.  190C)  and  Fundulus 
n.  sp.  from  Bermuda  has  up  to  approximately  350  with  a  mean 
of  181  droplets  per  egg  (Fig.  190F,  and  Able  et  al.,  in  prep.). 
Subspecific  variation  in  the  fundulid  F.  heteroclitus  is  pro- 
nounced and  population  means  range  from  10  to  180  droplets 
(Morin  and  Able,  1983). 

These  droplets  probably  provide  nutrition  late  in  embryonic 
development  (Smith,  1957;  Lentz  and  Trinkhaus,  1967;  Blaxter, 
1969a;  Temer,  1979).  The  chemical  composition  of  lipids  in 
the  oil  droplets  has  been  determined  by  Bailey  (1973).  The  oil 
droplets  are  clumped  together  at  ovulation  but  disperse  after 
fertilization.  Individual  oil  droplets  are  retained  in  the  yolk  sac 
after  hatching  in  several  Fundulus  species,  L.  parva  (see  Hardy, 
1978a),  R.  marmoratus  (McMillan,  1979)  and  postflexion  G. 
affinis  (Ryder.  1885).  The  eggs  of  all  known  oviparous  and  ovo- 
viviparous  cyprinodontiforms  have  a  small  perivitelline  space 
and  are  spherical  (except  in  Nothobranchius  in  which  the  egg  is 
oval,  Scheel,  1968). 

The  chorion  is  variable  in  thickness  and  surface  structure 
(Table  97).  In  most  of  the  oviparous  and  ovoviviparous  forms 
the  chorion  is  multilayered  and  thick,  whereas  in  many  vivipa- 
rous forms  it  is  considerably  reduced  (see  Flegler,  1977).  De- 
tailed studies  of  the  chorion  microstructure  are  available  for  the 
fundulid  F.  heteroclitus  (Kuchnow  and  Scott,  1977)  and  the 
rivulid  Cynolebias  bellottii  (Sterba  and  Muller,  1962;  Muller 
and  Sterba,  1963).  The  chorion  of  all  oviparous  and  ovovivipa- 
rous forms  have  adornments  of  some  type  on  the  surface.  In- 
stances where  they  have  been  reported  as  lacking  (F.  heteroclitus, 
F.  parvipinnis.  Foster,  1967;  F.  majalis.  Hardy,  1978a)  are  in- 
correct. Often  the  chorion  is  covered  with  filaments  either  uni- 
formly arranged  or  clustered  together  to  form  tufts  (Fig.  190, 
191;  Table  97).  The  filaments  can  vary  in  diameter  and  density 
between  species  (Fig.  190  and  191)  and  subspecies  (Dumont 
and  Brummetl,  1980;  Morin  and  Able,  1983).  Differences  in 
these  structures  in  F.  heteroclitus  appear  to  be  correlated  with 
spawning  site  preference  (Able,  1984).  Fundulus  majalis  has 
microfilaments  on  the  large  filaments  and  on  the  chorion  surface 
(Fig.  191  A,  B).  Some  species  have  other  structures  ("punctae" 
of  Foster,  1967)  which  appear  as  small  spherical  knobs  on  the 
surface  of  the  chorion,  occasionally  with  filaments  originating 
from  them  (Fig.  190D,  E;  19 ID,  E).  In  other  species  the  surface 
of  the  chorion  may  be  sculptured  (Table  97).  Fundulus  luciae 
has  numerous  circular  pits  in  the  chorion  surface  (Fig.  190D, 
E).  The  distribution  of  chorionic  modifications  (filaments,  mi- 
crofilaments, pits,  knobs)  within  the  Cyprinodontiformes  is  in- 
completely known  and  thus  it  is  difficult  to  assess  their  phylo- 
genetic  significance.  Several  species  of  fundulids  studied  possess 
punctae  or  knobs  (Table  97;  Figs.  190,  191)  while  these  are 
lacking  in  the  cyprinodontids  (see  Fig.  1 90A,  B)  thus  supporting 
the  separation  of  these  groups  by  Parent!  (1981). 

The  presence  of  chorionic  filaments  in  the  Cyprinodonti- 
formes is  a  synapomorphy  shared  with  the  Atheriniformes  and 
Beloniformes  as  discussed  in  this  volume  and  constitutes  one 
of  the  synapomorphies  serving  to  unite  the  Atherinomorpha. 
Further  studies  of  egg  morphology  in  the  oviparous  forms  will 


Table  96. 


Published  Descriptions  of  Cyprinodontiform  Early  Life 
History  Stages  Listed  by  Family  and  Genus. 


Family  and  genus 

Sources 

Suborder  Aplocheiloidei 

Aplocheilidae 

Fundulopanchax 

Peters,  1963 

Nothobranchius 

Peters,  1963 

Zahradka  and  Frank,  1976 

Rivulidae 

Rivulus 

McMillan,  1979 

Suborder  Cyprinodontoidei 

Profundulidae 

None 

Fundulidae 

Planet  erus 

Koster,  1948 

Fundulus 

Ryder,  1885 

Kuntz,  1914a.  1916 

Fish,  1932 

Armstrong  and  Child,  1965 

Foster,  1967.  1974          ' 

Byrne,  1978 

Hardy,  1978a 

Jones  and  Tabery,  1980 

Lucania 

Kuntz,  1916 

Foster,  1967,  1974 

Hardy.  1978a 

Leptolucania 

Foster.  1967 

Adinia 

Foster,  1967 

Koenig  and  Livingston,  1976 

Valenciidae 

None 

Goodeidae 

Crenichthys 

Kopec,  1949 

Cyprinodontidae 

Cubanichthys 

Troemner,  1932,  1941 

Cyprinodon 

Kuntz.  1916 

Foster,  1974 

Hardy,  1978a 

Mettee  and  Beckham,  1 978 

Jordanella 

Foster,  1967 

Anablepidae 

Anableps 

Turner,  1940c 

Jenynsia 

Turner,  1940d 

Poeciliidae 

Gambusia 

Ryder,  1885 

Kuntz,  1914a 

Xiphophorus 

Tavolga,  1949 

probably  provide  useful  insights  into  the  phylogeny  of  this  abun- 
dant and  diverse  group. 

Embryonic  development 

Embryonic  development  within  the  Cyprinodontiformes  is 
almost  as  variable  in  duration  and  number  of  physiological  and 
morphological  modifications  as  in  all  other  fishes  combined. 
The  incubation  time  may  be  as  short  as  4-8  days  in  C.  variegatus 
and  Jordanella  Jloridae  (see  Foster,  1967;  Hardy,  1978a)  to 
possibly  longer  than  a  year  in  some  of  the  "annual"  species. 
Parent!  (1981)  considers  this  annual  habit  to  have  developed 
more  than  once  within  the  Cyprinodontiformes.  This  is  sup- 


364 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  97.    Summary  of  Ego  Characteristics  of  Cyprinodontiform  Fishes. 


Species 

Egg  diameter 
(mm) 

Oil  globules 

Chorion  surface 

Source(s) 

Suborder  Aplocheiloidei 

Family  Rivulidae 

Cynolebias  bellotlii 
Cynolebias  ladigesi 
Cynolebias  whilei 
Cynolebias  melanotaenia 
Procatopus  nototaenia 

ca.  2.0 

7 

ca.  1.0 

7 

1.0 

present 

7 

present,  large 

7 

9 
anchorlike  filaments 
prickles 

palm  like  stems 
filaments 

Breder  and  Rosen,  1966 

Peters.  1963 

Breder  and  Rosen,  1966 

Scheel,  1962 

Scheel,  1961 

Family  Aplocheilidae 

Fundulopanchax  amieti 
Fundulopanchax  arnoldi 
Aplocheilus  blocki 
Aplocheilus  Imeatus 
Epiplatys  chaperi 
Epiplalys  senegalensis 
Nothobranchius  guentheri 
Nothobranchius  korthausae 

7 

7 

13 
2.0 

•> 

1.0 

9 

0.9-1.1 

? 
? 

7 

9 

present,  large 
7 

present,  large 
present,  large 

filaments 
hexagonal  pattern 
3  long  filaments 
7 

filaments  in  tuft 
7 

filaments 
filaments 

Carr,  1982 

Peters,  1963 

Jones.  1937 

Job,  1940 

Peters,  1963 

Breder  and  Rosen,  1966 

Peters,  1963 

Zahradka  and  Frank,  1976 

Suborder  Cyprinodontoidei 
Family  Fundulidae 
Planclerus  kansae 
Fundulus  chrysotus 
F.  cingulatus 
F.  confluentus 

F.  diaphanus 

F.  heteroclitus 

F.  luciae 
F.  majalis 

F.  notatus 

F.  nottii 
F.  olivaceus 

F.  parvipinnis 

F.  similis 

Lucania  goodei 
Lucania  parva 

Leptolucania  ommata 

Adinia  xenica 


Family  Poecilidae 

Gambusia  ajjinis 
Poecilia  reticulata 
Poeciliopis  lucida 
P.  monacha 
P.  prolifica 
P.  lurneri 
Tomeurus  gracilis 

Family  Goodeidae 
Crenichthys  baileyi 

Family  Valenciidae 
Valencia  hispanica 


2.3-2.4 

7 

filaments,  punctae 

Foster,  1967 

2.0-2.1 

7 

filaments,  punctae 

Foster,  1967 

1.5-1.6 

7 

filaments,  sculptunng 

Foster,  1967 

1.6-1.8 

10-15,  medium 

filaments,  punctae 

Harrington,  1959;  Foster, 
1967;  Hardy,  1978 

1.7-2.4 

10-15.  medium  and 

filaments,  punctae 

Foster,  1 967;  Wang  and 

40-80,  small 

Kemehan,  1979 

1.5-2.5 

10-180 

filaments,  variable 

Hardy,  1978a;  Morin  and 
Able.  1983 

1.7-2.2 

5-58 

filaments,  punctae 

Hardy,  1978a 

2.0-3.0 

50 

absent  ?  and  present 

Ryder,  1885;  Nichols  and 
Breder,  1927;  Hardy,  1978a 
Wang  and  Kemehan,  1979 

ca.  1.8 

many,  medium  also 

filaments  in  tuft 

Foster,  1967;  Jude,  1982a 

single 

and  punctae 

2.2-2.3 

numerous 

filaments 

Foster,  1967 

ca.  1.8 

many,  medium 

filaments  in  tuft 
and  punctae 

Foster,  1967 

2.8 

present 

filaments 

Ritterand  Bailey,  1908; 
Hubbs,  1965 

2.6-2.9 

7 

? 

Foster,  1967;  Martin  and 
Finucane,  1969 

1.3 

10-12,  medium 

filaments  in  tuft 

Foster,  1967 

1.0-1.3 

10-12(0.3-0.4  mm) 

filaments  in  tuft 

Kuntz,  1916;  Foster,  1967; 
Hardy,  1978a 

1.0-1.1 

7 

filaments  in  tuft, 
sculpturing 

Foster,  1967 

1.5-2.2 

many,  small 

filaments 

Foster,  1967;  Koenig  and 

1.6-2.1 

present 

1.7 

9 

1.4 

9 

2.0 

7 

1.0 

7 

1.0 

7 

9 

9 

absent 

7 

7 

7 

7 

7 
many,  long  filaments 


1.9-2.0 


2.5-2.6 


Livingston,  1976;  Hastings 
and  Yerger,  1971 

Hardy,  1978a 

Thibaultand  Schultz,  1978 

Thibault  and  Schultz,  1978 

Thibaultand  Schultz,  1978 

Thibaultand  Schultz,  1978 

Thibault  and  Schultz,  1978 
Breder  and  Rosen,  1 966 


Kopec,  1949;  Bnll,  1981 


Rachow,  1924;  Villwock, 
1960 


ABLE:  CYPRINODONTIFORMES 

365 

Table  97.    Continui 

;d. 

Egg  diameter 

Species 

(mm) 

Oil  globules 

Chorion  surface 

Source(s) 

Family  Cyprinodontidae 

Cubanichlhvs  cuhensis 

ca.  1.0 

? 

present 

Mayer,  1933 

Cubanichlhvs  pengellevi 

1.2-1.4 

few 

filaments  in  tuft 

Foster,  1969 

Aphamus  fasciatus 

7 

7 

filaments 

Breder  and  Rosen,  1 966 

Aphanius  memo 

1.4-1.5 

ca.  6,  medium 

filaments 

Mazza,  1902 

Cyprinodon  macularius 

ca.  2.0 

7 

7 

Constantz,  1981 

Cvprinodon  ncvadensis 

1.3-1.4 

7 

7 

Constantz,  1981 

Cyprinodon  variegatus 

1.1-1.7 

one  large 
many  minute 

filaments 

Foster,  1967;  Fanara, 
1964;  Wickler,  1959 

Floridichthys  carpio 

1.4 

7 

1-2  filaments,  long 

Kaill,  1967 

Jordanella  Jloridae 

1.3-1.4 

7 

filaments 

Henzelmenn,  1930; 
Foster.  1967;  Kaill,  1967 

Jordanella  pulchra 

1.0 

7 

7 

Cassel,  1981 

'^^^^^^^^^^^^K^ 

■-L- 

^                 "^WHR 

^^^if^ 

■''^■^-'^^^*^ 


Fig.  190.     Scanning  electron  micrographs  of  the  chonon  surface  of  the  cypnnodonlid  Cyprinodon  aharezi  (A,  B)  and  the  fundulid  Fundulus 
luciac  (D,  E).  Oil  droplets  are  shown  for  the  fundulids  Lucania  parva  (C)  and  Fundulus  n.  sp.  from  Bermuda  (F). 


366 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES -AHLSTROM  SYMPOSIUM 


Fig.  191.     Scanning  electron  micrographs  of  the  chorion  surface  of  the  fundulids  F.  majalis  (A,  B),  Fundulus  n.  sp.  from  Bermuda  (C),  the 
rivulid  Rinilus  mannoratus  (D,  E)  and  the  profundulid  Profundulus  punctalus  (F). 


ported  by  an  apparent  difference  in  the  manner  in  which  hatch- 
ing is  delayed.  Fundulus  confluentus  can  hatch  after  three  months 
of  "latency"  or  postponement  of  hatching  (Harrington,  1959; 
Harrington  and  Haeger,  1958)  while  the  embryos  continue  to 
grow  and  utilize  yolk  reserves.  Delayed  hatching  is  probably 
typical  for  many  North  American  fundulids,  as  seen  in  F.  het- 
eroclitus  (Taylor  et  al.,  1977)  and  Adinia  xenica  (Koenig  and 
Livingston,  1976).  The  incubation  period  is  known  to  be  influ- 
enced by  temperature  (see  Gabriel,  1944)  and  dissolved  oxygen 
(DiMichele  and  Taylor,  1980).  During  diapause,  which  occurs 
in  the  annual  killifishes  (Wourms,  1972a,  b,  c)  hatching  may  be 
delayed  for  up  to  six  months  in  nature  and  possibly  longer  than 
a  year  under  extreme  conditions.  During  this  time  growth  does 
not  occur,  cardiac  activity  ceases  and  the  yolk  is  not  depleted. 
The  length  of  the  incubation  period  may  be  controlled  by  tem- 
perature, photoperiod,  desiccation  and  oxygen  tension  cues  (see 
Matias,  1982). 

The  embryonic  development  of  several  aplocheilids  (Aploch- 
eilus),  a  rivulid  {Rivulus)  and  two  fundulids  (Adinia  and  F. 


heteroclitus)  has  been  described  in  detail  (Table  96).  Some 
authors  have  placed  special  systematic  significance  on  the  pat- 
tern of  vitelline  circulation  of  the  embryo  in  cyprinodonti forms 
(Foster,  1967;  Hubbs  and  Bumside,  1972)  and  other  atherino- 
morphs  (White  et  al.,  this  volume).  The  viviparous  poeciliids, 
anablepids,  jenynsiids  (placed  in  the  Anablepidae  by  Parenti, 
1981)  and  goodeids  have  a  variety  of  modifications  for  receiving 
nourishment  during  development  (reviewed  by  Wourms,  1981). 
The  phylogenetic  significance  of  independent  development  of 
viviparity  in  several  cyprinodontiform  lineages  is  discussed  in 
detail  by  Parenti  (1981). 

Larvae 

The  larvae  of  oviparous  cyprinodontiforms  are  incompletely 
known  (Table  96)  despite  the  fact  that  many  of  them  are  avidly 
bred  by  aquarium  hobbyists.  All  of  those  known  lack  the  preanal 
finfold  characteristic  of  the  beloniforms  (except  exocoetids) 
(Collette  et  al.,  this  volume)  and  have  a  longer  preanal  length 
than  the  atheriniforms  (White  et  al..  this  volume).  In  all  cy- 


ABLE:  CYPRINODONTIFORMES 


367 


Mur 


^(S;^ 


Fig.  192.     Larvae  of  (A)  the  aplocheilid  Nothobranchius  eggersi,  3.1  mm  SL;  (B)  the  nvulid  Rivulus  marmoratus,  4.6  mm  SL;  and  (C-E)  the 
fundulid  Funduliis  n.  sp.  from  Bermuda.  6.0  mm  SL. 


368 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


prinodontiforms  that  have  been  studied  the  caudal  fin  rays  form 
first  (Fig.  192)  and  often  this  occurs  before  hatching  (Foster, 
1967). 

Within  the  oviparous  cyprinodontoid  killifishes  the  presence 
and  location  of  melanophores  as  well  as  the  relative  location  of 
the  dorsal  finfold  may  be  useful  systematic  characters  (Foster, 
1968).  In  most  fundulid  larvae  the  dorsal  finfold  originates  pos- 
terior to  the  origin  of  the  anal  finfold  (Foster,  1967;  Fig.  192) 
with  the  possible  exception  of  Lucania parva  (see  Hardy,  1 978a). 
In  the  cyprinodontids  studied  however,  the  dorsal  finfold  orig- 
inates anterior  to  the  anal  finfold  (Foster.  1967).  The  larvae  of 
most  fundulids,  the  aplocheilid,  Nothobranchius  eggersi  and  the 
rivulid,  R.  marmoratus  also  possess  three  rows  or  stripes  of 
melanophores  (middorsal,  midlateral  and  midventral)  on  the 
body  (Fig.  192).  This  characteristic  is  shared  by  the  beloniform 
Oryzias  latipes  and  some  atheriniforms  (Martin  and  Drewry, 
1978)  and  suggests  that  this  character  may  be  symplesiomorphic 
within  the  Atherinomorpha.  In  cyprinodontids  these  rows  of 
melanophores  are  lacking  and  the  existing  melanophores  are 
scattered  evenly  over  the  body  or  appear  as  saddle-shaped  groups 
of  melanaphores  on  the  dorsolateral  surface  of  the  body  (see 
Foster,  1967;  Hardy,  1978a). 


Summary 

The  early  life  history  of  cyprinodontiforms  appears  to  offer 
many  possibilities  for  elaborating  on  their  phylogeny.  Several 
authors  (Rosen  and  Parenti,  1981;  Collette  et  al.,  this  volume; 
White  et  al.,  this  volume)  have  pointed  out  the  usefulness  of 
early  life  history  characters  in  defining  the  monophyletic  nature 
of  the  Atherinomorpha  and  the  orders  within.  Although  studies 
of  the  early  life  history  of  the  Cyprinodontiformes  are  not  as 
far  along,  they  may  offer  more  potential  for  several  reasons. 
First,  Foster  (1967,  1968)  has  already  pointed  out  the  value  of 
early  life  history  characters  in  resolving  the  phylogeny  of  the 
group.  Second,  based  on  this  review,  both  egg  morphology  and 
larval  characters  vary  within  the  group  and  thus  seem  to  offer 
real  promise  for  assessing  relationships.  Third,  many  killifishes 
are  easily  maintained  and  will  reproduce  in  aquaria  so  that  study 
material  should  be  easily  obtainable,  especially  given  their  pop- 
ularity in  the  aquarium  trade. 

Biological  Sciences  and  Center  for  Coastal  and  Envi- 
ronmental Studies,  Doolittle  Hall,  Rutgers 
University,  New  Brunswick,  New  Jersey  08903. 


Lampriformes:  Development  and  Relationships 
J.  E.  Olney 


THE  order  Lampriformes  (=Lampridiformes,  see  Robins  et 
al.,  1980)  is  composed  of  approximately  21  species  (Table 
98)  of  pelagic,  marine  fishes  with  worldwide  distribution  ex- 
cluding polar  seas.  Highly  evolved  and  extremely  divergent  in 
form  and  lifestyle,  these  species  occupy  meso-  and  epipelagic 
habitats  and  have  attained  a  remarkable  degree  of  specialization, 
of  which  the  most  notable  examples  are:  the  pectoral  muscu- 
lature of  Laniphs  (Rosenblatt  and  Johnson,  1976);  the  unique 
feeding  mechanism  of  S7r/e/'/7C'n«(Pietsch,  1978a);  the  ribbon- 
like body  form,  specialized  integument  and  rotating  eye  of  the 
trachipterids  (Walters,  1963;  Haedrick,  1974;  Oelschlager, 
1976a);  the  "horn"  of  Eiimecichthys  (Fitch,  1966;  Oelschlager, 
1979);  and  the  cephalopod-like  ink  gland  of  the  lophotids  and 
Radiicephalus  (Walters  and  Fitch,  I960;  Fitch  and  Lavenberg, 
1968;  Harrison  and  Palmer,  1968;  Saldanha  and  Pereira,  1977; 
and  others).  By  far  the  most  impressive  species  of  the  group  is 
the  oarfish,  Regalecus  glesne.  which  attains  lengths  of  over  8m, 
possesses  a  crimson  dorsal  fin  and  cockscomb-like  anteriormost 
dorsal  rays  and  is  the  probable  basis  for  many  historical  sightings 
of  sea  monsters  (Fitch  and  Lavenberg,  1968). 

Regan  (1907,  1924)  first  suggested  a  relationship  between 
Lophotus.  Eumecichthys,  Lampris,  V'elifer  and  Stylephorus.  all 
on  the  basis  of  the  common  possession  of  peculiar  characteristics 
of  the  protractile  mouth  and  assigned  these  genera  to  a  new 
order,  the  Allotriognathi  (from  the  Greek,  meaning  "strange 
jaw").  Presently,  the  order  consists  of  12  genera  (Velifer.  Me- 


tavelifer,  Lampris.  Zu,  Desmodema,  Trachipterus,  Radiicepha- 
lus, Lophotus.  Eumecichthys.  Stylephorus.  Regalecus  and 
Agrostichthys)  comprising  seven  families  (Table  98). 

Two  conflicting  proposals  exist  for  the  allocation  of  these 
fishes  and  nomenclatural  inconsistencies  persist.  Oelschlager 
(1976a,b,  1978a,  b,  1979;  also  see  Palmer,  1973)  retains  Regan's 
( 1 907)  ordinal  designation  and  defines  two  suborders  of  the 
Allotriognathi:  the  Bathysomi,  deep-bodied  fishes  with  sym- 
metrical caudal  fins,  well  developed  skeletons  and  musculature 
(represented  by  Lampris  and  the  veliferids);  and  the  Taenio- 
somi,  elongate  fishes  with  asymmetrical  caudal  fins,  weak  skel- 
etons and  musculature  (represented  by  Trachipterus.  Regalecus 
and  remaining  genera).  In  contrast.  Greenwood  et  al.  (1966) 
recognize  four  suborders  of  the  Lampriformes:  Lampridoidei, 
Veliferoidei,  Trachipteroidei  and  Stylephoroidei.  At  lower  taxo- 
nomic  levels,  Heemstra  (in  press)  considers  Lophotus  to  be 
monotypic  while  Briggs(l952)and  Oelschlager  (1 979)  recognize 
two  species  (Table  98).  In  addition,  a  number  of  nominal  species 
exist  within  the  genera  Regalecus.  Trachipterus  and  Lampris. 
Recently,  Heemstra  (in  press)  and  Heemstra  and  Kannemeyer 
(in  press)  have  treated  South  African  Lampriformes,  describing 
a  new  Zu  species  and  providing  synonymies  of  several  trachip- 
terids. In  general,  the  systematic  status  of  lampriform  fishes  is 
in  question  and  the  nomenclature  lacks  stability  owing,  in  part, 
to  the  rarity  of  examples  in  systematic  collections  and  the  fragile 
nature  of  these  fishes. 


OLNEY:  LAMPRIFORMES 


369 


Table  98.     Recorded  Meristics  of  Adult  Lampriform  Fishes.  Total  element  counts  are  reported  without  reference  to  ray  or  spine  designa- 
tion of  onginal  source.  Abbreviations  used  are:  ABS— absent;  PRE— precaudal  vertebrae;  CAUD— caudal  vertebrae;  TV  — total  vertebrae; 
DORS— dorsal  fin  rays;  ANAL— anal  fin  rays;  PECT— pectoral  fin  rays;  PELV— pelvic  fin  rays;  CAUD— caudal  fin  rays. 


pre 

CAUD 

TV 

DORS 

ANAL 

PECT 

PELV 

CAUD 

References 

Veliferidae 

Velifer  hypselopterus 

Metavelifcr  nnilliradialus 

16 

16 

17-18 
17-18 

33-34 
33-34 

34-36 
41-45 

25-26 
33-36 

15-16 
15-16 

7-8 
9 

36 
36 

Heemstra  (in  press);  Regan 
(1970);  Walters  (1960) 

Heemstra  (in  press);  Ste- 
phenson (1977);  Walters 
(1960) 

Lampridae 
Lampris  guttatus 

L.  immaculatus 
Trachipteridae 
7.U  cristatus 


T.  jacksonensis 

Radiicephalidae 
Radiicephalus  elongalus 


Lophotidae 
Lophotus  lacepedei 

L.  capcUci 
Eunwcichthys  fiski 

Stylephoridae 
Stylephorus  chordatus 

Regalccidae 
Regalecus  glesne 

Agrostichthys  parkeri 


21 


25 


46 


43 


48-52   33-42  21-24   13-17 


52-56   35-38   23-24   12-14 


22-24         39 


Z.  elongatus 

29-31 

— 

84-87 

Desmodema  polysliclum 

18-20 

- 

71-74 

D.  lorum 

21-25 

- 

106-111 

Trachiplcrus  fukuzaki 

25-28 

- 

69-72 

T.  allivetis 

35-40 



90-94 

T.  Irachiplerus 

35-39 

— 

84-96 

T.  arcticus 





99-102 

T.  ishikawae 

— 

— 

73 

62-69        120-150       ABS        10-12        3-7 


142-147  ABS  11-12  7 

120-124  ABS  12-14  ABS 

197  ABS  12-14  ABS 

153-174  ABS  11-13  5 


165-184       ABS        10-11 
145-185       ABS         9-11 


133-168       ABS 


9-11 
13 


31-34 


81-83 


166-173       ABS       13-14 


36-39      77-79      114-121       152-160        6-7 


9-10 


170 


6-7 

5 

5-6 


_  ,_  124-153  206-263  12-20  14-17  5 

-  -      -      238      18  14-16  6 
56  101  151-200  310-392  5-9  13-15  2-3 

-  -      50  115-122  16-17  10-11  - 
_  _  143-151  260-412  ABS  12-13  1 


400  ABS         8-11  1 


30-32  Herald  (1939);  McKenzie 

and  Tibbo  (1963), 
Regan  (1907a) 

—  Parin  and  Kukuyev  (1983) 

8-12  +  1-5       Fitch  (1964);  Heemstra 
and  Kannemeyer  (in 
press);  Palmer  (1961 ); 
Walters  and  Fitch  (1960, 
1964) 
12  +  5  Heemstra  and  Kannemeyer 

(in  press) 
7-10  Rosenblatt  and  Butler 

(1977) 
4-7  Rosenblatt  and  Butler 

(1977) 
7-9  +  6-7        Fitch  (1964,.  1967);  Walters 

and  Fitch  (1964) 
7-8  +  6  Fitch  (1964) 

8+5  Heemstra  and  Kannemeyer 

(in  press);  Palmer  (1961) 
8  +  5-6       Palmer  (1961) 

—  Heemstra  and  Kannemeyer 

(in  press);  Nishimura 
(1964);  Smith  (1956a) 

—  Heemstra  and  Kannemeyer 

(in  press) 

4  +  7  Harrison  and  Palmer 

(1968);  Heemstra  and 
Kannemeyer  (in  press); 
Karrer(1976) 

16-17  Heemstra  (in  press);  Sal- 

danha  and  Pereira 

(1977) 
17  Briggs  (1952);  Coin  and 

Erdman(1951) 
12-13  Abe  (1954);  Fitch  (1966); 

Heemstra  (in  press); 

Walters  and  Fitch  (1960) 

5-6  +  2  Pietsch  (1978a);  Regan 

(1924) 

3-4  Heemstra  (in  press);  Tru- 

nov  (1982);  Walters  and 
Fitch  (1960) 
2  Heemstra  (in  press);  Tru- 

nov  (1982);  Walters  and 
Fitch  (1960) 


F.g.  193.     Eggs  and  larva  (A-C)  of  Trach.pterus  sp.  (larva,  7.6  mm  NL)  after  Mito  (1961b).  (D)  Larva  of  Lophotus  sp.  (12.1  mm  NL)  after 
Sanzo(1940). 


OLNEY:  LAMPRIFORMES 


371 


Development 

Walters  and  Fitch  (1960),  Breder  and  Rosen  (1966),  Palmer 
(1973),  Nielsen  (1973)  and  Moser  (1981)  have  summarized  the 
state  of  knowledge  of  the  early  life  history  of  lampriform  fishes. 
Eggs  and  larvae  of  the  Veliferidae,  Radiicephalidae  and  Style- 
phoridae  are  unknown,  although  Karrer  (1976)  has  mentioned 
ripe  ovarian  eggs  of  Radiicephalus  clongatus.  Harrison  and 
Palmer  (1968)  presented  meristic  and  morphometric  data  on  a 
154  mm  SL  R.  elongatus  termed  a  postlarva  and  Regan  (1924) 
figured  a  26  mm  SL  larval  Stylephorus  chordatus.  Little  data  on 
young  stages  of  the  Lampridae  are  available.  Ehrenbaum  ( 1 905- 
1 909)  and  Gudger  ( 1 930)  presumed  pelagic  eggs  based  on  ovar- 
ian examination;  Gudger  (1930),  D' Ancona  (1933b) and  Oelsch- 
lager  (1976b)  figured  juvenile  stages  of  Lampris  giutatus;  and 
Parin  and  Kukuyev  (1983)  illustrated  a  young  specimen  of  L. 
irnmaculatus.  Within  the  Lophotidae,  larvae  of  Eumecichthys 
are  unknown  while  Fitch  (1966)  reported  on  ovarian  eggs  in  E. 
fiskt  and  Parin  and  Pokhilpkaya  (1968)  figured  juvenile  stages. 
Sanzo  ( 1 939b,  1 940)  and  Sparta  (1954)  have  described  eggs  and 
early  larvae  of  a  species  oi  Lophotus  considered  by  Oelschlager 
(1979)  to  be  L.  lacepedei.  Eggs  and  larvae  of  trachipterid  and 
regalecid  fishes  have  received  considerable  attention  although 
early  life  history  stages  of  Agrostichthys  and  Desmodema  are 
unknown.  Eggs  and  early  larvae  of  Zu  cristatus  were  described 
by  Sanzo  (1918),  Sparta  (1933)  and  Olney  and  Naplin  (1980). 
Eggs  and  larvae  of  Trachipterus,  probably  representing  several 
species,  were  described  by  Emery  (1879),  Lo  Bianco  (1908a), 
Jacino  (1909),  Ehrenbaum  (1905-1909),  Sparta  (1933),  Mito 
(1961b)  and  Sardou  ( 1 966).  Eggs  and  larval  stages  of  Regalecus 
were  figured  and  described  by  Sanzo  (1925),  Sparta  (1933)  and 
Robertson  (1975a).  In  summary,  published  information  on  the 
development  of  eggs  and  larvae  of  four  of  the  1 2  lampriform 
genera  is  available.  In  the  following  discussion,  these  published 
data  as  well  as  additional  material  are  utilized  to  summarize 
the  important  characteristics  of  eggs,  larvae  and  young  of  lam- 
priform fishes  and  provide  illustrations  of  larvae  of  four  addi- 
tional genera. 

Egg  and  embryonic  morphology.  —  Data  on  morphology  and  de- 
velopment of  lampriform  eggs  are  incomplete  (Table  99)  but 
indicate  that  eggs  are  large  (1.7-4.0  mm  egg  diameter,  range 
excludes  measurements  of  ovarian  eggs,-see  Table  99),  spherical, 
pelagic,  often  brightly  colored  (generally  in  amber,  pink  or  red 
hues)  and  possess  thick,  resilient  chorions.  Up  to  three  weeks 
may  be  required  in  incubation  ( 1 8-20  days  for  R.  glesne,  Sparta, 
1933).  As  a  result,  eggs  are  distinctive  and  easily  recognized  in 
plankton  collections  (Fig.  1 93B,  C)  especially  in  advanced  stages 
of  development  (Orton,  1955a;  Olney  and  Naplin,  1980). 
Sanzo  (1940)  reported  both  homogeneous  (Lophotus)  and 
segmented  yolks  (Zu  and  Regalecus)  but  recent  observations 
indicate  homogeneous  yolks  in  all  known  forms  (Robertson, 
1975a;  Olney,  unpublished  data).  Egg  diameters,  presence  or 
absence  of  oil  droplets,  chorionic  ornamentation  and  micro- 
structure  may  delimit  some  species  (Table  99).  Scanning  elec- 
tron micrographs  of  cross-sections  of  the  chorions  of  Zu  cris- 
talus  and  an  unidentified  trachipterid  species  (Fig.  194)  indicate 
variability  in  chorion  thickness  and  layering  which  may  be  of 
systematic  value.  In  general,  however,  confirmed  identification 
of  lampriform  eggs  requires  late  stages  with  advanced  embryos 
(Olney  and  Naplm,  1980). 


™*       JM 

[Hi 

m 

■1 

K 

leKv 

g_ 

^mm^^ 

^■i 

12138 

VMS 

X26e0 

10U 

565 

Fig.  1 94.  Scanning  electron  micrographs  of  chorionic  microstructure 
in  lampriform  eggs.  (A)  trachipterid,  chorion  thickness  1 1.04  ^m.  (B) 
Zu  cristatus,  chorion  thickness  13.3  ^m. 


Lampriform  embryos  exhibit  precocious  development.  In 
Trachipterus.  Zu,  Lophotus  and  Regalecus.  anterior  dorsal  rays, 
pelvic  rays,  distinctive  pigment  and  total  myomeres  are  appar- 
ent and  distinguish  these  forms  (Sparta,  1933;  Mito,  1961b; 
Sanzo,  1940;  Olney  and  Naplin,  1980).  Some  disparity  exists, 
however,  in  descriptions  of  late  embryos.  Sparta  (1933)  depicts 
late  embryonic  R.  glesne  with  anterior  elements  reduced  but  the 
fourth  elongate  while  Robertson  (1975a)  figures  R.  glesne  em- 
bryos off  New  Zealand  with  an  elongate  first  element  followed 
by  three  reduced  rays. 

Larval  morphology.  —  Al  hatching,  larvae  of  lampriform  fishes 
possess  a  number  of  distinctive  characteristics  including;  well 
developed,  protrusible  jaws;  diflierentiated  guts  with  an  open 
lumen  and  little  or  no  yolk  material;  elongate  anterior  dorsal 
elements  which  insert  between  the  posterior  eye  margin  and  the 
shoulder  and  are  usually  ornamented  with  broad,  spatulate  and 


372 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  99.    Some  Characteristics  of  Eggs  of  Lampriform  Fishes. 


Egg 

diameler 

Descriplive 

Species 

(mm) 

features 

Reference 

L.  gultatus 

— 

Ovarian;  thick 
chorion  with 
amber  tint 

This  Study 

Z.  cristatus 

2.17-2.27 

Thick  chorion 

Sanzo,  1918; 

with  amber  tint; 

Olney  and 

oil  globule  ab- 

Naplin, 1980 

sent 

Trachipterus  sp. 

1.78 

Yolk  with  scat- 
tered melano- 
phores 

Mito.  1961b 

T.  trachipterus 

2.9-3.2 

Thick  chorion;  oil 
globule  absent; 
yolk  scattered 
melanophores 

Sparta,  1933 

R.  glesne 

2.4-2.5 

Numerous,  scat- 
tered oil  glob- 
ules; yolk  with 
scattered  mela- 
nophores 

Sanzo,  1925 

R.  glesne 

3.25-4.05 

Chorion  with  pink 

Robertson, 

tint 

1975a 

Lophotiis  sp. 

2.48-2.64 

Oil  globules  ab- 
sent; chorion  or- 
namented with 
small  spines 

Sanzo,  1940 

E.fiski 

1.5-2.0 

Ovarian;  transpar- 
ent 

Fitch,  1966 

Table  100.    PatternsofPterygiophoreInterdigitationin  Anteri- 
or Interneural  Spaces  of  Young  Lampriform  Fishes.  Within  inter- 
neural  spaces,  P  indicates  a  predorsal  element  and  numerals  indicate 
numbers  of  pterygiophores. 


highly  pigmented  serial  swellings;  stout,  well  developed  pelvic 
eleinents  often  with  fleshy  sheaths  and  highly  pigmented  ter- 
minal swellings;  and  snout  to  vent  lengths  approximately  40- 
60%  NL  (Figs.  193,  195).  Larvae  of  Metavelifer  and  Lampris 
are  slender  at  small  sizes  but  Lampris  larvae  (and  presumably 
veliferids)  rapidly  increase  in  body  depth  (Fig.  196).  By  10.6 
mm  SL  (Fig.  197),  larval  Lampris  have  assumed  the  character- 
istic adult,  deep-bodied  form.  Larvae  of  the  remaining  genera 
are  also  slender  at  hatching  but  become  rapidly  elongate  with 
growth.  In  these  taxa,  gut  length  may  vary  during  ontogeny, 
increasing  to  80-90%  SL  in  Lopholus.  Eumecichthys  and  Re- 
galecus.  Gut  length  at  transformation  distinguishes  these  genera 
from  7ai.  Trachipterus,  Dcsmodema  and  Radiicephalus. 

Anterior  dorsal  rays  are  supported  by  a  fleshy  base  in  early 
larval  stages  (Figs.  193,  195)  and  by  modified  radials  first  ap- 
pearing as  cartilage.  The  fleshy  base  and  associated  radials  sup- 
porting these  rays  are  sometimes  damaged  in  capture  and  torn 
away  from  the  cranium.  This  artifactual  condition  is  occasion- 
ally referenced  in  older  literature  as  a  "nuchal  lobe""  (Hubbs, 
1925:475)  and  "nuchal  pennant"  (Walters  and  Fitch,  1960:442- 
443).  In  larval  trachipterid,  lophotid  and  regalecid  fishes,  the 
first  two  dorsal  elements  are  supported  by  elongate,  stout  radials, 
the  second  of  which  serially  supports  succeeding  radials  of  vary- 
ing number  (Table  100).  These  elements  interdigitate  in  inter- 
neural spaces  and  pterygiophore  interdigitation  patterns  vary  at 
the  generic  level  (Rosenblatt  and  Butler,  1977;  Table  100).  Law- 
pris  and  Metavelifer  possess  a  predorsal  element  which  inter- 
digitates  in  the  first  neural  space.  These  genera  (and  presumably 
Velifer)  are  unique  in  the  possession  of  this  character  (Table 
100).  In  addition,  the  morphology  of  the  anterior  modified  ra- 
dials varies  with  ontogeny.  The  rostrum  or  "horn"  of  Eume- 
cichthys e\ongaies  duringgrowth  to  twice  its  original  length  (Far- 


Inlemeural  space 


Species 


Z.  cristatus 
Trachipterus  sp. 
R.  glesne 
Desmodema  spp. 
R.  elongatus 
M.  muitiradiatus 
L.  gultatus 


1 

7-9 

1 

2 

1 

8-9 

1 

2 

1 

11 

2 

2 

7 

T 

1 

13-14 

1-2 

1-2 

P  +  1 

1 

1 

1 

P  +  1 

1 

1 

1 

in  and  Pokhil'skaya,  1968).  Oelschlager  (1976a)  considers  the 
lophotid  "horn'"  to  be  supported  by  modifications  of  the  frontals 
and  the  supraoccipital  (termed  the  "fronto-occipital  carina"), 
however  these  highly  modified  dorsal  rays  are  likely  supported 
by  dorsal  fin  radials  which  may  fuse  to  cranial  bones  in  adults. 
In  Dcsmodema.  anterior  dorsal  rays  are  elongate  in  juveniles 
but  lost  at  transition  (Rosenblatt  and  Butler,  1977).  Presumably, 
pterygiophores  supporting  these  rays  and  interdigitating  anterior 
to  the  first  neural  spine  (Table  100)  are  retained  in  adults. 

Development  of  pelvic  fin  elements  is  precocious  in  all  known 
larval  Lampriformes  (Figs.  193,  195-200)  and  characterized  by 
reduction  of  ray  number,  length  or  both  in  some  genera.  In  S. 
chordalus  (Figs.  199  and  200),  a  single  elongate  pelvic  element 
in  early  larvae  becomes  increasingly  long  and  stout  with  de- 
velopment, persisting  until  around  70  mm  SL.  Adult  5.  chor- 
datus  (Regan,  1924;  Table  98)  lack  a  pelvic  fin.  A  similar  de- 
velopmental sequence  may  occur  in  Desmodema  which  loses 
pelvic  elements  by  173  mm  SL  (Rosenblatt  and  Butler,  1977) 
and  in  R.  elongatus  (Harrison  and  Palmer,  1968).  In  Regalecus 
and  Agrostichthys  {Oe\sch\a.ge:v,  1978a)  the  pelvic  fin  is  retained 
as  a  persistent  larval  floatation  device  that  also  serves  in  loco- 
motion and  taste  perception. 

The  pectoral  fin  is  the  last  to  complete  difTerentiation  in  larval 
lampriformes.  In  our  material,  pectoral  rays  are  incompletely 
developed  at  18.4  mm  SL  in  R.  elongatus  (Fig.  198);  21.4  mm 
SL  in  5.  chordatus  (Fig.  200);  and  29  mm  SL  in  Trachipterus 
sp.  Pectoral  development  is  most  rapid  in  L.  gultatus  which 
possesses  adult  counts  by  10.6  mm  SL  (Fig.  197).  Adults  ofthis 
species  possess  a  strongly  developed,  lunate  pectoral  (Rosenblatt 
and  Johnson,  1976)  which  may  also  be  important  in  locomotion 
of  larvae.  Pectoral  morphology  and  insertion  vary  consider- 
ably among  lampriform  genera  and  are  of  systematic  value, 
although  no  comprehensive  treatment  exists. 

Among  lampriform  genera,  caudal  morphology  exhibits  the 
greatest  potential  for  taxonomic  and  systematic  evaluation.  Ro- 
senblatt and  Butler  (1977)  have  demonstrated  the  utility  ofthis 
character  in  distinguishing  juvenile  and  adults  of  Trachipterus 
and  Desmodema  and  the  details  of  caudal  morphology  (Table 
98  and  see  Gosline,  1961;  Hulley  and  Rau,  1969;  Oelschlager, 
1974;  Patterson,  1968;  Pietsch,  1978a;  and  Rosen,  1973  for 
illustrations  of  caudal  skeletons  in  various  lampriform  genera) 
clearly  delimit  all  other  taxa,  with  the  possible  exception  of 
Agrostichthys  {OehcMager,  1978b).  Diff"erentiation  of  caudal 
elements  occurs  early  in  development,  rendering  caudal  mor- 
phology an  important  larval  identification  criterion.  Although 
full  developmental  series  are  not  available  for  most  forms,  the 


OLNEY:  LAMPRIFORMES 


373 


Fig.  195.     Larvae  of  Zu  cnstaius.  (A)  6.5  mm  NL  and  Regalecus  glesne,  (B)  5.4  mm;  (C)  45.8  mm  SL,  all  after  Sparta  1933. 


374 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  196.     Larvae  of  Lampris  gultalus.  (Upper)  4.7  mm  NL,  MCZ  58990;  (Lower)  8.6  mm  SL,  MCZ  58989. 


OLNEY:  LAMPRIFORMES 


375 


Fig.  197.     Larva  of  Lampns  gutlalus  10.6  mm  SL.  MCZ  58991. 


caudal  is  complete  (or  nearly  so)  by  8.6  mm  SL  in  L.  guttatus 
and  1 2-20  mm  SL  in  R.  elongatus.  S.  chordatus,  Zu  cristatus. 
Trachipienis  spp.  and  R.  glesne. 

The  highly  protrusible  jaws  of  lampriform  fishes  develop  pre- 
cociously and  jaw  structures  vary  from  the  long,  tubular  mouth 
of  S.  chordatus  (¥\gs.  1 99  and  200;  Pietsch  1978a)  to  only  mod- 
erately specialized  in  I'elifer.  All  lampriform  genera  possess 
prcmaxillae  with  long  ascending  processes  which  fit  withm  the 


nasal  and  ethmoid  regions  and  slide  forward  during  jaw  pro- 
traction. Larval  lampriforms.  especially  Stylephorus  (Figs.  199 
and  200).  are  easily  recognized  by  this  feature  although  these 
upper  jaw  specializations  may  not  be  unique  to  lampriform 
genera  (Rosen,  1973).  In  trachipterid,  radiicephalid  and  regal- 
ecid  fishes,  the  premaxilla  has  a  high,  broad  ascending  process 
which  is  often  conspicuous  in  capture-damaged  larvae. 

Lampriforms  are  highly  pigmented  in  all  life  history  stages 


376 


ONTOGENY  AND  SYSTEMATICS  OF  nSHES-AHLSTROM  SYMPOSIUM 


Fig.  198.  Larvae  of  Melavelifer  multiradiatus.  (Upper)  5.7  mm  NL,  MCZ  59717;  and  Radiicephalus  elongatus.  (Lower)  18.4  mm  SL,  ZMUC 
uncatalogued.  The  vent  is  indicated  by  an  arrow.  The  posterior  portion  of  the  ink  gland  is  seen  as  a  concentration  of  melanophores  along  the 
ventral  margin  just  posterior  to  the  vent. 


and  larval  pigment,  especially  in  the  form  of  melanophores 
present  laterally  and  along  the  dorsal  and  ventral  margins,  is 
useful  in  identification  of  some  genera  (Figs.  193,  195-200). 
Melanophores,  concentrated  on  spatulate  swellings  in  elongated 


dorsal  and  pelvic  rays  help  to  distinguish  larval  lampriforms 
although  care  should  be  taken  since  elongate,  sometimes  pig- 
mented appendages  are  found  in  the  larvae  of  a  number  of 
unrelated  taxa  (Govoni  et  al.,  1984).  Among  these  taxa,  how- 


OLNEY:  LAMPRIFORMES 


377 


Fig.  199.     Larvae  of  Slylephorus  chordatus.  (Upper)  3.8  mm  NL,  MCZ  59718;  (Lower)  7.6  mm  SL,  MCZ  59719. 


ever,  only  lophiiform,  bothid,  zeid  and  serranid  larvae  have 
elongate  dorsal  and  pelvic  elements. 

Specialized  ink  glands  filled  with  dark  brown  fluid  are  char- 
acteristic of  lophotid  and  radiicephalid  fishes,  and  are  conspic- 
uous in  larval  R.  elongatus  (Fig.  198),  young  Lophotus  (Fig. 
201)  and  presumably  Eumectchthys  (Walters  and  Fitch,  1960; 
Harrison  and  Palmer,  1968).  Although  this  unpaired,  internal 
structure  is  not  considered  a  larval  specialization,  its  early  ap- 
pearance in  larval  R.  elongatus  and  juvenile  Lophotus  suggests 
that  the  ink  gland  may  be  functional  in  young  fishes. 

Development  from  larval  to  juvenile  stages  is  gradual  in  Lam- 
pris  (Figs.  196  and  197;  Oeschlager,  1976b)  but  ontogenetic 
variability  is  marked  and  abrupt  in  trachipterid  genera  and  Sty- 
lephorus  chordatus.  This  rapid  transition  from  prejuvenile  to 
juvenile  morphology  has  been  termed  metamorphosis  in  Des- 
wof^ewfl  (Rosenblatt  and  Butler,  1977)  and  Trachlptenis  (Huhhs, 
1925).  In  D.  polystictum.  metamorphosis  is  characterized  by 
changes  in  ventral  profile,  elongation  of  caudal  vertebrae,  in- 


crease in  eye  size,  eruption  of  mandibular  teeth,  and  loss  of 
spots,  pelvic  fins  and  the  posterior  nostril  (Rosenblatt  and  But- 
ler, 1977).  Examination  of  S.  chordatus  material  indicates  a 
similar  rapid  transition,  characterized  by  the  loss  of  elongate 
dorsal  rays,  three  ventral  caudal  rays  and  stout  pelvic  elements 
and  by  a  marked  change  in  eye  morphology  from  a  normal, 
non-telescopic  eye  to  the  specialized  adult  condition  (Pietsch, 
1978a).  Similar  metamorphic  change  may  occur  in  other  lam- 
priform  taxa,  however  full  developmental  series  are  not  avail- 
able. 

Meristics. —  Mtn%X\c  variability  is  useful  in  delimiting  lampri- 
form  taxa  (Tables  98,  100).  Precaudal,  caudal  and  total  vertebral 
counts  distinguish  genera  and  sometimes  species  (i.e.,  D.  poly- 
stictum vs  D.  lorum;  T.  fuku:aki  vs  T.  altivelis)  and  total  myo- 
mere counts  can  be  used  to  identify  early  larvae  (Olney  and 
Naplin,  1980).  Total  vertebral/myomere  counts  of  less  than  53 
characterize  Lampris.  Slylephorus  and  veliferids  and  are  the 


378 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  200.     Larvae  oi  Stylephorus  chordalus  21.4  mm  SL,  ZMUC  uncataloged. 


primary  basis  for  the  present  identification  of  larval  L.  guttatus 
(Figs.  196  and  197).  Aletavelifer  multiradialits  {Fig.  198)  and  5. 
chordalus  (Figs.  1 99  and  200).  Total  vertebral/myomere  counts 
range  from  62-200  (Table  98)  in  trachipterid,  radiicephalid, 
lophotid  and  regalecid  fishes,  but  care  should  be  taken  since 
these  elongate  forms  are  often  damaged  in  capture  and  the  pos- 
teriormost  myomeres  are  difficult  to  discern  in  larvae. 

Rays  of  median  fins  are  either  stout,  unsegmented,  spine-like 
elements  or  typical  soft  rays.  Previous  researchers  have  been 
inconsistent  in  their  treatment  of  these  elements  as  spines  or 
rays  and  ontogenetic  variability  likely  exists.  For  these  reasons, 
total  element  counts  are  reported  without  reference  to  spine  or 
ray  designation  (Table  98).  Furthermore,  because  of  the  lack  of 
developmental  series  in  collections,  few  data  exist  on  the  se- 
quence of  development  of  these  and  other  fin  elements.  As  a 
result,  dorsal  element  counts  delimit  the  genera  of  veliferid  fishes 
for  example  (Table  98),  but  are  not  developed  in  a  5.7  mm  NL 
Hawaiian  specimen  (Fig.  198).  Identification  of  this  specimen 
(Fig,  198)  as  M.  midtiradiatus  is  based  on  distributional  records 
(Walters,  1960;  Heemstra,  in  press).  In  larval  L.  guttatus  (Fig. 
196;  identification  based  on  distributional  records  of  Parin  and 
Kukuyev,  1983),  total  dorsal  elements  are  developed  by  8.6  mm 
SL  but  counts  indicate  some  overlap  with  veliferid  species  (Ta- 
ble 98).  In  the  elongate  forms,  dorsal  element  counts  are  less 
valid  identification  criteria  since  complete  differentiation  of  ele- 
ments occurs  late  in  development  (approximately  44  mm  SL  in 
Lophotus  lacepedei,  HML  6851;  83  mm  SL  in  Eumecichthys 
fiski.  MCZ  42264). 

The  absence  of  the  anal  fin  characterizes  adult  trachipterid 
and  regalecid  fishes  (Table  98),  but  its  absence  in  early  larvae 
cannot  be  considered  diagnostic.  In  genera  possessing  an  anal 
fin,  differentiation  of  elements  is  evident  in  our  material  at  18 
mm  SL  in  R.  elongatus  (Fig.  198);  7.5  mm  NL  in  S.  chordatus 
(Fig.  199);  33.5  mm  SL  in  E.fiski;  5.7  mm  NL  in  M.  multiradia- 
tus  (Fig.  198);  and  8.6  mm  SL  in  L.  guttatus  (Fig.  196).  Size  at 
first  differentiation  of  anal  elements  of  Lophotus  spp.  is  un- 
known but  total  element  counts  can  serve  to  delimit  young 
Radiicephalus  and  Lophotus  (Table  98;  compare  Figs.  198,201). 


Larvae  of  these  two  forms  can  be  easily  confused  due  to  the 
common  possession  of  the  distinctive  ink  gland  (Figs.  1 98,  20 1 ). 
Total  number  of  pectoral  rays  overlap  considerably  among 
lampriform  fishes  and  are  of  limited  diagnostic  value  (Table 
98).  Total  pelvic  elements  are  of  potential  use  in  identification 
but  ontogenetic  variability  is  great  and  care  should  be  taken 
until  descriptions  of  full  transformation  series  are  available. 
Total  caudal  elements  are  diagnostic  among  some  lampriform 
genera  (Table  98),  and,  as  previously  discussed,  details  of  caudal 
morphology  are  important  larval  identification  criteria. 

Relationships 

Our  present  knowledge  of  the  egg  and  larval  taxonomy  of 
lampriform  fishes  is  inadequate  to  the  task  of  fully  understand- 
ing phylogenetic  relationships.  Although  larval  stages  have  been 
described  for  8  of  1 2  genera  (those  of  Agrostichthys.  Desmo- 
dema,  Velifer  and  Eumecichthys  remain  unknown),  full  de- 
velopmental series  and  detailed  studies  of  developmental  os- 
teology and  morphology  are  lacking.  Among  those  taxa  for  which 
some  ontogenetic  data  are  available,  selected  characters  may 
elucidate  relationships  within  the  Lampriformes  and  between 
this  group  and  other  teleostean  fishes.  These  are:  (1)  Egg  mor- 
phology. The  distinctive  eggs  of  lampriforms  (Table  99,  Figs. 
193,  194)  are  likely  specializations  for  epipelagic  incubation 
(Breder,  1962)  and,  if  considered  a  derived  condition,  tend  to 
support  the  conclusion  of  a  common  ancestry  for  the  group. 
Complicating  this  interpretation  is  the  lack  of  data  on  egg  mor- 
phology in  all  lampriform  genera  (Table  99)  as  well  as  the  com- 
mon possession  of  somewhat  similar  (although  probably  inde- 
pendently evolved)  egg  morphology  in  other  fishes  (Orton, 
1955a);  (2)  Precocious  embryonic  development.  At  hatching,  all 
known  lampriform  larvae  possess  fully  developed,  protrusible 
jaws;  functional,  differentiated  guts;  and  pigmented  eyes.  This 
complement  of  precociously  developed  features  shared  by  lam- 
priform taxa  may  be  a  specialization  for  early,  successful  feeding 
in  the  low  prey  densities  of  the  epipelagic  habitat.  To  my  knowl- 
edge, only  exocoetoid  fishes  exhibit  similar  development;  (3) 
Elongate  anterior  dorsal  elements.  All  known  lampriform  larvae 


OLNEY:  LAMPRIFORMES 


379 


Fig.  20 1 .  Photomicrograph  of  the  posteriormost  ponion  of  the  ink 
gland  in  young  Lophotus  lacepedei  (HML  6851,  45  mm  SL).  Ink  gland 
is  seen  as  the  dark,  tubular  body  overlymg  the  hindgut  and  vent.  The 
vent  is  indicated  by  an  arrow. 


possess  elongate  anterior  dorsal  elements  which  are  ornamented 
with  spatulate,  pigmented  swellings  in  some  genera.  As  with 
other  fishes  with  ornamented  larval  appendages  (Govoni  et  al., 
1984),  variation  in  ornamentation  may  be  due  to  capture  dam- 
age. As  a  result,  the  absence  of  elaborate  ornamentation  in  early 
larvae  of  L.  guttatus  (Fig.  196),  M.  imdtiradiatus  (Fig.  198)  and 
S.  chordatus  (Figs.  199  and  200)  could  be  artifactual;  (4)  Pelvic 
fin  elements.  Precocious  appearance  of  ventral  fin  elements  which 
are  stout,  elongate  and  supported  by  well  developed  pelvic  bones 
is  observed  in  all  known  lampriform  larvae  (Figs.  193,  195- 
200).  Variation  among  genera  occurs  in  element  number  and 
fate  at  metamorphosis.  In  I'elifer and  Lampris,  pelvic  elements 
are  numerous  and  well  developed  in  adults.  In  remaining  genera, 
reductive  trends  are  evident  and  only  regalecids  retain  strongly 
developed  and  specialized  pelvic  fins  (Oelschlager,  1978a);  (5) 
Minute  spines  on  dorsal  elements.  Small,  laterally  projecting 
spines  are  conspicuous  in  some  young  lampriform  fishes  and 
have  been  reported  in  juveniles  by  Walters  and  Fitch  (1960: 
443),  Rosenblatt  and  Butler  ( 1 977:844),  and  Heemstra  and  Ken- 
nemeyer  (in  press).  In  our  material,  these  minute  spines  are 
conspicuous  in  larval  Zu.  Trachiptenis.  Regalecus.  Lophotus 
and  Radiicephalus  as  well  as  juvenile  specimens  of  Desmodema 
and  Eumecichthys.  Larval  Lampris.  Metavelifer  and  Stylephonis 
lack  these  characters;  (6)  Multiple  pterygiophores  interdigitate 
in  first  two  interneural  spaces.  In  all  our  lampriform  material, 
only  L.  guttatus  and  M.  multiradiatus  have  fewer  than  seven 
pterygiophores  which  interdigitate  in  interneural  spaces  1  and 
2  (Table  100).  In  addition,  only  Lampris  and  Metavelifer  (and 
presumably  Velifer)  possess  a  single  predorsal  element.  Inter- 
digitation  sequences  in  Velifer.  Lophotus.  Eumecichthys.  Sty- 
lephorus  and  Agrostichthys  are  unknown;  and  (7)  Metamorpho- 


sis. The  absence  of  abrupt  ontogenetic  transition  delimits  Lampris 
(and  presumably  veliferids)  from  other  lampriform  genera. 

The  distribution  of  ontogenetic  characters  1-7  among  lam- 
priform genera  may  be  instructive  when  considering  suggestions 
by  previous  authors  of  evolutionary  trends  within  the  order. 
The  indication  of  monophyly  by  Regan  (1907,  1924)  and  the 
adoption  of  this  hypothesis  by  Greenwood  et  al.  (1966)  and 
Oelschlager  (1976a)  is  supported  by  the  common  possession  of 
characters  1-4  among  all  known  lampriform  larvae.  Rigorous 
testing  of  this  hypothesis  utilizing  ontogenetic  data,  however, 
must  await  a  more  complete  knowledge  of  egg  and  larval  de- 
velopment among  Lampriformes  and  between  these  fishes  and 
other  groups.  Rosen  (1973)  suggested  that  relationships  among 
trachichthyoids,  berycids,  zeoids  and  lampriforms  seem  plau- 
sible. Ontogenic  characters  (1-4)  which  appear  to  unite  the  di- 
verse lampriform  genera  are  variously  present,  absent  or  un- 
known in  trachichthyoid,  berycid  and  zeoid  fishes  and  present 
no  clear  picture  of  inter-relationships.  Larvae  of  Diretmus  and 
Diretmoides  (Post  and  Quero,  1981)  lack  these  characters  and 
are  distinguished  by  pronounced  occipital  and  preopercular 
spines.  Polymixia  sp.  ( 1 0.0  mm  SL;  MCZ  58964)  lack  characters 
2  and  3  (eggs  of  Polymixia  are  unknown)  but  possess  well 
developed  ventral  fins.  These  fins  may  not  be  present  at  hatch- 
ing, however.  Juvenile  Cyttus  traversi  (James,  1976b)  possess 
elongate,  ornamented  and  pigmented  pelvic  and  anterior  dorsal 
elements,  although  the  sequence  of  development  of  these  struc- 
tures is  unknown.  The  rhomboidal  body  shape,  symmetrical 
caudal  and  jaw  structure  of  C.  traversi  resemble  deep-bodied 
lampriform  genera. 

Rosen  and  Patterson  (1969),  Rosen  (1973)  and  Oelschlager 
(1974,  1976a,  1978a,  b,  1979)  have  examined  osteological  and 
functional  aspects  of  adult  lampriform  morphology  and  com- 
mented on  relationships.  Recent  fishes  are  represented  by  a 
series  of  highly  modified  forms  of  which  i'elifer  is  believed  to 
be  the  least  specialized.  Veliferids  are  considered  to  be  more 
closely  related  to  Lampris  than  to  any  other  genus  on  the  basis 
of  similar  body  form,  caudal  morphology,  meristics  and  the 
possession  of  a  predorsal  element.  No  apomorphous  character 
serves  as  a  criterion  for  monophyly  in  the  Veliferidae  (Oel- 
schlager, 1976a).  Ontogenetic  characters  5-7,  however,  may  be 
useful  in  defining  relationships  between  the  two  series  of  families 
[Oelschlager's  (1976a)  Bathysomi  and  Taeniosomi]  within  the 
order. 

Among  the  elongate  genera,  Agrostichthys  is  considered  most 
closely  related  to  Regalecus  (Oelschlager,  1978a,  b).  Desmo- 
dema  and  Zu  represent  an  apomorphous  sister  group  of  Tra- 
chiptenis. considered  the  most  primati  ve  trachipterid  genus  (Ro- 
senblatt and  Butler,  1977).  Radiicephalus  appears  to  be  the  least 
specialized  among  all  elongate  lampriforms  although  it  shares 
several  specialized  features  (ink  sac,  caudal  filament)  in  common 
with  lophotids  and  Stylephorus  (Harrison  and  Palmer,  1968). 

Virginia  Institute  of  Marine  Science,  College  of  William 
AND  Mary,  Gloucester  Point,  Virginia  23062. 


Mirapinnatoidei:  Development  and  Relationships 

E.  Bertelsen  and  N.  B.  Marshall 


FISHES  of  the  Mirapinnatoidei  are  soft-rayed,  scaleless, 
oceanic  teleosts  with  elongated  body,  jugular  pelvic  fins  of 
4-10  rays,  a  single  dorsal  fin  opposed  to  the  anal  fin  with  origin 
behind  mid-standard  length,  pectorals  lateral,  caudal  fin  with 
10  -(-  9  principal  rays,  cleft  of  mouth  oblique  to  subvertical, 
premaxillae  excluding  maxillae  from  gape,  jaws  no  more  than 
slightly  protrusible,  branchiostegal  rays  3-5  on  epihyal,  4  on 
ceratohyal,  swimbladder,  functional  only  in  larvae,  with  two 
posterior  retia  mirabilia  that  supply  an  anterior  gas  gland.  An 
isolated  phylogenetic  lineage  of  uncertain  systematic  position 
but  apparently  most  closely  related  to  the  Megalomycteroidei 
and  the  Cetomimoidei. 

Development 

These  fishes  were  originally  placed  in  two  families  by  Ber- 
telsen and  Marshall  ( 1 956):  ( 1 )  Mirapinnidae,  with  a  single  genus 
and  species  Mirapinna  esau(Fig.  202)  based  on  a  single  subadult 
female  39.5  mm  SL  caught  at  the  surface  off  the  Azores  and  (2) 
Eutaeniophoridae,  with  two  genera  Eutaeniophorus  and  Para- 
taeniophorus  (Figs.  202  and  203)  both  known  only  in  larval  and 
metamorphosis  stages  less  than  55  mm  SL  that  are  epipelagic 
in  tropical  and  subtropical  parts  of  all  oceans. 

Examination  of  more  recent  material  indicates  that  these  fish- 
es are  better  regarded  as  members  of  a  single  family.  Mirapin- 
nidae, containing  the  above  mentioned  3  genera.  Adults  prob- 
ably are  mesopelagic.  The  genera  and  species  are  distinguished 
by  meristic  and  morphometric  characters  as  well  as  differently 
developed  dermal  structures  (Table  101).  Hair-like  outgrowths 
of  the  epidermis  are  found  all  over  the  head,  body  and  fins  of 
Mirapinna  esaii.  The  longest  hairs  measure  from  about  1 .0  to 
1.5  mm  in  length  and  bear  stalked  glandular  cells.  The  skin  of 
Eutaeniophorus  and  Parataeniophorus  is  densely  covered  with 
minute  papillae  less  than  about  0.05  mm  in  length  (Bertelsen 
and  Marshall,  in  preparation).  Skin  of  the  caudal  fin  of  Eutae- 
niophorus and  Parataeniophorus  is  prolonged  into  a  ribbon-like 
streamer  reaching  lengths  of  200-300%  SL.  Upper  and  lower 
lobes  of  the  caudal  fin  overlap  in  Mirapinna. 

Specimens.  — \nc\ud\ng  a  number  of  unpublished  records  the 
material  of  Mirapinnatoidei  known  to  us  consists  of:  One  Mir- 
apinna esau:  the  holotype,  a  39.5  mm  juvenile  female;  about 
100  Eutaeniophorus  festivus  8.0-53  mm;  two  Eutaeniophorus 
n.  sp.  (in  preparation)  12  and  16  mm;  32  Parataeniophorus 
gulosus  8-35  mm;  3  Parataeniophorus  hrevis  13.5,  29  and  46 
mm;  2  Parataeniophorus  n.  sp.  (in  preparation)  9  and  1 1  mm; 
about  40  unidentified  small  larvae  (most  probably  E.  festivus) 
5-12  mm.  Eggs  of  Mirapinnatoidei  are  unknown  and  no  larval 
Mirapinna  has  been  recorded;  [according  to  our  reexamination 
a  specimen  of  about  16  mm  referred  to  this  species  by  Four- 
manoir,  (1971b)  is  a  Parataeniophorus  sp.].  All  the  specimens 
have  small  immature  gonads.  A  light  brown  pigmentation  of 
the  skin  appears  at  a  larval  length  of  about  20  mm  and  some 
of  the  35-53  mm  largest  specimens  are  dark  brown  and  are 
considered  post-metamorphic  juveniles.  However  the  transfor- 


mation from  larval  to  juvenile  appearance  is  quite  gradual  with- 
out any  distinct  specialized  metamorphic  stage. 

The  youngest  Eutaeniophorus  larva  described  (6.5  mm  SL) 
has  remains  of  a  yolk  sac,  nearly  unpigmented  eyes,  no  rudiment 
of  pelvic  fins,  continuous  embryonic  fins  without  trace  of  fin 
rays  and,  except  for  a  ventral  series  of  melanophores,  the  body 
is  completely  unpigmented  (Bertelsen  and  Marshall,  1958).  Full 
numbers  of  rays  of  the  unpaired  fins  may  be  delected  at  8-9 
mm  SL.  Rudiments  of  pelvic  fins  are  present  at  6-7  mm  SL, 
the  number  of  rays  discemable  at  about  10  mm  SL.  Pectoral 
fin  rays  are  not  well  differentiated  until  lengths  of  about  20  mm 
SL.  The  caudal  streamer,  characteristic  of  Eutaeniophorus  and 
Parataeniophorus,  is  present  as  a  short  rudiment  in  the  5-6  mm 
youngest  larvae;  it  increases  with  increasing  SL.  It  is  broken  in 
most  specimens  of  more  than  about  10  mm.  The  greatest  lengths 
observed  are  86%  SL  in  a  Parataeniophorus  hrevis  of  22.6  mm, 
about  200%  SL  in  two  specimens  of  Eutaeniophorus  of  12-15.5 
mm  SL  (unpublished  data),  and  no  less  than  300%  in  an  E. 
festivus  of  35  mm  (Fig.  203). 

All  Eutaeniophorus  larvae  are  very  slender  with  body  depth 
less  than  10%  SL  except  for  the  largest  specimens.  Body  depth 
in  Parataeniophorus  species  is  less  than  1 5%  SL.  Predorsal  lengths 
(snout  to  first  dorsal  finray)  in  these  genera  is  69  to  77%  SL  (cf 
Table  101), 

All  larvae  have  a  fine  peppering  of  melanophores  on  head 
and  body,  slightly  increasing  in  density  with  increasing  SL,  with 
no  distinct  grouping  except  for  a  slight  increase  in  density  to- 
wards the  tail,  on  the  dorsal  part  of  the  peritoneum,  and  along 


Table  101.    Characteristics  of  Mirapinnatoidei. 


Eutaenio- 

Mirapinna 
esau 

Parataeniophorus 
hrevis              gulosus 

phorus 

festivus 

Texture  of  skin 

Hairs 

Minute  papillae 

Caudal  streamer 

Absent 

Present 

Ungth  in  %  SL: 

Head 

27 

10-17 

12 

10-13 

Predorsal 

61 

71-7: 

69 

10-11 

Longest  pectoral  finray 

14 

3-9 

7 

4-7 

Longest  pelvic  finray 

40 

15-25 

16-39' 

12-19^ 

No.  of  finrays: 

Pectoral 

13 

ca.  16-18 

19-20 

20-24 

Pelvic 

8 

8-9 

9-10 

4-5 

Dorsal 

16 

15-20 

28-33 

16-20 

Anal 

14 

14-17 

23-29 

15-18 

No.  of  vertebrae: 

Predorsal 

22 

21-27 

28 

31-36 

Under  dorsal  fin 

10 

8-11 

13 

8-11 

In  caudal  peduncle 

16 

10 

8 

7-8 

Total 

48 

42-46 

49 

47-55 

'  Shorter  in  specimens  less  than  15  mm. 
-  Shorter  in  specimens  less  than  25  mm. 


380 


BERTELSEN  AND  MARSHALL:  MIRAPINNATOIDEI 


381 


■■■      \    I     /  /  / 


«a^ 


10  mm  il !  M  ^  \  ,\  V 


""'^-. 


Fig.  202.     (A)  Mirapinna  esau.  holotype.  39.5  mm  SL;  (B)  Paralaeniophorus  brevis.  29  mm  SL;  and  (C)  Parataemophorus  brevis.  holotype, 
13.5  mm  SL.  A  and  C  from  Bertelsen  and  Marshall  (1956),  B  drawn  by  Kai  L.  Elsman. 


the  myosepta.  Density  of  pigment  is  greater  on  the  caudal 
streamer  and  the  caudal  fin  rays  at  the  base,  the  fully  developed 
streamer  has  a  median  longitudinal  band  of  pigment  and  a  nearly 
black  ventral  border.  The  two  species  of  Parataemophorus  differ 
in  pigmentation  from  Emaeniaphorus  feslivus  in  having  a  distal 
patch  of  pigment  on  each  pelvic  fin.  No  other  distinguishing 
characters  in  pigmentation  have  been  found. 


Relationship.s 

Reference  to  Bertelsen  and  Marshall  ( 1 956),  Myers  and  Frei- 
hofer  (1966)  and  Goodyear  (1970)  shows  that  both  the  mira- 
pinnatoid  and  megalomycteroid  fishes  have  the  following  com- 
mon features;  ( 1 )  they  are  small,  elongated  fishes  with  a  relatively 
small  head  and  mouth;  (2)  the  suspensoria  are  inclined  forwards 
and  there  is  a  single  supramaxilla  in  the  upper  jaw;  (3)  they  have 


382 

A 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


'„CLI  ■3(.W."1  Ij'A 


10  mm 


Fig.  203.     (A)  Eutaeniophorus  festivus.  paratype,  35  mm  SL,  with  complete  caudal  streamer,  106  mm  in  length;  (B)  Eutaeniophorus  festivus. 
holotype.  53  mm  SL;  and  (C)  Paralaeniophorus  gulosus.  paratype,  21  mm  SL.  All  from  Bertelsen  and  Marshall  (1956). 


soft  rays,  which  are  unbranched  [except  in  the  caudal  fin  of 
megalomycteroids?  Myers  and  Freihofer's  (1966)  drawing  of 
Megalomycter  leevani  shows  the  complete  caudal  rays  ending 
in  actinotrichia,  and  they  state  that  the  dorsal  and  anal  rays  are 
unsegmented];  (4)  the  dorsal  and  anal  fins  are  opposed  and 
inserted  on  the  posterior  half  of  the  body;  (5)  the  pectorals  are 
laterally  set  and  have  numerous  rays  (D.  15-33,  A.  14-29  in  the 
mirapinnatoids;  D.  15-26,  A.  14-20  in  the  megalomycteroids); 

(6)  the  pelvic  fins  are  inserted  below  or  before  the  base  of  the 
pectorals,  but  are  reduced  or  absent  in  the  megalomycteroids, 
whereas  the  pelvics  of  the  mirapinnatoids  are  well  developed; 

(7)  the  numbers  of  branchiostegal  rays  (on  the  epihyal  and  cer- 
atohyal)  are  3-5  +  4-5;  (8)  the  vertebrae  number  41-54  (45- 
48  in  the  megalomycteroids). 

The  main  differences  between  the  two  groups  concern  the  skin 
(papillate  or  "hairy"  in  the  mirapinnatoids,  scaled  in  the  me- 
galomycteroids), olfactory  organs  (very  large  in  the  latter,  small 
in  the  former)  [Goodyear's  (1970)  specimen  was  a  ripe  male  but 
Myers  and  Freihofer  (1966)  did  not  determine  the  sex  of  their 
specimens.  It  is  possible  that  the  females  have  yet  to  be  found 
and  are  microsmatic].  The  gape  markedly  oblique  in  the  mir- 
apinnatoids, somewhat  oblique  or  horizontal  in  the  megalo- 
mycteroids. 


The  mirapinnatoids  resemble  the  cetomimoids  in  having  soft 
rays,  a  scaleless  skin,  opposed  dorsal  and  anal  rays  on  the  pos- 
terior part  of  the  body  and  the  same  numbers  and  arrangement 
of  branchiostegal  rays  (mirapinnatoids  3-5  +  4,  cetomimoids 
3-4  +  4-5).  There  is  also  a  marked  resemblance  between  the 
swimbladder  of  Barbourisia.  which  regresses  after  a  presumed 
functional  stage  in  the  early  life  history,  and  the  swimbladder 
of  the  mirapmnatoids  (see  Bertelsen  and  Marshall,  1956).  In 
both  there  are  two  posterior  retia  mirabilia  that  run  forward  to 
an  anterior  gas  gland. 

One  main  difference  between  these  two  suborders  concerns 
the  head,  which  whether  relatively  large  or  small  in  the  ceto- 
mimoids, bears  long  jaws  with  a  more  or  less  horizontal  gape. 
This  contrasts  strongly  with  the  relatively  short,  upturned  jaws 
of  the  mirapinnatoids.  (Even  so,  it  may  well  be  that  the  fishes 
of  these  suborders  and  the  megalomycteroids  feed  largely  on 
copepods.)  Secondly,  in  the  two  cetomimoids  that  have  pelvic 
fins  {Rondcletia  and  Barbourisia)  these  are  abdominal  in  posi- 
tion whereas  those  of  the  mirapinnatoids  are  jugal. 

Beside  the  similarities  considered  above,  the  mirapinnatoids, 
megalomycteroids  and  cetomimoids  resemble  each  other  in  the 
disposition  of  the  red  muscle  component  of  their  axial  muscles. 
Down  the  entire  length  of  their  myotomes  red  muscle  fibres 


BERTELSEN  AND  MARSHALL:  MIRAPINNATOIDEI 


383 


cover  at  least  the  main  "V"  of  each  element,  and  such  an  ar- 
rangement seems  to  be  unusual  in  teleosts.  Similar  wide  red 
muscle  coverage  of  the  myotomes  is  found  also  in  the  stomia- 
toids  and  giganturoids  (Marshall,  unpublished)  and  apparently 
alsoinmaleceratioidangler-fishes(Marshall,  1971).Othergroups 
will  probably  prove  to  have  this  kind  of  red  muscle  arrangement 
but  the  most  usual  condition  in  teleosts  is  a  narrow  concentra- 
tion of  red  muscle  on  either  side  of  the  horizontal  septum  down 
the  entire  length  of  the  fish.  However,  in  alepisauroids  the  ver- 
tical extent  of  red  muscle  expands  towards  the  tail,  where  it  may 
cover  most  of  the  myotomes  (Marshall,  1971;  Johnson,  1982). 
The  above  treatment  of  adult  characters  indicates  that  the 
mirapinnatoids  are  most  closely  related  to  the  megalomycter- 
oids.  Next  to  the  latter  they  are  most  nearly  allied  to  the  ceto- 


mimoids.  As  will  be  seen  from  the  title  of  this  paper,  we  have 
followed  Greenwood  et  al.  (1966)  in  placing  all  three  suborders 
in  the  order  Cetomimiformes  away  from  the  Acanthopterygii. 
Whether  they  can  be  gathered  into  a  larger  ordinal  grouping,  as 
in  the  Lampridiformes  (Rosen  and  Patterson,  1969)  or  in  the 
Beryciformes  (Rosen,  1 973),  is  a  matter  for  further  comparative 
studies  (see  also  Zehren,  1979).  Nothing  is  known  of  larval 
megalomycteroids  and  cetomimoids.  Larval  forms  of  other 
groups  seem  to  have  no  affinities  to  larval  mirapinnatoids. 

(E.B.)  Zoological  Museum,  University  of  Copenhagen, 
Copenhagen  2100  0.  Denmark.  (N.B.M.)  6  Park  Lane, 
Saffron-Walden.  Essex,  England. 


Beryciformes:  Development  and  Relationships 
M.  J.  Keene  and  K.  a.  Tighe 


IN  the  classification  of  Greenwood  el  al.  (1966),  the  order 
Beryciformes  was  divided  into  3  suborders;  the  Stephan- 
oberycoidei  with  3  families,  the  Polymixoidei  with  I  family  and 
the  Berycoidei  with  8  families.  Rosen  and  Patterson  (1969)  re- 
moved the  Polymixiidae  from  the  Beryciformes,  assigned  it  to 
a  new  order,  the  Polymixiiformes  and  placed  this  order  in  the 


Paracanthopterygii.  Rosen  and  Patterson  (1969)  also  moved  the 
Cetomimidae,  Barbourisiidae  and  Rondeletiidae  to  the  Bery- 
ciformes in  the  suborder  Cetomimoidei.  Woods  and  Sonoda 
(1973)  considered  the  order  Berycomorphi  to  contain  the  fam- 
ilies Polymixiidae,  Diretmidae,  Monocentridae,  Anomalopidae, 
Trachichthyidae,  Holocentridae,  Berycidae,  Sorosichthyidae,  and 


Table  1 02.    Merlstic  Ranges,  OsTEOLooirAL  Characters,  Number  of  Genera, 
All  data  are  from  Woods  and  Sonoda  (1973),  Ebeling  and 


AND  Number  of  Species  for  Families  in  the  Order  Beryciformes. 
Weed  (1973)  or  Zehren  (1979)  unless  noted. 


Osteologjcal  characters 

Principle 

.  Number 

Number 

caudal 

Branchi- 

Verte- 

Orbito- 

Subocular 

Supra- 

of 

of 

Pelvic 

Dorsal 

Anal 

rays 

Pectoral 

ostegals 

brae 

sphenoid 

shelf 

maxillarv 

genera 

species 

Berycidae 

I,  7  or 
I,  10-12 

iv-vn, 

13-18 

IV, 

12-30 

16-17 

13-18 

7-9 

24 

Present 

Present 

2 

2 

ca.  10 

Anoplogasteridae 

1,6 

0,  17-19 

0,  8-9 

17 

14-16 

8 

28 

Present 

Present 

1 

1 

1 

Diretmidae 

1,6 

0,  17-19 

0,  18-24 

17 

16-20 

7-9 

27-31 

Present 

Present 

1 

2 

3 

Soroshicthyidae' 

1,5 

X,  8 

11.8 

16  or  17 

13 

7 

? 

9 

7 

7 

1 

1 

Trachichthyidae 

1,6 

IV-VIII, 
12-18 

II-IIl, 
8-12 

17 

14-20 

8 

26-29 

Present 

Present 

2 

5 

ca.  14 

Anomalopidae 

0,  6-7 

n-iv, 

14-19 

II,  10-13 

19 

16 

7  or  8 

25-30 

Present 

Present 

1  or  2 

3 

4 

Monocentridae 

I,  2-3 

IV-VII, 

11-12 

IV-V,10 

0.  10-12 

19 

14 

8 

27 

Present 

Present 

1 

1 

2 

Hispidoberycidae- 

1,  6 

III,  9 

17 

12 

8 

34 

Present 

Absent 

1 

1 

1 

Holocentridae 

I.  5-8 

X-XIII, 
11-16 

IV,  9-16 

17 

14-17 

7-8 

26-27 

Present 

Present 

2 

9 

ca.  70 

Gibberichthyidae 

0,5-6 

V-VIII, 
8-9 

IV-V, 
7-9 

17 

13-15 

8 

28-31 

Absent 

Absent 

1 

1 

2 

Stephanoberycidae 

0,5 

O-III. 
11-14 

O-III, 
10-13 

19 

11-18 

7-8 

30-33 

■Absent 

Absent 

1 

3 

3 

Melamphaidae 

I,  6-9 

I-III, 

9-18 

I,  7-10 

17 

14-17 

24-31 

Absent 

Absent 

Oor  1 

5 

ca.  30 

Rondeletiidae 

0,  5-6 

0,  13-16 

0,  13-15 

17 

9-10 

8 

24-27 

Absent 

Absent 

1 

1 

2 

Barbourisiidae 

0.  6 

0,21-22 

0,  16-18 

17 

14 

8 

ca.  42 

Absent? 

Absent? 

1 

1 

1 

Cetomimidae 

Absent 

14-ca.  30 

13-ca.  30 

11-19 

16-20 

8-10 

51-52 

Absent? 

Absent? 

1 

5 

10 

Paradirctmidae' 

1,5 

X-XI?, 
15? 

III,  13 

ca.  13 

16 

7 

9 

7 

7 

7 

1 

1 

'  Whitley,  l'>45. 
'Kotlyar,  1981, 
'Whitley.  1946, 


384 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  103.    References  Giving  Descriptions  and/or  Figures  of 
Early  Life  History  Stages  of  the  Order  Beryciformes. 


Pre-juveniles/ 

Family 

Eggs                     Larvae 

juveniles 

Holocentridae 

-     McKenney,  1959 

McKenney,  1959 

Jones  and 

Jones  and 

Kumaran,  1962 

Kumaran,  1962 

Aboussouan,  1966b 

Greenfield,  1965 
Randall  et  al., 
1982 

Melamphaidae 

-     Ebeling,  1962 

Ebeling  and  Weed, 

1963,  1973 
Pcrtseva-Ostrou- 
mova  and  Rass, 
1973 
Moser,  unpub- 
lished 

Diretmidae 

-     Post,  1976 

Post  and  Quero, 
1981 

Anoplogasteridae 

—             — 

Parr,  1933 

Trachichthyidae 

—     Ahlstrom  (notes) 

Parr,  1933 

Crossland,  1981 

Johnson,  1970 

Gibberichthyidae 

Robins  and 

de  Sylva,  1965 
Thorp,  1969 
de  Sylva  and 

Eschmeyer. 

1977 

Berycidae 

Anomalopidae 

Monocentridae 

Sorosichthyidae 

Paradiretmidae 

H  ispidoberycidae 

Stephanoberycidae 

Barbourisiidae 

Rondeletiidae 

Cetomimidae 


Anoplogasteridae.  while  Ebeling  and  Weed  (1973)  considered 
the  order  Xenoberyces  to  contain  the  families  Melamphaidae, 
Gibbeinchthyidae,  and  Stephanoberycidae.  Both  pairs  of  au- 
thors gave  diagnostic  characters,  and  compared  and  contrasted 
their  orders.  Zehren  (1979),  after  studying  the  comparative  os- 
teology and  phylogeny  of  the  beryciform  families  of  Greenwood 
et  al.  ( 1 966),  also  concluded  that  the  Polymixiidae  did  not  belong 
in  the  Beryciformes.  Nelson  (1976)  included  the  families  So- 
rosichthyidae and  Paradiretmidae  in  the  suborder  Berycoidei 
but  did  not  treat  them  further.  Kotlyar  (1981 )  described  a  new 
species  of  beryciform  which  he  felt  deserved  status  as  a  new 
family,  the  Hispidoberycidae.  He  tentatively  aligned  his  new 
family  within  the  Berycoidei.  The  Beryciformes  are  presently 
defined  on  the  basis  of  several  primitive  characters  such  as  the 
presence  of  an  orbitosphenoid  and  subocular  shelf  (in  most 
forms)  and  a  high  number  of  pelvic  and  caudal  rays  as  well  as 


several  derived  characters  such  as  the  presence  of  dorsal,  anal 
and  pelvic  spmes.  and  the  presence  of  spinous  procurrent  caudal 
fin  rays.  However,  none  of  the  characters  is  unique  to  the  order 
and  the  monophyly  of  the  order  is  still  in  question.  Meristics, 
osteological  characters,  and  the  number  of  genera  and  species 
in  each  beryciform  family  are  shown  in  Table  102. 

Although  the  systematics  of  the  Acanthopterygii  is  in  a  state 
of  flux,  the  order  Beryciformes  presently  contains  3  suborders; 
the  Stephanoberycoidei  with  3  families,  the  Berycoidei  with  10 
families,  and  the  Cetomimoidei  with  3  families. 

The  Beryciformes  are  considered  by  Greenwood  et  al.  (1966) 
to  be  the  basal  stock  from  which  some  of  the  more  advanced 
acanthopterygians  have  evolved.  Beryciformes  are  marine  and 
occur  in  all  oceans.  Some  species  are  semibenthic  inhabiting 
coral  reefs,  rocky  shores,  and  shelf  or  slope  waters  (Woods  and 
Sonoda,  1973)  while  others  are  epipelagic,  mesopelagic.  bathy- 
pelagic,  or  bathybenthic  (Ebeling  and  Weed,  1973). 

Development 

There  is  no  published  information  on  early  life  history  stages 
for  the  Monocentridae,  Anomalopidae,  Berycidae,  Sorosichthy- 
idae, Paradiretmidae,  Hispidoberycidae,  Stephanoberycidae, 
Barbourisiidae,  Rondeletiidae,  and  Cetomimidae  (Table  103). 
Although  information  is  lacking  on  the  eggs  of  the  Beryciformes, 
there  is  some  on  other  early  life  history  stages  of  the  Holocen- 
tridae, Melamphaidae,  Anoplogasteridae,  Diretmidae,  Trach- 
ichthyidae and  Gibberichthyidae. 

The  Holocentridae  contains  two  subfamilies,  the  Holocen- 
trinae  and  the  Myripristinae.  Prejuveniles  and  early  life  history 
stage  series  are  known  for  at  least  one  species  in  each  subfamily. 
McKenney  (1959)  gave  a  detailed  description  of  the  early  life 
history  of  Holoccntrus  ve.xi/lanus  based  on  specimens  less  than 
2.0  mm  to  adults,  while  both  Aboussouan  (1966b)  and  Jones 
and  Kumaran  (1962)  figure  and  discuss  larvae  of  Holocentrus 
sp.  less  than  5.0  mm  SL.  Jones  and  Kumaran  (1962)  also  figure 
and  describe  larval  stages  ranging  from  2.7  to  6.7  mm,  for  My- 
ripristis  mirdjan  [specific  identification  questioned  by  Greenfield 
(1965)].  McKenney  (1959)  figured  the  prejuvenile  or  rhynch- 
ichthys  stage  of  Holocentrus  vexillarius  while  Jones  and  Ku- 
maran ( 1 962),  Greenfield  (1965),  and  Randall  et  al.  (1982)  figure 
the  rhynchichthys  stage  for  several  myripristine  species.  The 
following  characterization  of  holocentrid  development  is  based 
on  the  data  of  McKenney  (1959)  and  Jones  and  Kumaran  (1962). 

Holocentrid  larvae  are  characterized  by  a  relatively  large  head 
with  well-developed  preopercular,  rostral,  and  median  cranial 
spines.  Pigmentation  is  extensive  on  the  peritoneum  and  there 
is  a  ventral  line  of  melanophores  in  the  postanal  region.  The 
long  preopercular  spines  develop  first  and  are  well  developed 
at  1.8  mm  TL  (Fig.  204A).  At  2.2  mm  the  posteriorly  directed 
cranial  spine  is  rapidly  forming  and  by  2.8  mm  the  rostral  spine 
is  apparent.  The  5.0  mm  H,  vexillarius  and  4.7  mm  Mynphstis 
sp.  (Fig.  204B,  C)  both  exhibit  strong  rostral,  median  cranial, 
preopercular  and  opercular  spination  that  develops  into  the  head 
armor  found  in  the  rhynchichthys  stage  (Fig.  204D,  E).  There 


Fig.  204.  (A)  Preflexion  larva  oi  Holocentrus  vexillarius.  1.8  mm  NL  (source:  McKenney,  1959);  (B)  Flexion  larva  o[  Holocentrus  vexillarius. 
5.0  mm  NL  (source;  McKenney,  1959);  (C)  Preflexion  larva  of  A/ir/pmrKsp..  4.7  mm  NL  (source;  Jones  and  Kumaran,  1962);  (D)  Rhynchichthys 
prejuvenile  of  Holocentrus  vexillarius.  24.9  mm  SL  (source;  McKenney,  1959);  and  (E)  Rhynchichthys  prejuvenile  of  Myriprislis  sp.,  16.3  mm 
SL  (source:  Jones  and  Kumaran,  1962). 


KEENE  AND  TIGHE:  BERYCIFORMES 


385 


386 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  205.     Urval  senes  oi  Melamphaes  lugubris  (A)  5.3  mm  NL,  (B)  6.2  mm  SL.  (C)  10.4  mm  SL  and  (D)  15.2  mm  SL  (source:  Southwest 
Fisheries  Center,  CalCOFI,  original,  illustrated  by  B.  Y.  Sumida). 


KEENE  AND  TIGHE:  BERYCIFORMES 


387 


Fig.  206.  (A)  Postflexion  larva  of  Scopelogadus  bispinosus.  8.0  mm  SL;  (B)  Postflexion  larva  of  Poromitra  sp..  13.5  mm  SL;  (C)  Postflexion 
larva  of  Poromitra  megalops.  10.0  mm  SL;  (D)  Postflexion  larva  of  Scopelobery.x  sp.,  6.5  mm  SL;  (E)  Postflexion  larva  of  Melamphaes  lugubris, 
8.3  mm  SL;  (F)  Postflexion  larva  of  Melamphaes  lyphlops.  9.4  mm  SL  (source:  Ebeling,  1962). 


are  minor  differences  in  the  spine  patterns  of  the  two  species 
illustrated,  and  the  Holocenlnis  spination  is  somewhat  more 
developed.  All  of  these  spines  are  lost  as  the  fish  develops  into 
a  juvenile  in  the  Myripristinae,  while  the  Holocentrinae  retain 
only  large  preopercular  spines. 

A  considerable  amount  is  known  about  at  least  some  of  the 
early  life  history  stages  of  all  melamphaid  genera  except  Sio. 
Notes  from  Moser  and  Ahlstrom,  and  an  examination  of  me- 
lamphaid larval  specimens  from  the  Southwest  Fisheries  Cen- 
ter (SWFC),  allow  the  following  conclusions  to  be  made  about 
early  melamphaid  larvae  2-10  mm  for  Poromitra.  Melam- 
phaes. Scopelogadiis,  and  Scopeloberyx.  Specimens  in  this  range 
tend  to  have  a  relatively  more  elongate  and  slender  body  shape 
than  later  larval  stages.  The  pelvics  develop  rapidly  followed 
closely  by  the  pectoral  fins.  The  pelvic  fin  origin  is  more  anterior 
than  in  later  stages,  and  the  pelvic  rays  are  quite  long,  fragile 
and  darkly  pigmented.  This  condition  persists  longer  and  is 
more  striking  in  some  species  such  as  M.  lyphlops  than  in  others. 
In  early  larvae  of  Melamphaes.  Scopelogadiis.  and  Scopelobe- 
ryx. two  pigment  spots  occur  near  the  posterior  end  of  the  dorsal 
and  anal  fin  anlagen  (Fig.  205A).  These  pigment  spots  spread 
both  anteriorly  and  postenorly  during  growth  to  form  longitu- 
dinal rows  of  pigment  along  the  dorsal  and  ventral  surfaces  of 
the  body  (Fig.  205B).  In  some  species,  these  areas  of  initial 
pigmentation  spread  laterally  to  form  a  band  of  pigment  between 
the  dorsal  and  anal  fin  bases  in  later  larval  stages.  Additional 
pigmentation  occurs  on  the  cranium  and  peritoneum  in  all  four 
genera,  and  in  the  form  of  a  spot  at  the  posterior  end  of  the 
caudal  peduncle  in  at  least  Melamphaes  and  some  Poromitra. 
In  these  early  stages,  the  second  or  third  dorsal  fin  ray  tends  to 
be  much  longer  than  the  others,  extending  to  the  region  of  the 
caudal  peduncle.  This  elongate  ray  is  known  to  occur  in  Me- 
lamphaes. Scopeloberyx.  and  Scopelogadus.  Usually  damaged. 


this  elongate  ray  is  not  evident  after  5-10  mm  but,  even  in 
adults,  the  second  or  third  dorsal  ray  (spine)  is  the  largest.  By 
5-10  to  20  mm  SL  melamphaid  larvae  exhibit  body  shapes 
and  other  characters  such  as  meristics  and  preopercular  spi- 
nation that  allow  them  to  be  separated  into  genera  (Ebeling, 
1962;  Fig.  206A-F).  Development  is  gradual  and  direct;  there 
are  no  known  prejuvenile  stages.  Additional  larvae  were  illus- 
trated and  are  published  here  without  further  comment  (Fig. 
207). 

Early  life  history  stages  are  known  for  all  three  species  con- 
tained in  the  two  genera  of  the  Diretmidae.  Post  (1976)  discusses 
the  systematics  and  early  life  history  of  two  of  these  species, 
and  Post  and  Quero  (1981)  in  their  familial  revision,  describe 
a  new  genus  and  species,  give  the  early  life  history  of  all  three 
species,  and  provide  a  key  for  the  identification  of  juveniles. 
The  larvae  of  all  three  species  are  relatively  elongate  at  4-5  mm 
sizes  but  rapidly  develop  a  relatively  deeper  body.  All  three 
species  also  possess  a  short  stout  spine  over  each  eye,  a  longer 
cranial  spine  directed  posteriodorsad  on  each  side  of  the  head, 
and  a  long  preopercular  spine  directed  posterioventrad  (Fig. 
208A).  The  head  spine  configuration  is  quite  similar  to  that  of 
.4.  cormita.  described  below,  and  is  gradually  lost  during  growth. 

The  monotypic  Anoplogasteridae  contains  the  highly  spe- 
cialized mesopelagic  predator  .-inoplogaster  cormtta.  Specimens 
over  about  100  mm  SL  are  jet  black  with  large  fangs  while 
specimens  less  than  80  mm  SL  are  metallic  grey  with  black 
pigmentation  developing  along  the  ventral  midline,  do  not  have 
such  large  teeth,  and  exhibit  a  pattern  of  head  spination  not 
found  in  larger  individuals  (Woods  and  Sonoda,  1973).  USNM 
collections  contain  many  individuals  from  4.5  mm  TL  larvae 
to  adults,  upon  which  the  following  characterization  of  the  early 
life  history  stages  is  based. 

A  4.5  mm  prefiexion  larva  has  the  caudal  fin  elements  de- 


KEENE  AND  TIGHE:  BERYCIFORMES 


389 


Fig.  207.  (A)  Preflexion  larva  of  Scopeloberyxsp..  4.4  mm  XL;  (B)  Postflexion  larva  of  Scopeloberyx opisthopterus.  9.1  mm  SL;  (C)  Postflexion 
larva  of  Scopeloberyx  robuslus.  1 3.0  mm  SL;  (D)  Preflexion  larva  of  Poromitra  crassiceps  complex,  7.9  mm  SL;  (E)  Postflexion  larva  of  Melamphaes 
lepnis.  19.5  mm  SL;  all  drawn  by  B.  Washington. 


veioping.  The  dorsal,  anal  and  pectoral  fins  are  already  devel- 
oped, while  pelvic  fin  buds  are  present.  The  pattern  of  head 
spination  described  below  is  already  well  fisrmed.  A  6.0  mm 
postflexion  specimen  (Fig.  208B)  has  all  fins  completely  devel- 
oped except  for  the  pelvics  and  procurrent  caudal  elements. 
There  is  pigmentation  on  the  head,  lateral  surface  of  the  body 
and  caudal  peduncle,  while  the  abdominal  area  is  pale  with 
scattered  melanophores.  A  small  pigmented  area  occurs  on  the 
pectoral  bases.  A  serrate  frontal  ridge  bordenng  the  anterior  of 
each  eye  terminates  in  a  short  stout  supraocular  spine.  Ridges 
continuing  posteriodorsad  on  the  cranium  terminate  in  long 
serrate  spines  probably  arising  from  the  parietals.  The  pre- 
opercles  end  in  strong  serrate  spines  directed  posterioventrad. 
By  9.0  mm  SL,  the  pelvics  have  become  well-developed  and 
the  head  spination  is  still  strong.  A  small  dense  patch  of  me- 
lanophores occurs  on  the  ventral  surface  of  the  body  justantenor 
to  the  origin  of  the  pelvic  fins.  With  increasing  growth  (28  mm 
SL),  this  pigmentation  darkens  and  expands,  extending  forward 
in  a  continuous  band  to  the  tip  of  the  isthmus.  Additional  pig- 
mentation occurs  at  the  ongin  of  the  pelvic  fins,  around  the 
vent,  just  posterior  to  the  anal  fin  on  the  caudal  peduncle,  and 
in  a  transverse  bar  on  the  abdomen  midway  between  the  pelvic 
origin  and  the  vent.  The  increase  in  dark  pigmentation  and  the 
reduction  in  cranial  and  preopercular  spines  in  larger  juveniles 
is  described  by  Woods  and  Sonoda  (1973). 

Crossland  (1981)  illustrated  a  trachichthyid  larva,  probably 
of  OpliYus  I'longatus.  taken  off  northeastern  New  Zealand  (Fig. 
208C).  Larger  larvae  of  the  same  species  had  the  skin  on  the 
dorsal  surface  of  the  head  and  body  covered  with  tiny  spines. 
Ahlstrom  (notes)  sketched  early  Trachichthys  mento  larvae  that 
are  fairly  deep-bodied  at  3.5  to  4.5  mm,  with  the  pectoral  fin 
showing  precocious  development.  A  dark  spiny  pigmented  band 
extending  from  the  region  of  the  anal  to  the  dorsal  occurs  in  3.5 
mm  TL  specimens.  This  spination  covers  areas  on  both  sides 
of  the  dorsal  fin,  parts  of  the  head,  thoracic  region  and  jaws.  In 
a  preflexion  6.4  mm  specimen,  the  fin  rays  are  mostly  developed, 
and  the  body  is  stockier,  approaching  the  shape  of  the  adult  and 
is  covered  with  minute  spines.  The  holotype  of  Korogaster  nanus 
Parr  1933,  synonomized  by  Woods  and  Sonoda  (1973)  in  Ho- 
plostethus.  is  19  mm  long  (Fig.  208D),  possesses  unbranched 


rays  in  the  pectoral,  pelvic  and  caudal  fins,  and  has  dermal 
papillae  and  small  spines  all  over  its  body.  This  specimen  and 
the  second  specimen  of  Korsogaster  Tepone:<i  by  Johnson  (1970) 
(Fig.  208E)  are  juveniles  of  the  family  Trachichthyidae. 

The  most  striking  early  life  history  of  any  beryciform  is  ex- 
hibited by  the  prejuvenile  kasidoron  stage  of  gibberichthyids 
(Figs.  209,  210).  This  stage  is  characterized  by  a  long  trailing 
pelvic  appendage  which  is  part  of  a  modified  third  pelvic  ray 
and  is  present  in  specimens  from  at  least  7.5  to  21  mm  TL.  It 
is  lost  by  30  mm  SL  (de  Sylva  and  Eschmeyer,  1977).  Dunng 
early  growth,  this  trailing  appendage  becomes  more  ornate  and 
resembles  the  trailing  tentacles  of  siphonophores  or  Sargassum 
weed  at  about  15  mm  SL.  Up  until  about  at  least  20  mm,  the 
prejuveniles  inhabit  epipelagic  waters  but  by  30  mm  individuals 
have  lost  the  pelvic  appendage  and  taken  up  a  mesopelagic  to 
upper  bathypelagic  existence.  The  anterior  dorsal  and  anal  fin 
elements  are  soft  rays  during  the  kasidoron  stage,  but  develop 
into  strong  fin  spines  in  the  adult.  There  is  also  a  marked  de- 
velopment of  bony  head  ridges  in  the  adults,  that  is  not  found 
in  the  stages  20  mm  and  smaller  (de  Sylva  and  Eschmeyer, 
1977). 

Relationships 

Rosen  and  Patterson  (1969)  and  Rosen  (1973)  emphasized 
the  futility  of  the  present  classification  of  the  Beryciformes  and 
the  rest  of  the  Acanthopterygii,  because  it  relies  on  grouping  of 
primitive  characters  to  express  relationships.  Realizing  this. 
Zehren  (1979)  did  a  phylogenetic  analysis  of  the  Beryciformes 
to  attempt  to  determine  whether  or  not  the  order  is  monophy- 
letic(Fig.  211). 

Besides  supporting  Rosen  and  Patterson's  removal  of  the 
Polymixiidae  from  the  Beryciformes,  Zehren's  analysis  super- 
ficially suggests  that  the  remaining  ten  families  form  a  mono- 
phyletic  group.  However,  he  cautions  that  since  none  of  the 
derived  character  states  that  he  uses  is  unique  to  the  ten  families, 
their  monophyly  is  uncertain. 

Zehren's  results  and  discussion  suggest  that  the  Holocentridae 
do  not  appear  to  be  closely  related  to  the  other  nine  families 
studied.  Woods  and  Sonoda  (1973)  felt  that  the  Holocentridae 
were  very  different  from  the  other  Beryciformes  and  Rosen  (1973) 


390 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  208.  (A)  Postflexion  larva  of  Direlmus  argenteus.  ca.  6  mm  SL  (source:  Post,  1976);  (B)  Postflexion  larva  oi  Anoplogaster  cornuta.  6.0 
mm  SL  (USNM  244902)  drawn  by  B.  Washington;  (C)  Flexion  larva  of  Opiivusus  elongatus  ?  5.3  mm  NL  (source:  Crossland,  1981);  (D) 
Unidentified  trachichthyid  juvenile,  19.0  mm  SL  (source:  Parr,  1933);  (E)  Unidentified  trachichthyid  juvenile,  21.5  mm  SL  (source:  Johnson, 
1970). 


KEENE  AND  TIGHE:  BERYCIFORMES 


391 


Fig.  209.     Kasidoron  prejuvenile  of  Gibberichthys  pumilus,  1  5.3  mm  SL  (source:  de  Sylva  and  Eschmeyer,  1977). 


considered  the  holocentrids  to  be  a  distinct  major  subgroup 
within  the  order.  Rosen  (pers.  comm.  to  Zehren)  believes  that 
the  Holocentridae  should  be  placed  within  the  Perciformes. 

Another  result  of  Zehren's  study  is  that  the  Berycidae  appear 
to  be  the  primitive  sister  group  to  the  other  eight  families  and 
should  be  placed  in  their  own  suborder,  the  Berycoidei.  The 
Trachichthyidae,  Diretmidae.  Anoplogasteridae,  Anomalopi- 
dae  and  Monocentridae  are  closely  related  and  should  be  placed 
in  the  suborder  Trachichthyoidei,  as  suggested  by  Parr  (1933). 
The  Gibberichthyidae,  Stephanoberycidae  and  Melamphaidae 
also  appear  closely  related  and  form  the  suborder  Slephanob- 
erycoidei. 

Despite  the  efforts  of  Rosen  and  Zehren,  there  are  still  prob- 


lems with  beryciform  classification.  Only  adult  characters  have 
presently  been  used,  but  early  life  history  data  is  pertinent  in 
two  instances.  In  the  cladogram,  a  common  ancestry  is  suggested 
for  the  Diretmidae,  Anoplogasteridae,  Trachichthyidae.  An- 
omalopidae  and  Monocentridae  with  no  character  state  to  sep- 
arate them.  The  larval  head  spine  pattern  in  the  Diretmidae  and 
Anoplogasteridae  is  similar  and  distinctive,  and  may  help  to 
resolve  the  cladogram.  Gibberichthys  with  its  kasidoron  stage 
may  appear  to  be  vastly  different  from  the  Melamphaidae,  but 
the  occurrence  of  very  long  branched  pelvics  in  larval  Poromitra 
suggest  a  possible  relationship  (de  Sylva  and  Eschmeyer,  1977). 
In  summary,  further  phylogenetic  studies  of  the  order  Bery- 
ciformes  are  needed  in  order  to  determine  if  the  order  is  mono- 


392 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  210.     Kasidoron  larva  of  Gibberichthys  pumilus.  6.2  mm  NL,  DANA  Sta.  3543  (source:  de  Sylva.  pers.  comm.). 


Adioryx    diodema 
Holocentrus     rufus 
Plectrypops    retrospinis 
Ostichthys    trachypoma 
Myripristis     sp. 
Centroberyx    affinis 
4 — Beryx  splendens 

Scope  log  ad  us   mi  zolepis 
Scope logodus    unispinis 
Melamphaes    macrocepholus 
Melamphoes    eulepis 
Scopeloberyx    sp. 
Poromitra    pumilus 
Stephanoberyx    monae 
Gibberichthys    pumilus 
Diretmus    orgenteus 
Anopiogaster    cornuto 
Hoploslethus   mediterroneus 
Gephy  robery  X    da  r  wi  ni 
Trachichthys    australis 
Paratrachichthys    sp. 
Photoblephoron    palpebrotus 
'Anomalops    katopteron 
Monocentrus    japonicus 
Cleidopus    gtorioe- maris 

Fig.  211.     Cladogram  showing  the  relationships  of  the  beryciform 
families  studies  by  Zehren  (1979). 


phyletic,  to  determine  the  relationships  between  the  various 
suborders,  and  to  determine  the  relationships  of  the  Beryci- 
formes  to  other  orders  of  fishes.  Inclusion  of  early  life  history 
characters  in  these  studies  would  be  useful.  However,  the  lack 
of  early  Ufe  history  data  for  ten  of  the  beryciform  families  may 
prove  a  stumbling  block  in  these  efforts. 

Division  of  Fishes,  National  Museum  of  Natural  History, 
Smithsonian  Institution,  Washington,  District  of 
Columbia  20560. 


Zeiformes:  Development  and  Relationships 

K.  A.  TiGHE  AND  M.  J.  KJEENE 


THE  order  Zeiformes  is  diagnosed  by  a  series  of  derived 
characters  that  are  not  unique  to  the  order  (Heemstra. 
1980):  presence  of  dorsal  spines  in  most  forms;  presence  of  anal 
and  pelvic  spines  in  most  forms;  reduced  number  of  pelvic  and 
caudal  rays;  absence  of  orbitosphenoid;  absence  of  subocular 
shelf;  gills  3'/2  (no  slit  behind  last  hemibranch);  mouth  more  or 
less  protrusible;  no  supramaxilla.  Other  characteristics  of  the 
order  as  presented  by  Heemstra  ( 1 980)  are  primitive  characters 
that  shed  little  information  on  the  relationships  of  the  order. 
The  literature  on  Zeiformes  is  scattered  and  inadequate.  Only 
the  family  Zeidae  has  been  examined  on  a  world-wide  basis 
(Bray,  1983).  Information  on  most  species  is  descriptive,  with 
little  known  about  ranges,  life  history  stages,  abundance,  ecol- 
ogy, and  relationships.  Zeiformes  are  marine  and  various  species 
occur  in  the  tropical  and  temperate  parts  of  all  oceans  in  coastal, 
benthic,  epipelagic,  mesopelagic,  bathypelagic,  and  bathyben- 
thic  waters  (Wheeler  et  al.,  1973).  Families  are  distinguished  by 
presence  of  vertically  elongate  or  small  or  no  scales,  relative 
body  depth,  relative  mouth  size,  degree  of  development  of  anal 
and  pelvic  fin  spine(s),  number  of  lateral  lines,  and  morphology 
of  the  eye-jaw  region.  Generic  and  specific  designations  are  based 
mainly  on  morphometnc,  meristic,  specialized  scale,  and  color 
characters  (Heemstra,  1980). 

The  order  Zeiformes  is  presently  placed  in  superorder  Acan- 
thopterygii,  near  the  Beryciformes  and  other  groups  that  have 
not  attained  the  perciform  level  of  structural  organization. 
Greenwood  et  al.  (1966)  included  the  Parazenidae,  Grammi- 
colepididae,  Zeniontidae,  Oreosomatidae,  Zeidae,  Caproidae, 
and  Macrurocyttidae  in  the  Zeiformes.  Heemstra  (1980)  revised 
the  Zeidae  of  South  Africa  and  gives  a  key  to  all  the  zeiform 
families  above  except  the  Caproidae  which  he,  like  earlier  work- 
ers (Rosen,  1973),  feels  is  only  superficially  similar  to  zeiforms 
and  therefore  should  not  be  included  in  the  order.  He  also  pro- 
vides diagnoses  for  the  order  and  four  of  the  remaining  families. 
Parazen  pacificus,  not  reported  from  South  Africa,  is  described 
by  Mead  (1957).  Keys  to  South  African  zeids  and  grammico- 
lepidids  are  given  by  Heemstra  ( 1 980),  along  with  a  key  to  adult 
oreosomatids  of  the  southern  Atlantic  and  Indian  Oceans  sup- 
plied to  him  by  Karrer  and  Eschmeyer.  Meristic  ranges,  number 
of  species,  and  number  of  genera  for  the  six  families  presently 
in  the  Zeiformes  are  given  in  Table  104. 

Development 

Early  life  history  information  on  most  zeiform  species  is  non- 
existent (Table  105).  There  is  some  information  on  prejuvenile 
stages  (specialized  ontogenetic  stages  between  larvae  and  ju- 
veniles) for  Oreosomatidae  and  Grammicolepididae,  but  none 
on  earlier  stages.  Early  life  history  data  for  Zeus  faber  from  egg 
through  juvenile  is  quite  extensive,  but  such  information  is 
incomplete  or  nonexistent  for  other  zeid  species.  For  the  Cap- 
roidae. larvae  of  Aniigonia  capros  and  .-1.  ruhescens  are  known, 
as  are  all  the  early  stages  of  Capros  aper.  Nothing  is  known  for 
the  Parazenidae  and  Zeniontidae. 

Eggs  are  known  for  two  species  of  zeids.  They  are  spherical. 


have  a  single  oil  droplet,  nonsegmented  yolk,  and  a  smooth 
chorion.  Eggs  of  Zeus  faber  range  from  1 .8-2. 1  mm  in  diameter 
with  an  oil  droplet  diameter  of  .32-.40  mm  (Sanzo,  1956;  De- 
khnik,  1973;  Robertson,  1975a).  Those  of  Zenopsis  nebulosus 
are  2.0-2.25  mm  with  a  droplet  of  .275-.375  mm  (Robertson, 
1975a).  Eggs  of  Capros  aper  are  about  1.0  mm  in  diameter, 
spherical,  and  have  a  smooth  chorion,  unsegmented  yolk  and 
a  single  oil  droplet  (Arbault  and  Boutin,  1968a;  Sanzo,  1956). 
Eggs  of  all  other  species  of  zeiform  fishes  are  unknown. 

Newly  hatched  larvae  of  Zeus  faber  were  described  by  Sanzo 
(1931b).  Pigmentation  is  extensive  over  body,  head  and  yolk 
sac  with  the  pigmentation  extending  to  the  margin  of  the  dorsal 
finfold  and  also  on  the  base  of  the  anal  finfold  for  most  of  its 
length  (Fig.  212A).  Only  the  tip  of  the  caudal  region  is  unpig- 
mented.  The  pectoral  and  pelvic  fin  buds  are  present  upon  hatch- 
ing. Preflexion  larvae  retain  the  extensive  body  pigmentation, 
rapidly  become  deep-bodied,  and  show  a  precocious  develop- 
ment of  the  pelvic  fins  (Fig.  212B).  Postflexion  larvae  have 
almost  all  fin  elements  developed  (Fig.  212C)  and  are  rapidly 
assuming  the  characters  of  the  adult. 

Larval  stages  are  known  for  both  genera  in  the  family  Cap- 
roidae. Newly  hatched  larvae  of  Capros  aper  (Fig.  21 2D)  have 
large  stellate  melanophores  on  the  dorsal,  lateral  and  ventral 
surface  of  the  body  with  a  few  melanophores  on  the  head  and 
associated  with  the  oil  globule.  Preflexion  larvae  (Fig.  212E) 
become  very  deep-bodied  with  an  increase  in  head  size.  Pig- 
mentation densely  covers  the  entire  body  except  for  the  caudal 
region.  A  medial  serrated  ridge  occurs  on  the  cranium  and  other 
paired  serrate  ridges  develop  along  the  lower  jaw  and  in  the 
supraocular  region.  Numerous  preopercular  spines  also  develop 
during  this  stage.  Minute  spines  associated  with  the  developing 
scales  cover  the  entire  body  (Fage,  1918).  Transformation  to 
the  juvenile  is  gradual  and  completed  by  a  size  of  15-20  mm 
SL. 

Larvae  of  Antigonia  were  described  by  Uchida  (1936)  and 
Nakahara  (1962).  The  larvae  are  relatively  deep-bodied  with 
pigmentation  on  the  peritoneum  and  head.  The  median  serrate 
cranial  spine,  serrate  preopercular  spines,  and  serrate  ridges  on 
the  frontal,  mandibular  and  preopercular  regions  are  character- 
istic of  both  A.  rubescens  and  A.  capros  (Fig.  213A,  B),  but  are 
totally  lost  before  reaching  juvenile  sizes  of  25  mm.  There  are 
several  differences  between  the  larvae  of  the  two  species  but  the 
most  obvious  is  the  presence  of  a  vertically  directed  spine  in 
the  occipital  region  of  A.  rubescens. 

At  least  some  grammicolepidids  exhibit  striking  proportional 
changes  during  growth.  Smaller  Grammicolepis  brachtusculus 
are  very  deep-bodied  relative  to  larger  ones  based  on  an  ex- 
amination of  specimens  70  to  400  mm  SL  (Quero,  1979).  Young 
Xenolepidichihys  dalgleishi  also  have  a  relatively  deeper  body 
than  larger  specimens  (Myers,  1937)  and  possess  long  filamen- 
tous extensions  on  some  of  the  dorsal  spines  and  on  the  first 
anal  spine  (Smith,  1949;  Fig.  279).  These  shorten  greatly  with 
growth  as  shown  by  Myers'  (1937)  71  mm  SL  specimen. 

Oreosomatid  adults  have  mainly  overlapping  cycloid  or  cte- 


393 


394 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  104.    Meristic  Ranges,  Number  of  Genera,  and  Number  of  Species  for  the  Families  Placed  in  the  Zeiformes  by  Greenwood  et 

AL.  (1966).  All  data  are  from  Heemstra  (1980)  unless  noted. 


Branched 

caudal 

Branchi- 

Number  of 

Number  of 

Pelvic 

Dorsal 

Anal 

rays 

ostegals 

Vertebrae 

Pecloral 

genera 

species 

Grammicolepididae 

I,  6 

V-VII,  27-34 

II,  27-35 

13 

7 

37-46 

13-16« 

3 

5 

Zeidae' 

0-1,  6-10 

VII-X,  22-37 

I-IV,  20-39 

11 

7 

29-42 

11-14, 
17-18 

6 

13 

Oreosomatidae- 

I,  5-6 

V-VIII,  29-35 

II-IV,  27-34 

11 

7 

(3 

genera)' 

35-43 

17-22 

4(5?) 

10 

Zeniontidae 

1.5-6 

VI-VII,  25-29 

O-II,  23-32 

11 

7-8 

25-27 

17" 

3? 

4? 

Parazenidae' 

0,  7 

VIII,  26-30 

I,  31-33 

9 

7 

34 

15-16 

1 

1 

Caproidae" 

I,  5 

Vll-X,  26-37 

III,  23-34 

10 

5-6 

21-23' 

I. 

11-14 

2 

6 

Macrurocyttidae' 

I,  3 

V,  27 

T> 

7 

ca.  5 

7 

15 

1 

1 

'  Bray,  1983. 

-  Karrer  and  Eschmeyer,  Ms. 

'Mead.  1957. 

'  Berry.  1959a. 

>  Fowler.  1934. 

•McAlhsIer.  1968. 

'  Rosen,  1973. 

"  Myers,  1937. 

'Quero,  1978. 


noid  scales  while  the  pelagic  prejuveniles  are  oval  in  outline  and 
possess  a  leathery  skin  with  distinct  hardened  cones  or  scaley 
knobs  laterally  and  ventrally  (Myers,  1960;  Eschmeyer  et  al., 
1 983).  This  stage  is  exhibited  by  Abe  and  Kaji's  ( 1 972)  specimen 
of  Oreosoma  atlanlicum  (Fig.  2 1 3C).  Karrer  and  Eschmeyer  (in 
press)  report  prejuveniles  of  Pseudocytlus  as  large  as  100  mm 
and  suggest  that  metamorphosis  can  be  delayed.  In  one  species, 
the  transformation  is  incomplete  and  the  species  remains  in  the 
midwater  prejuvenile  habitat  and  becomes  mature  there. 

In  the  Zeidae.  only  Cyttus  traversi  is  presently  known  to  have 
a  prejuvenile  stage.  This  stage  has  a  relatively  much  deeper  body 
than  the  adult,  and  bears  long  filamentous  extensions  with  nu- 
merous appendages  from  the  dorsal  spines  and  pelvic  spine  and 
rays  (James,  1976b:  fig.  1).  The  prejuvenile  stage  occurs  near 


the  surface  in  coastal  waters  (James,  1976b)  while  specimens 
greater  than  about  100  mm  have  been  caught  near  the  bottom 
at  depths  from  200  to  800  meters  (Heemstra,  1980). 

Macrurocytlus  acanthopodiis  was  described  by  Fowler  ( 1 934) 
and  was  placed  in  the  order  Zeiformes  by  Greenwood  et  al. 
(1966).  Its  small  size,  elongate  pelvic  spine,  and  stout  dorsal 
spine  (Fig.  21 3D)  array  suggest  that  this  may  be  a  juvenile  or 
prejuvenile  form,  perhaps  of  the  Zeniontidae  as  suggested  by 
Heemstra  (1980). 

Relationships 

The  present  classification  of  the  Zeiformes  is  based  only  on 
characters  of  the  adults.  Heemstra  (1980)  includes  five  families 
in  the  Zeiformes  but  speculates  that  the  Grammicolepididae 


Table  105.    References  Giving  Descriptions  and/or  Illustrations  of  Eggs,  Larvae  and  Prejuveniles  of  the  Order  Zeiformes. 


Family 


Eggs 


Pre-juveniles 


Grammicolepididae 
Zeidae 


Caproidae 


Oreosomatidae 


Zeniontidae 
Parazenidae 
Macrurocyttidae 


Sanzo,  1931b 
Sanzo,  1956 
Dekhnik,  1973 
Robertson,  1975a 


Cunningham,  1889 
Holt,  1897,  1899 
Sanzo,  1956 
Arbault  and  Boutin,  1968a 


Ehrenbaum,  1905-1909 
Schmidt,  1908 
Sanzo,  1931b 
Sanzo,  1956 
Banarescu,  1964 
Crossland,  1982 
Holt,  1897,  1899 
HefTord,  1910 
Clark,  1914 
Fage,  1918 
Sanzo,  1956 
Uchida,  1936 
Nakahara,  1962 


Smith,  1949 
James,  1976b 


Cuvier,  1829 
Abe,  1957 

Kobayashi  et  al.,  1968 
Abe  and  Kaji.  1972 


Fowler,  1934 


TIGHE  AND  KEENE:  ZEIFORMES 


395 


Fig.  212.  Zeiform  larvae.  (A)  Yolk-sac  larva  of  Zeus  faber,  4.3  mm  NL  (source:  Sanzo,  1931b);  (B)  Preflexion  larva  of  Zeus  faber.  4.3  mm 
NL  (source:  Crossland,  1982);  (C)  Postflexion  larva  of  Zeusfaher,  7.2  mm  SL  (source:  Crossland.  1982);  (D)  Yolk-sac  larva  of  Capros  aper.  2.9 
mm  NL  (source:  Sanzo.  1956);  and  (E)  Preflexion  larva  of  Capros  aper.  5.0  mm  NL  (source:  Sanzo,  1956). 


396 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  213.  Zeiform  larvae.  (A)  Preflexion  larva  of  Anligonia  rubescens.  4.5  mm  TL  (source:  Uchida,  1936);  (B)  Poslflexion  larva  of  Anligonia 
capros,  4.75  mm  TL  (source;  Nakahara,  1962);  (C)  Prejuvenile  of  Oreosoma  allanlicum.  61  mm  SL  (source:  Abe  and  K.aji,  1972);  and  (D) 
Holotype  o(  Macrurocyttus  acanthopodus  Fowler  1934,  43  mm  SL  (source:  Fowler,  1934). 


TIGHE  AND  KEENE:  ZEIFORMES 


397 


may  prove  to  be  incorrectly  placed  there  because  they  differ 
considerably  in  the  configuration  of  their  jaw  elements,  scales, 
number  of  vertabrae.  and  have  a  higher  number  of  caudal  rays. 
Heemstra's  decision  to  exclude  the  Caproidae  from  the  Zei- 


formes  is  supported  by  evidence  from  Rosen  (1973),  who  dis- 
cusses some  similarity  between  zeoids  and  caproids  but  states 
that  the  pelvic  count  of  1  spine  and  5  rays,  3  anal  spines,  and 
the  reduced  vertebral  number  21-23  are  a  combination  of  char- 


398 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


acters  found  among  percoids.  The  possession  of  normal  abdom- 
inal parapophyses,  lack  of  ventral  ridge  scales  or  bucklers,  and 
a  percoid  type  of  caudal  skeleton  suggest  to  Rosen  that  caproids 
appear  to  fit  the  present  definition  of  a  perciform  while  other 
zeoids  do  not. 

These  findings  support  the  movement  of  the  Caproidae  higher 
in  Acanthopterygian  classification.  The  very  different  larvae  of 
the  two  caproid  genera  suggest  that  a  thorough  reexamination 
of  the  osteology  of  adult  representatives  of  these  genera  could 
be  necessary  before  the  family  is  placed  somewhere  else. 

There  has  been  no  phylogenetic  systematic  study  of  the  order 
Zeiformes.  Inclusion  of  early  life  history  characters  would  prob- 
ably be  useful  in  such  a  study,  but  these  are  unknown  for  most 
members  of  the  order. 


group.  Rosen  has  suggested  that  the  Zeiformes  do  not  represent 
a  monophyletic  lineage,  but  are  best  included  within  the  Te- 
traodontiformes  with  which  they  are  united  by  seven  synapo- 
morphies.  Within  Rosen's  classification,  the  Caproidae  are  the 
sister  group  to  the  rest  of  the  Tetraodontiformes.  In  addition, 
the  rest  of  the  zeiform  families  are  united  with  the  plectognath 
fishes  by  four  synapomorphies  while  the  plectognath  families 
are  monophyletic  on  the  basis  of  six  synapomorphies.  Evidence 
from  early  life  history  characters  supporting  this  classification 
is  very  limited  due  to  the  lack  of  knowledge  of  the  early  life 
history  of  most  of  these  fishes,  but  the  similarity  in  morphology 
and  pigmentation  between  newly  hatched  Zens  faher  larvae  and 
tetraodontid  larvae  does  provide  some  support  for  Rosen's  hy- 
pothesis. 


Addendum:  After  this  paper  went  to  press.  Rosen  (1984)  pub- 
lished a  phylogenetic  analysis  of  the  families  (except  Macru- 
rocyttidae)  herein  included  in  the  order  Zeiformes  which  re- 
sulted in  a  drastic  change  in  the  systematic  placement  of  this 


Division  of  Fishes,  National  Museum  of  Natural  History, 
Smithsonian  Institution,  Washington,  District  of 
Columbia  20560. 


Gasterosteiformes:  Development  and  Relationships 
R.  A.  Fritzsche 


THE  actinopterygian  fish  order  Gasterosteiformes  contains  a 
diverse  assemblage  of  specialized  fishes.  There  are  about 
220  species  arranged  into  10  or  II  families  (Fritzsche.  1982). 
Historically  this  group  has  been  divided  into  two  or  three  orders, 
under  such  names  as  Lophobranchii,  Thoracostei,  Solenich- 
thyes,  Catosteomi,  Hemibranchii,  Hypostomides,  Gasterostei- 
formes, Syngnathiformes,  and  Pegasiformes  (Boulenger,  1904; 
Berg,  1940;  and  Starks,  1902).  Pietsch  (1978b)  presented  infor- 
mation which  suggests  that  Pegasiformes  are  intermediate  be- 
tween the  Gasterosteiformes  and  Syngnathiformes.  Pegasids  are 
intermediate  in  (1)  snout  development  and  in  the  condition  of 
the  nasal  bones;  (2)  retention  of  the  parietals;  (3)  retention  of 
three  circumorbital  bones;  (4)  presence  of  a  dorsal  strut  join- 
ing the  ceratohyal  and  epihyal;  (5)  reduction  in  number  of  ele- 
ments of  the  branchial  arches;  (6)  the  presence  of  two  pairs  of 
pleural  ribs;  and  (7)  retention  of  support  for  a  spinous  dorsal 
fin  (Pietsch,  1978b).  He  proposed  a  tentative  classification  unit- 
ing all  three  groups  into  the  single  order  Gasterosteiformes.  This 
order  is  characterized  by  (1)  branchiostegal  rays  reduced  to  1- 
5;  (2)  absence  of  supramaxillary,  orbitosphenoid,  and  basi- 
sphenoid;  (3)  postcleithrum  reduced  to  single  bone  or  absent; 
(4)  pelvic  girdle  never  attached  directly  to  cleithra;  (5)  rather 
small  mouth,  often  at  end  of  more  or  less  tubular  snout;  and 
(6)  armor  of  dermal  plates  covers  most  members  (Fritzsche, 
1982).  Pegasids  form  the  primitive  sister-group  of  the  Soleno- 
stomidae  and  Syngnathidae.  These  families  share  a  number  of 
derived  character  states  including  (1)  feeding  mechanism;  (2) 
metapterygoid  absent;  (3)  hyoid  apparatus  short,  bearing  elon- 
gate, filamentous  branchiostegal  rays;  (4)  gill  opening  restricted 
to  a  small  hole  on  the  dorsolateral  surface  behind  head;  (5)  gill 
filaments  tufted  or  lobe-like;  (6)  articular  processes  of  mobile 
vertebral  centra  absent;  (7)  posttemporal  co-ossified  with  cra- 
nium; (8)  postcleithrum  absent;  and  (9)  head  and  trunk  encased 
by  bony  plates,  tail  encircled  by  bony  rings  (Pietsch,  1978b). 
The  Pegasidae,  Solenostomidae  and  Syngnathidae  form  the 
primitive  sister-group  of  the  Macrorhamphosidae,  Centriscidae, 


Aulostomidae,  and  Fistulariidae  and  the  resulting  classification 
is  as  follows: 

Order  Gasterosteiformes 
Suborder  Gasterosteoidei 

Superfamily  Aulorhynchoidea 
Family  Aulorhynchidae 


B 


Fig.  214.  Eggs  of  some  gasterosteiforms;  (A)  Gasterosleus  acuteatus 
(from  Kuntz  and  RadclifTe,  1917);  (B)  Fistularta  pelimha  (from  Mito, 
1 96 1  a);  (C)  Macrorhamphosus  scolopa.x  (horn  Hardy,  1 978a,  after  Spar- 
ta, 1936);  (D)  l/ippocumpus  ereclus  (from  Hardy,  1978a). 


FRITZSCHE:  GASTEROSTEIFORMES 


399 


B 


"T-^Mcji.^  -,■<:,;  -.-v.v?  >»,'.c v>7.:- 


j*tf»i^  iJo^T-vsrrrr 


-■■*g 


Fig.  215.     Larvae  of  some  gasterosteoids.  (A,  B)  Aulorhynchus  flavidus.  8  mm  TL  and  23  mm  TL  (from  Marliave,  1 976);  (C)  Apeltes  quadracus. 
6.0  mm  TL  (from  Ryder,  1887);  (D)  Apeltes  quadracus.  10.5  mm  TL  (from  Hardy,  1978a). 


Family  Hypoptychidae 
Superfamily  Gasterosteoidea 

Family  Gasterostcidae 
Suborder  Syngnathoidei 
Infraorder  Syngnatha 
Superfamily  Pegasoidea 

Family  Pegasidae 
Superfamily  Syngnathoidea 

Family  Solenostomidae 

Family  Syngnathidae 
Infraorder  Macrorhamphosa 
Superfamily  Macrorhamphosoidea 

Family  Macrorhamphosidae 

Family  Centriscidae 
Superfamily  Aulostomoidea 

Family  Aulostomidae 

Family  Fistulariidae 


The  taxonomy  within  this  order  is  poorly  understood.  The 
lack  of  agreement  regarding  relationships  within  the  Gasteros- 
teiis  aculeatus  complex  (Bell,  1976)  and  whether  or  not  Macro- 
rhamphosus  contains  only  one  species  (Ehrich,  1 976)  are  two 
examples.  Recent  studies,  such  as  that  of  Fritzsche  (1980),  have 
shown  that  many  species  of  syngnalhids  are  morphologically 
plastic.  This  plasticity  has  been  the  cause  of  a  proliferation  of 
species  and  subspecies  descriptions  in  the  literature.  The  process 
of  sorting  out  the  nominal  species  still  continues  for  most  taxa 
included  in  Gasterosteiformes. 

Gasterosteiforms  are  found  in  freshwater,  estuarine,  and  ma- 
rine habitats  through  tropical  and  temperate  regions.  Most  species 
are  relatively  small  and  cryptically  colored.  They  have  no  real 
fishery  importance  and  usually  are  thought  of  as  interesting 
aquarium  fishes  or  simply  curiosities,  e.g.  the  seahorse.  Since 
commercial  importance  is  lacking,  there  is  very  little  literature 
dealing  with  the  early  life  histories  of  these  fishes  except  for 


400 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


.  ,^      i_i — i — L-iLJLJ) 


Fig.  216.  Larvae  of  some  pegasoids  and  syngnathoids.  (A)  Pegasidae,  2.4  mm  (from  Leis  and  Rennis,  1983);  (B)  Solenoslomus  sp.,  5.1  mm 
NL  (ongmal  illustration  by  Wayne  A.  Laroche);  (C)  Syngnathus  fuscus.  ca.  3.5  mm  TL  (from  Ryder,  1887);  (D)  Hippocampus  japomcus.  ca.  6 
mm  TL  (from  Nakamura,  1937). 


anecdotal  accounts  or  descriptions  of  chance  collections  of  eggs 
or  young. 

Development 

There  are  published  descriptions  of  the  eggs  of  Aulorhynchus 
(Limbaugh,  1962;  Ida,  1976),  Hypoptychiis(\s\\i%aku  1957;  Ida, 
1976),  gasterosteids  (notably  Kuntz  and  Radclifle,  1917;  Vrat, 
1949;  Swamp,  1958),  Solenoslomus  (Padmanabhan,  1961), 
Macrorhamphosus  (Sparta,  1936),  and  Fistulana  (Delsman, 
1921;  Mito,  1961a;  Watson  and  Leis,  1974).  There  are  few 
descriptions  of  the  eggs  of  syngnathids  due  to  the  unique  male 


brooding  habits  of  this  group,  however,  Hudson  and  Hardy 
(1975)  provided  a  good  description  of  Hippocampus  erectus 
eggs.  Most  accounts  simply  include  the  number  and  size  of  eggs 
in  the  male's  pouch  (e.g.,  Fritzsche,  1980).  Gudger  (1905)  pro- 
vided a  fairly  extensive  treatment  of  the  embryology  of  Syng- 
nathus floridae. 

Larvae  (usually  just  one  or  two  and  not  a  series)  have  been 
described,  for  .-l«/o/-/;i«c/!i« (Limbaugh,  1962;  Marliave,  1976), 
gasterosteids  (Kuntz  and  Radcliffe,  1917;  Vrat,  1949;  Swarup, 
1958),  pegasids  (Jones  and  Pantulu,  1 958;  Jones  and  Kumaran, 
1967;  Leis  and  Rennis,  1983),  Solenoslomus  (Padmanabhan, 


FRITZSCHE:  GASTEROSTEIFORMES 


401 


B 


Fig.  217.     Larvae  of  some  macrorhamphosoids  and  some  aulostomoids.  (A)  Macrorhamphosus  scolopax.  3.0  mm  TL  (from  Hardy,  1978a, 
after  Sparta,  1936);  (B)  Centriscidae.  2.7  mm  (from  Leis  and  Rennis.  1983);  (C)  Fistularia  petimba.  7.08  mm  (from  Mito,  1961a). 


1961),  syngnathids  (most  notably  D'Ancona,  1 933c;  Nakamura, 
1937;  Takai  and  Mizokami,  1959;  James,  1970;  Russell,  1976; 
Dawson  et  al.,  1979),  macrorhamphosids  (D'Ancona,  1933d; 
Sparta,  1936;  Mohr,  1937),  centriscids  (Mohr,  1937;  Leis  and 


Rennis,  1983)  and  F/5n//ana  (Jungersen,  1910;  Delsman,  1921; 
Mito,  1961a;  Leis  and  Rennis,  1983).  Larvae  have  not  been 
described  for  Hypoplychus  and  Aulostoinus. 
Osteological  development  has  not  been  studied  for  most  gas- 


Table  106.    Meristic  Characters  for  Families  of  the  Gasterosteiformes  (adapted  from  Pietsch,  1978b). 


Hypo. 

ptychi- 

Oaster- 

Soleno- 

Syng- 

Macrorham- 

Fislu- 

Character 

Aulorhynchidae 

dae 

ostcidac 

Pegasidac 

stomidac 

nathidae 

phosidae 

Cenlnscidae 

Aulostomidae 

lanidae 

Circumorbital  bones 

3 

T 

3 

3 

0 

2-3 

1 

1  (  +  2'') 

1  (O'^) 

0 

Branchiostegal  rays 

4 

4 

3-4 

5 

1  (bifid) 

1-3 

4-5 

3-5 

4 

5 

Vertebrae 

52-56 

55 

28-42 

19-22 

33 

37-77 -h 

24 

20 

59-64 

76-87 

Elongate  anterior 

0 

0 

0 

6 

3 

3 

5 

5-6 

4 

4 

vertebrae 

Pleural  nbs  (pairs) 

0-22 

9 

9-16 

-) 

0 

0 

0 

0 

0 

0 

Dorsal-fin  rays 

XXIV-XXVI 
+  9-10 

20 

III-XVI 

-1-  6-14 

5 

V  -h  18-23 

0-60 

IV-VII 
+  9-11 

III 

+  10-12 

VIII-XIII 

-1-  21-26 

14-20 

Anal-fin  rays 

I  -1-  9-10 

20 

I  -F  6-12 

5 

16-23 

0-6 

19-20 

11-12 

22-27 

14-19 

Caudal-fin  rays 

13 

13 

11-12 

8 

16 

0-11 

23 

11 

20 

22-24 

Pectoral-fin  rays 

10-11 

9 

9-23 

10-18 

18-27 

0-23 

15 

10-12 

15-16 

13-18 

Pelvic-fin  rays 

1  +  4 

0 

I  +  0-2 

I  +  2-3 

I  +6 

0 

1  +  4 

I  +  4 

6 

5-6 

402 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  218.  Larvae  of  some  pegasoids  and  syngnathoids.  (A)  Eurypegasus  papilio.  7.0  mm  (from  Leis  and  Rennis,  1983);  (B)  Solenoslomus  sp., 
11.5  mm  SL  (original  illustration  by  Wayne  A.  Laroche);  (C)  Syngnathus  schlegeli.  size  unknown  (from  Chyung,  1977);  (D)  Yozia  bicoarctata. 
ca.  10-1 1  mm  SL  (from  Dawson  et  al.,  1979);  (E)  Hippocampus  japomcus.  ca.  6.5  mm  TL  (from  Nakamura,  1937). 


FRITZSCHE:  GASTEROSTEIFORMES 


403 


B 


Fig.  219.     Larvae  of  some  macrorhamphosoids  and  aulostomoids.  (A)  Macrorhamphosus  scolopax.  9.0  mm  TL  (from  Hardy,  1978a,  after 
D'Ancona,  \97>'!i(X)\(Q)  Aeoliscus stngalus.  7.9  mm  (from  Leisand  Rennis,  \9S3y,  {C)  Fisliilaria pelimba.  15.6  mm  (from  Leis  and  Rennis,  1983). 


terosteiforms.  Kindred  (1921)  presented  a  classic  study  on  the 
chondrocranium  of  Syngnalhus  fusciis.  Padmanabhan  (1961) 
published  information  on  the  development  of  jaws  in  Solenos- 
tomus  cyanopterus.  Development  of  the  bony  rings  on  the  body 
ofSyngnaihus  typhle  was  studied  in  detail  by  Czolowska  ( 1 962). 

Considering  the  diversity  of  habitats  and  spawning  behaviors 
found  within  the  group,  it  is  difficult  to  identify  a  character  or 
suite  of  characters  that  typifies  all  members  of  this  order.  Some 
gasterosteiforms  spawn  in  open  water  and  produce  buoyant  eggs 
(e.g.,  Fistularia.  Watson  and  Leis,  1974);  others  such  as  the 
sticklebacks  and  tubesnouts  (Gasterosteidae  and  Aulorhynchi- 
dae)  construct  nests  out  of  vegetation  for  receipt  of  the  eggs; 
while  others  such  as  the  seahorses,  pipefishes,  and  ghost  pipe- 
fishes (Syngnathidae  and  Solenostomidae)  brood  the  eggs  within 
specialized  structures  located  on  one  of  the  parents.  Syngnathids 
have  a  most  unusual  adaptation  in  having  a  specialized  patch 
or  pouch  (marsupium)  developed  on  the  males  for  receipt  and 
incubation  of  eggs.  Those  groups  containing  species  that  broad- 
cast spawn  or  have  nests  produce  larvae  that  go  through  the 
typical  developmental  pattern  of  pelagic  larvae.  Those  that  brood 
eggs,  such  as  the  more  advanced  syngnathids,  may  retain  the 
eggs  and  developing  larvae  until  the  young  have  reached  a  ju- 
venile stage  of  development. 

In  general,  eggs  of  most  gasterosteiforms  are  spherical,  how- 
ever, those  of  Hippocampus  have  been  described  as  being  dis- 


tinctly pear-shaped  (Hudson  and  Hardy.  1975)  or  ellipsoidal 
(Nakamura,  1937)  (Fig.  214).  The  eggs  typically  have  numerous 
oil  droplets  in  the  yolk  (Gudger,  1905;  Kuntz  and  Radcliffe, 
1917).  However,  those  of  Fistularia  lack  oil  droplets  (Watson 
and  Leis,  1974),  and  Macrorhamphosus  has  a  single  oil  globule 
(Lo  Bianco,  1909;  Fage,  1918).  The  perivitelline  space  is  narrow 
in  Solenostomus  (Padmanabhan,  1961),  gasterosteids  (Hardy, 
1978),  and  Fistularia  (Mito,  1961a),  while  it  is  relatively  wide 
in  Hippocampus  (Hardy,  1 978a).  The  yolk  is  not  segmented  and 
is  typically  yellow  in  syngnathids  (James,  1970),  rose- violet  in 
Macrorhamphosus  (Hardy,  1 978a),  and  clear  in  Fistularia  (Mito, 
1961a).  The  chorion  is  typically  smooth,  however,  small  at- 
tachment threads  have  been  reported  for  some  gasterosteids 
(Hardy,  1978a).  Most  gasterosteiforms  have  eggs  about  1.0mm 
in  diameter  except  that  Solenostomus  eggs  are  about  0.6  mm 
(Padmanabhan,  1961)  and  Hippocampus  eggs  may  approach 
4.0  mm  in  one  dimension  (Hardy,  1978a). 

Larvae  of  most  gasterosteiforms  (except  gasterosteids)  have 
a  very  distinctive,  elongate  snout  bearing  a  small  upturned  mouth 
which  reflects  a  trenchant  character  of  the  adults  (Figs.  2 1 5- 
219).  Meristic  characters  are  quite  variable  in  this  order  (Table 
106).  Myomere  counts  range  from  a  low  of  19  in  pegasids  to  87 
in  Fistularia  (Leis  and  Rennis,  1983).  Fin  ray  meristics  are 
equally  variable  and  some  groups  lack  one  or  all  of  the  fins 
(Table  106).  Syngnathids,  for  example,  may  have  0  to  60  dorsal 


404 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


fin  rays.  Size  at  hatching  has  not  been  well  documented  for 
gasterosteiforms.  Gasterosteids  and  Solenostomus  may  hatch  at 
3.0  mm  TL  (Padmanabhan,  1961;  Hardy,  1978a),  while  Aulo- 
rhynchus  hatch  at  5.5-8.0  mm  TL  (Marliave,  1976).  Presence 
of  bony  plates  rather  than  scales  is  the  rule  in  this  order.  These 
plates  are  typically  present  and  easily  seen  by  the  time  notochord 
flexion  is  complete  (Figs.  218  and  219).  Several  groups  develop 
small  spinules  in  the  skin  on  the  body  early  in  development. 
Macrorhamphosus  develops  spinules  at  about  6  mm  TL  (Sparta, 
1936).  All  species  of  Fisliilaha  go  through  a  so-called  "villosa 
stage"  (Liitken,  1880)  during  which  they  are  covered  with  small 
spinules  (Fig.  219C).  Pigmentation  of  species  for  which  larvae 
have  been  described  varies  from  very  heavy  pigmentation  in 
Gasterosteidae  and  Macrorhamphosidae  to  rather  light  pig- 
mentation in  Syngnathidae  and  Fistulariidae.  The  young  of  sev- 
eral species  of  syngnathids  have  conspicuous  dark  bars  (D'An- 
cona,  1933c;  Takai  and  Mizokami,  1959;  and  Fritzsche,  1980) 
(Fig.  2 1 8C).  Dawson  et  al.  ( 1979)  reported  the  presence  of  elon- 
gate dermal  appendages  in  young  of  the  syngnathid  genus  Yozia 
(Fig.  218D).  They  believed  that  these  appendages  have  a  buoy- 
ant function  for  aid  in  distribution  of  the  pelagic  young. 

Relationships 

Besides  the  hypothesis  of  relationships  proposed  by  Pietsch 
(1978b),  there  are  several  other  recent  hypotheses.  Greenwood, 
et  al.  (1966)  proposed  the  following  classification  scheme: 

Order  Gasterosteiformes 
Suborder  Gasterosteoidei 

Family  Gasterosteidae 

Family  Aulorhynchidae 

Family  Indostomidae 
Suborder  Aulostomoidei 

Family  Aulostomidae 

Family  Fistulariidae 

Family  Macrorhamphosidae 

Family  Centriscidae 
Suborder  Syngnathoidei 

Family  Solenostomidae 

Family  Syngnathidae 
Order  Pegasiformes 

Family  Pegasidae 

The  family  Indostomidae  has  at  various  times  been  thought  to 
be  related  to  the  gasterosteiforms  (Bolin,  1936b;  Berg,  1940). 
But,  Pietsch  (1978b)  has  pointed  out  that  the  specific  relation- 
ship of  this  family  must  await  further  investigation.  I  have, 
therefore,  not  included  this  monotypic  family  (I ndostoiniis  par- 
adoxus) in  this  account. 

Banister  (1967)  proposed  a  classification  based  on  his  osteo- 
logical  studies  as  follows: 

Order  Aulorhynchiformes 

Family  Aulorhynchidae 

Family  Gasterosteidae 
Order  Aulostomiformes 
Suborder  Aulostomoidei 

Family  Aulostomidae 

Family  Fistulariidae 

Family  Solenostomidae 

Family  Syngnathidae 
Suborder  Centriscoidei 

Family  Macrorhamphosidae 

Family  Centriscidae 


His  scheme  differs  little  from  previous  ideas  except  in  use  of 
new  ordinal  names  (to  reduce  confusion?)  and  inclusion  of  the 
closely  related  macrorhamphosids  and  centriscids  in  their  own 
suborder.  Characters  of  his  Centriscoidei  are  ( I )  separate  meta- 
pterygoid  present  and  anterior  end  of  quadrate  normal;  (2)  nasals 
large  and  elongated;  (3)  five  or  more  modified  anterior  vertebrae; 
(4)  supraethmoid  contributes  little  to  dorsum  of  snout;  (5)  post- 
temporal  pyramidal;  (6)  caudal  fin  skeleton  uniform,  with  single 
large  hypural  plate;  (7)  vertebral  number  low  (about  20);  (8)  no 
sign  of  reduction  in  pharyngeal  skeleton;  and  (9)  intemeurals 
for  vertebrae  five  and  six  absent.  Banister's  (1967)  hypothesis 
of  relationships  has  not  been  published. 

Nelson  (1976)  proposed  a  classification  that  was  similar  to 
that  of  Greenwood  et  al.  (1966)  except  that  the  families  Gas- 
terosteidae and  Aulorhynchidae  were  recognized  as  forming  the 
order  Gasterosteiformes  while  the  remainder  of  the  families 
were  placed  in  Syngnathiformes.  This  separation  was  done 
pending  clarification  of  relationships  and  establishment  of 
monophyly.  As  noted  earlier,  Pietsch  (1978b)  was  able  to  link 
the  two  groups  based  on  the  intermediate  nature  of  the  pega- 
soids. 

Ida  (1976)  demonstrated  that  the  monotypic  Hypopiychus 
dybowskii  Steindachner  resembled  gasterosteids  and  aulorhyn- 
chids  in  osteology,  mode  of  life,  and  reproduction.  He,  therefore, 
removed  this  species  from  the  Perciformes  and  placed  it  close 
to  the  Gasterosteidae  and  Aulorhynchidae  in  the  suborder  Gas- 
terosteoidei of  his  order  Syngnathiformes. 

Early  life  history  stages  have  contributed  little  to  the  devel- 
opment of  the  above  hypotheses  of  relationships.  Pietsch  ( 1 978b) 
showed  that  snout  structure  of  Pegasus  and  Macrorhamphosus 
is  very  much  alike  at  small  sizes  even  though  it  is  quite  different 
in  adults.  Ida  (1976)  used  egg  morphology  as  one  of  the  char- 
acters supporting  his  placement  of  the  Hypoptychidae  close  to 
the  Gasterosteidae. 

Considering  the  paucity  of  developmental  descriptions  for 
species  of  the  Gasterosteiformes,  it  is  difficult  to  test  existing 
hypotheses  of  relationships  using  developmental  characters. 
However,  it  is  interesting  to  note  the  sequence  of  fin  formation 
seems  to  support  the  close  relationship  of  the  Gasterosteidae 
and  Aulorhynchidae.  Aulorhynchus  forms  the  pectoral  fins  first, 
followed  by  the  caudal,  second  dorsal  and  anal  fins  (Marliave, 
1976).  The  gasterosteid  Apeltes  follows  the  same  sequence 
(Hardy,  1978a).  Gasterosteus  forms  the  pectoral  fins  after  the 
anal  fin  (Hardy,  1978a).  Few  developmental  sequences  are  known 
for  the  other  gasterosteiforms.  Those  that  are  available  show 
that  for  the  pegasids,  macrorhamphosids  and  syngnathids  the 
sequence  begins  with  the  development  of  the  dorsal  fin  followed 
by  the  anal,  caudal  and  pectoral.  It  may  well  be  that  the  sequence 
of  fin  formation  will  provide  evidence  for  the  retention  of  the 
Gasterosteidae  and  Aulorhynchidae  in  their  own  order  or  sub- 
order. Additionally  Macrorhamphosus.  Acoliscus  and  Fislularia 
develop  a  dorsal  finfold  that  extends  on  to  the  head  which  might 
be  given  as  evidence  in  support  of  Pietsch's  (1978b)  infraorder 
Macrorhamphosa.  However,  pegasids  also  have  this  anteriorly 
placed  finfold  (Leis  and  Rennis,  1983).  This  coupled  with  the 
low  myomere  numbers  for  pegasids  and  macrorhamphosids  may 
indicate  that  these  two  groups  should  be  placed  closer  together 
than  is  presently  indicated  in  Pietsch's  treatment.  This  question 
must  remain  unresolved  pending  further  descriptive  and  com- 
parative work  on  gasterosteiform  larvae. 

Studies  of  the  relationships  of  Gasterosteiformes  to  other  taxa 
have  been  dominated  by  unsupported  hypotheses.  Gosline(197I) 
proposed  that  the  "origin  for  both  gasterosteoids  and  synga- 


FRITZSCHE:  GASTEROSTEIFORMES 


405 


thoids  (sic)  suggest  one  or  two  origins  in  the  percopsiform  — 
beryciform  area."  The  mixture  of  advanced  and  primitive  char- 
acters shown  by  gasterosteiforms  suggested  to  Banister  (1967) 
evolution  "from  a  primitive  myctophoid  type  offish  .  .  .  towards 
an  acanthopterygian  grade."  McAllister  (1968)  suggested  "the 
Gasterosteiformes  are  derivable  from  the  Perciformes"  and  ".  .  . 
the  Syngnathiformes  from  the  subperciforms,  such  as  Beryci- 
formes  and  Zeiformes."  In  fact  he  suggests  that  Antigonia  or 
Capromimus  would  appear  to  be  close  to  the  ancestors  of  the 
Syngnathiformes.  None  of  these  authors  presented  evidence  for 
support  of  their  ideas.  Examination  of  the  description  o^  Anti- 
gonia larvae  by  Nakahara  ( 1 962)  shows  that  this  fish  bears  little 
resemblance  to  the  early  stages  of  described  gasterosteiforms. 
Larval  Antigonia  are  characterized  by  well-developed,  serrated 


preopercular  and  cranial  spines.  These  spines  are  never  seen  in 
gasterosteiform  larvae.  However,  the  description  of  the  larvae 
of  Capros  aper  (Russell,  1976)  indicates  that  the  most  charac- 
teristic feature  of  them  is  the  occurrence  of  small  spines  all  over 
the  body  surface.  Additionally  the  larvae  of  C.  aper  are  darkly 
pigmented.  These  two  characteristics  are  also  found  in  some 
gasterosteiform  larvae,  e.g.,  Macrorhamphosus.  It  is  therefore 
tempting  to  use  these  characters  in  support  of  McAllister's  hy- 
pothesis, however  we  will  have  to  wait  for  further  information 
on  both  gasterosteiforms  and  zeiforms  before  we  can  support 
or  refute  this  hypothesis. 

Department  of  Fisheries,  Humboldt  State  University,  Ar- 
CATA,  California  95521. 


Scorpaeniformes:  Development 
B.  B.  Washington,  H.  G.  Moser,  W.  A.  Laroche  and  W.  J.  Richards 


THE  Scorpaeniformes  are  the  fourth  largest  order  of  fishes 
encompassing  about  20  families  (depending  on  classifi- 
cation used),  250  genera  and  over  1 ,000  species.  Representatives 
of  the  order  are  widely  distributed  from  tropical  to  arctic  and 
antarctic  waters.  Most  scorpaeniforms  are  benthic  or  epibenthic 
with  representatives  ranging  from  freshwater  to  the  deep  ocean. 

The  morphologically  diverse  "mail-cheeked  fishes"  are  named 
for  the  bony  suborbital  stay  which  extends  posteriorly  from  the 
third  infraorbital  to  the  preopercle.  The  suborbital  stay  is  the 
only  known  character  that  defines  the  order;  however,  some 
workers  have  suggested  that  the  stay  evolved  independently  in 
several  lineages  and  may  not  indicate  monophyly  (Matsubara, 
1943;  Quast,  1965;  Poss,  1975).  The  classification  of  the  scor- 
paeniforms is  controversial,  not  only  in  terms  of  monophyly 
but  also  at  the  subordinal  and  familial  levels.  Discussion  of  the 
taxonomic  status  and  current  hypotheses  of  relationships  is  pre- 
sented in  Scorpaeniformes:  Relationships  (this  volume). 

Modes  of  reproduction  vary  widely  within  the  scorpaeni- 
forms. Many  families  spawn  individual  pelagic  eggs  (Anoplo- 
pomatidae,  Congiopodidae,  Hoplichthyidae  and  Triglidae),  while 
others  spawn  demersal  clusters  of  adhesive  eggs  ( Agonidae.  Cot- 
tidae,  Cyclopteridae  and  Hexagrammidae).  Where  known,  most 
scorpaenids  produce  pelagic  egg  masses  enclosed  in  a  gelatinous 
matrix.  Notable  exceptions  include  the  scorpaenid  genus  Se- 
bastes  and  the  comephorids  of  Lake  Baikal  which  give  birth  to 
live  young. 

Larvae  of  only  about  20%  of  scorpaeniform  genera  and  ap- 
proximately 10%  of  the  species  are  known.  Because  of  the  wide 
diversity  of  form,  we  are  not  able  to  characterize  a  typical  scor- 
paeniform larva.  Early  life  stages  of  many  scorpaeniforms  are 
characterized  by  strong  head  spination  as  depicted  in  the  gen- 
eralized scorpaenid  larva  Sebastes  (Fig.  220).  However,  the 
expression  of  head  spination  is  variable  within  the  order  with 
elaborations  and  losses  in  many  groups. 

For  the  purposes  of  this  paper,  we  consider  the  Scorpaeni- 
formes to  be  monophyletic  and  utilize  the  broad  suborders  Scor- 


paenoidei  and  Cottoidei  as  a  framework  for  presentation  and 
discussion.  Because  of  the  order's  morphological  diversity  and 
the  lack  of  an  agreed  upon  classification,  discussion  of  larval 
taxonomy  is  focused  upon  each  family.  The  scorpaeniform  fam- 
ily Cyclopteridae  is  presented  in  the  subsequent  article  in  this 
volume. 

SCORPAENOIDEI 

Eggs 

Eggs  are  known  for  seven  of  the  scorpaenoid  families  recog- 
nized in  Washington  et  al.  (this  volume),  however,  they  are 
known  only  for  a  few  species  (Table  107).  Most  scorpaenoid 
families  are  oviparous  and  spawn  pelagic  eggs;  however,  repro- 
ductive modes  are  varied  in  the  Scorpaenidae.  In  the  scorpaenid 
subfamilies  Scorpaeninae,  Pteroinae,  and  Sebastolobinae  the 
eggs  are  extruded  in  bilobed  gelatinous  egg  masses  which  float 
at  the  surface.  The  eggs  are  slightly  elliptical  and  have  homo- 
geneous yolk,  a  narrow  perivitelline  space,  and  a  smooth  cho- 
rion. A  single  oil  globule  is  present  in  Pterois  (0.16-0.17  mm) 
and  Sebastolobus  (0. 18-0.20  mm);  Scorpaena  lacks  an  oil  glob- 
ule. In  the  choridactyline  genus  Inimicus,  eggs  are  extruded 
singly,  are  spherical,  and  lack  an  oil  globule  (Table  107).  Mem- 
bers of  the  scorpaenid  subfamily  Sebastinae  are  viviparous  and 
give  birth  to  large  broods  of  young  which  are  comparable  in 
stage  of  development  to  first-feeding  larvae  of  oviparous  scor- 
paenids. The  eggs  are  retained  in  the  lumen  of  the  ovary  after 
ovulation,  range  between  0.75  and  1.9  mm,  have  homogeneous 
yolk,  a  narrow  perivitelline  space,  smooth  chorion,  and  one  to 
many  oil  globules.  For  the  other  families  for  which  eggs  are 
known,  the  eggs  are  pelagic  with  none  to  multiple  oil  globules 
(Table  107). 

Larvae 

At  least  one  larval  stage  is  known  for  64  of  the  more  than 
600  species  of  scorpaenoids  and  for  20  of  the  100+  genera. 
Major  reviews  of  larval  scorpaenoids  include  Sparta  (1956b) 


406 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  107.    Summary  of  Eggs  and  Larval  Size  Characteristics  of  the  Scorpaeniformes  based  on  Available  Literature  (excluding 

Cyclopteridae). 


Family/subfamily 
species 


Type  of  egg 
pelagic  (P). 
demersal 
(D)  or  vivip- 
arous (V) 


Egg  size 
(mm) 


Number 

of  oil 
globules 


Largest  oil 

globule  size 

(mm) 


Body  length  (mm)  at 


Hatchmg 


Transfor- 
mation 


SCORPAENOIDEI 

Scorpaenidae 
Sebastinae' 

Sebastes  capensis 
S.  fasciatus 

V 

V 

- 

- 

S.  marinus 

V 

1.5 

- 

S.  viviparous 
S.  hubbsi 
S.  inermis 
S.  longispinis 

V 
V 
V 

V 

1.36 

1 

S.  marmoratus 
S.  nigricans 

V 
V 

0.75-0.95 
1.6 

1 

many  to 
1 

S.  oblongus 

S.  pachycephalus 

V 
V 

1.56-1.60 
1.5-1.9 

many 
many 

S.  schlegeU 
S.  steindachneri 
S.  taczanowskii 
S.  const ellal  us 

V 
V 

V 

V 

- 

- 

S.  cortezi 
S.  crameri 

V 

V 

— 

— 

S.  dallii 

S.  entomelas 

V 

V 

- 

- 

S.  flavidus 


S.  helvomaculatus 


S.  jordani 
S.  levis 

S.  macdonaldi 
S.  melanops 

V 
V 
V 

V 

S.  melanostomus 

V 

S.  ovalis 

V 

S.  paucispinis 
S.  pinniger 

V 
V 

S.  rufus 

V 

S.  zacentrus 

V 

Sebastes  Type  A 
Helicolenis  dactyloplerus 

V 

V 

Scorpaeninae 
Pontinus  Type  A 
Pontinus  Type  B 


0.2 


1.0 


ca.  0.20 


3.8 
ca.  5.8 

6.7-7.2 

5.4-5.8 
ca.  4.4 

4.5 
5.8-6.1 

ca.  4,5 
6,9-7.0 

7.2-7.5 
6.0-7.0 

ca.  6. 1 
ca,  4,8 
ca,  5.4 
4.0-5.0 

4.1 

ca.  5.7 


5.0 
4,5- 


4.5 


4.1 

5,4 
5.0 
4.0- 


4.6 


5.0 


4.5 

4.9-5.1 

4.6 
4.0 

4.6-4.8 

ca.  4.3 

4.2 
2.2 


<2.3 


6.2-7.0 
8.5-10.0 


7.i 
ca.  6 
ca.  7 
6.4-7 

ca,  8 


ca. 
ca. 


8.5 


<7.1 

7.0-8.3 
8.0-9.3 


6.2-8.0 
9.9-12.9 


7.7-8.0 


6.2-7,2 

ca.  6.8 

7.2-9.7 
ca.  7.8 

6.1-7.6 

7.4-8.5 


7.0- 
6.0- 


■7.6 
■7.9 


ca.  20 


8.5-11.8       ca.  24 


10.6       - 


ca.  18 


ca.  17 
ca.  10 

12-14 
>13 


ca.  17 
16-21 


<20 
21.7-30.6 


23.6-26.7 


12.0-18.6 


8.0-10.0 

27-30 

7.6-10.4 

ca.  19 

7.7-9.0 

ca.  15 

— 

23,2-30.6 

ca.  16 


15 
12.8- 


18.4 


13.7-19.6 


>19 


4.1-4.6         ca.  15 
<  5.0-5.5      ca.  10 


Moseret  al..  1977 

Moseret  al..  1977;  Fa- 
hay,  1983 

Moseret  al.,  1977;  TSn- 
mg,  1961 

T4nmg,  1961 

Uchida  et  al.,  1958 

Harada,  1962 

Takai  and  Fukunaga, 
1971 

Tsukahara,  1962 

Fujita,  1957b,  1959 

Fujita,  1958 

Shiokawa  and  Tsukahara, 
1961 

Sasaki,  1974 

Sasaki,  1974 

Sasaki,  1974 

Moser  and  Butler,  in 
press 

Moseret  al.,  1977 

Westrheim,  1975;  Rich- 
ardson and  Laroche. 
1979 

Moser  and  Butler,  1981 

Laroche  and  Richardson, 
1981;  Moser  and  But- 
ler, in  press 

DeLacy  et  al.,  1964;  Lar- 
oche and  Richardson, 
1980 

Richardson  and  Laroche. 
1979;  Westrheim,  1975 

Moseret  al.,  1977 

Moseret  al..  1977 

Moseret  al.,  1977 

Laroche  and  Richardson, 
1980 

Moser  and  Ahlstrom, 
1978 

Moser  and  Butler,  in 
press 

Moseret  al.,  1977 

Waldron,  1968;  Richard- 
son and  Laroche,  1979 

Moser  and  Butler,  in 
press 

Laroche  and  Richardson, 
1981;  Westrheim,  1975 

Moseret  al.,  1977 

Graham,  1939;  Sparta, 
1956b;  Tuning,  1961; 
Moser  et  al,,  1977;  Fa- 
hay,  1983 

Moseret  al.,  1977 
Moseret  al.,  1977 


WASHINGTON  ET  AL.:  SCORPAENIFORMES 


407 


Table  107.    Continued. 


Type  of  egg 
pelagic  (P). 

demersal 
(D)  or  vivip 

arous  (V) 

Egg  si/e 
(mm) 

Number 

of  oil 
globules 

Largest  oil 

globule  size 

(mm) 

Body  length  (mm)  at 

Family/subfamily 
species 

Hatching 

FIcMon 

Transfor- 
mation 

References 

Scorpaena  guttata 

P 

1.22- 
1.29  X 
1.16- 
1.19 

0 

1.9-2.0 

4.5-5.7 

>13 

DavKJ,  1939;  Onon, 
1955d;  Moseret  al., 
1977 

S.  notata 

P 

0.76  X  0.88 

0 

— 

<2.7 

ca.  6.0 

— 

Sparta,  1956b 

S.  porcus 

P 

0.84  X  0.92 

0 

— 

1.72 

ca.  6.7 

ca.  12 

Sparta,  1956b 

S.  scrofa 

P 

0.68  X  0.88 

0 

— 

<2.8 

ca.  6.0 

ca.  17 

Sparta,  1956b 

Scorpaena  Type  A 
Scorpaenodes  xyris 

- 

- 

- 

- 

ca.  2.0 
1.8 

4.0-5.5 
4.0-5,4 

>12 

11-14 

Moseret  al.,  1977 
Moseret  al.,  1977 

Pteroinae 

Pterois  lunulata 

P 

0.81-0.83 

1 

0.16-0.17 

1.52-1.58 

- 

- 

Mitoand  Uchida.  1958; 
Mito,  1963 

Dendrochtrus  brachypte- 

P 

- 

0 

- 

ca.  1.1 

- 

- 

Fishelson,  1975 

rus 

Sebastolobinae 
Sebastolobus  alascanus 

S.  allivelis 
Setarchinae 

Eclreposehastes  imus 
Choridactylinae 

Inimicus  japonicus 


Minoinae 

Minous  sp.  (?) 
Triglidae 

Chelidonichthys  cuculus 

C.  gurnardus 

C.  kumu 

C.  lastoviza 

C.  lucerna 

C.  obscurus 

Lepidotrigla  alata 

L.  aspera 

L.  japonica 

L.  microptera 

Prionotus  carolinus 

P.  evolans 

Peristediidae 

Peristedion 

Congiopodidae 

Congiopodus 

leucopaecilus^ 
C.  spinifer^ 


C.  torvus^ 
Platycephalidae 

Platycephalus  indicus^ 

Platycephalidae  spp. 


P 

P 


1,2-1.4 


1.2-1.4 


1.31-1.43 


0.18-0.20     ca.  2.6 


0.18-0.20     ca.  2.6 


<2.8 


6.0-7.3 


6.0-7.3 


ca.  5.5 


3.18-3.27        6.4-8.2 


ca.  1.8 


3.7-5,9 


14-20  Pearcy.  1962;  Moser, 
1974;  Moser  et  al., 
1977 

14-20        Moser  etal.,  1977 

ca.  28        Moseret  al.,  1977 


ca.  10.4      Fujita  and  Nakahara, 

1955;  Mito,  1963;  Sha 
etal.,  1981 


>9.0  Leis  and  Rennis,  1983 


p 

1,45-1,65        1 

p 

1,45-1,5          1 

p 

1,20-1.27        1 

p 

1.29-1.33        1 

p 

1.25-1.36        1 

p 

— 

p 

1,22-1.25        1 

p 

1,16                 1 

p 

1,20-1.40        1 

p 

1.26-1.31        1 

p 

0.94-1,15        1 

0,19-0,33  3,2  9,0 

0.25-0,27  3,12-3,26 

0,24  -  - 

0,26-0,28  3.2  9.0 

0.25  2.78-2,92 

0.21-0.22  3.2  7.0 

0.25-0.26  - 

0.25-0.28  - 

0-25        -  2.6-2.8 

-  6.3 


— 

Padoa,  1956e 

17.0 

Padoa,  1956e 

— 

Uchida  etal.,  1958 

— 

Padoa,  1956e 

17.0 

Padoa.  1956e 

— 

Padoa,  1956e 

— 

Mho,  1963 

19.0 

Padoa,  1956e 

— 

Mito,  1963 

— 

Mito,  1963 

8.6 

Fritzsche,  1978;  Fahay 

1983 

8.7 

Fahay,  1983 

1.7 

1.9-2.2 
1.82 


<11.5 


NA 
NA 


5-6 


15.0  Padoa,  1956e;  Breder  and 

Rosen,  1966 

-  Robertson,  1974,  1975a 

>12.4        Brownell,  1979;  Gilchrist, 
1904;  Gilchrist  and 
Hunter,  1919;  Robert- 
son, 1975a 


p? 

1.7-1.8 

0 

NA 

— 

— 

— 

Gilchrist,  1904 

p 

0.88-1.2 

1 

0.19-0.25 

1.78-2.3 
-2.1 

7.3 
3.9-5.2 

13 

Ueno  and  Fujita,  1958; 

Change!  al..  1980 
Uisand  Rennis,  1983 

408 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  107.    Continued. 


Family/subfamily 
species 


Type  of  egg 
pelagic  (P), 

demersal 
(D)orvivip-         Egg  size 

arous  (V)  (mm) 


Number  Largest  oil 

of  oil  globule  size 

globules  (mm) 


Body  length  (mm)  at 


Hatching 


Transfor- 
mation 


Hoplichthyidae 

Hoplichthys  haswellP 
Hoplichlhys  sp.^ 

Dactylopteridae 

Daclylopterus  volitans 

Daicocus  petersoni 
Dactyloptena  sp. 


0.85-0.90     1 


-0.8 


0.15 


0.14 


e8 


Robertson,  1975a 
Okiyama  (unpubl.  MS) 


<7 

ca. 

16 

Fntzsche,  1978;  Sanzo, 
1933c;  Padoa,  1956e 

4.3 

ca. 

10 

Senta,  1958 

3.9-6.5 

ca. 

10 

Leisand  Rennis,  1983 

COTTOIDEI 

Agonidae 

Agonomalus  mozinoi 

D 

-1.0 

— 

— 

5.5 

Agonopsis  chiloensis^ 

— 

— 

— 

— 

— 

Agonus  cataphractus 

D 

1.7-2.2 

Several 
co- 
alesce 

0.7-0.75 

6.3-8.0 

A.  decagonus-' 

— 

— 

— 

— 

— 

Aspidophowides 

— 

— 

— 

— 

— 

monopterigius^ 

A.  olriki' 

— 

— 

— 

— 

— 

Bolhragonus  swani 

D 

2 

— 

— 

7.5 

Pallasina  barbata- 

— 

— 

— 

— 

— 

Xeneretmus  latifrons' 

- 

- 

— 

— 

~7 

Anoplopomatidae 

A  noplopoma  fimbria 


Blepsias  cinhosus-  — 

Chitonotus  pugetensis^  D 

Clinocottus  acuticeps  D 

C.  analis^  D 

C.  embryum  — 

C.  globiceps  D 

C.  recalvus  D 

Cottus  asper  D 

C  bairdi  D 

C.  carolinae  D 

C.  cognatus  D 

C  nozawae  D 

C.  reinii^  D 


2.0-2.1         - 


Comephoridae 

Comephoms  baicalensis 
C.  dybowskii 

V 

V 

N/A 
N/A 

Cottidae 

Artedius  creaseri 
A.  fenestralis 

- 

: 

A.  harringtoni 
A.  lateralis 

D 

1.07 

A.  meanyi 
Ascelichthys  rhodorus 

D 

1.7- 

0.22 


1.02-1.05     1  large  0.3 

5-8  small 

1.0-1.2         -  - 

1.2-1.3         several         0.18 
large 

1.5-2.0         -  - 

1.25-1.35     - 


1-3 
2.6-3.3 


-10-12 


-10 


11-14 


—  Marliave,  1978 

—  de  Ciechomski,  1981 

—  14  mm      Russell,  1976;  Ehren- 

baum,  1904;  Mcintosh 
and  Pnnce,  1890 

—  Ehrenbaum,  1905-1909 

—  Dannevig,  1919;  Bigelow 

and  Schroeder,  1953 

—  Dunbar,  1947 
>16  Marliave,  1975 

—  Marliave,  1975 

—  Marliave,  1975 

>33  Ahlstrom  and  Stevens, 

1976;  Hart,  1973;  Ko- 
bayashi,  1957 


3.1-3.5 
2.0-2.6 


9.4 

— 

>48 

Chemyayev,  1975 

8.2 

-13 

-21 

Chemyayev,  1971 

-3.5 

5.7-7.9 

13-14 

Washington,  1981 

3.5-3.8 

5.9-6.8 

12-13 

Washmgton.  1981;  Mar- 
liave, 1975 

-3.0 

5.2-6.4 

12-14 

Washmgton,  1981 

3.9-4.5 

5.0-6.3 

9.5-10.5 

Washmgton,  1981;  Mar- 
liave, 1975;  Budd, 
1940 

-3 

6.3-9.4 

15-20 

Washington,  1981 

6.0 

8.8-9.0 

12-15 

Malarese  and  Marliave. 
1982 

— 

Sll 

— 

Marliave.  1975;  Richard- 
son, 1981a 

2.9-3.0 

" 

>I6 

Misitano,  1980;  Richard- 
son and  Washington, 
1980 

3.1-3.3 

5.5-7.3 

12.6-15.0 

Washington,  1981;  Wash- 
mgton, pers.  obs. 

4.2-4.5 

— 

— 

Budd.  1940;  Washington, 
1981 

-4.0 

6.4-9.6 

13-14 

Washington.  1981 

5.1-5.4 

6.2-8.1 

12.9-13.5 

Washington.  1981 

4.6-4.7 

_ 

9-11 

Moms.  1951 

5.5-6.3 

-7.0 

— 

Stein,  1972;  Richardson 
and  Washmgton,  1980 

6.3-6.9 

— 

9-10 

Heufelder,  1982 

6.86 

— 

9.5-10 

Wallus  and  Granneman, 
1979 

5.7-6.3 

— 

8-11 

Wallus  and  Granneman. 
1979 

10.5 

_ 

_ 

Watanabe,  1976 

— 

— 

— 

Watanabe,  1976 

WASHINGTON  ET  AL.:  SCORPAENIFORMES 


409 


Table  107.    Continued. 


Family ''subfamily 
species 


Type  of  egg 
pelagic  (P). 
demersal 
(D)  or  vivip- 
arous (V) 


Number  Largest  oil 

Egg  size  of  oil  globule  size 

(mm)  globules  (mm) 


Body  length  (mm)  at 

Transfor- 

Hatching 

Flexion                mation 

References 

7-8 

Heufelder.  1982 

4.9-5.2 

5.2-7.0         7.6- 

-7.8 

Misitano,  1978;  Richard- 
son and  Washington. 
1980 

5.5-5.8 

12 

Russell,  1976 

-4.0 

_                   _ 

Russell,  1976 

5.6-6.1 

—                   — 

Kyushin,  1970 

C,  riceP 
Enophrys  bison 


Rhamphocottus  richard- 
soni 


Tnglops  murrayP 
T.  pingelP 

Cottocomephoridae 

Abyssocottus  bergianus' 
A.  godlewskiP 
A.  korolnejjp 


D 
D 


E.  bubalis  ( Taurulus)  D 

E.  HUjeborgi  ( TaurulusY  D 

Gymnocanthus  herzen-  D 

steini 

G.  venlralis^  — 

Hemileptdotus  gitbertP  — 

H.  hemilepidons  D 

H.  jordani  — 

H.  spinosus  — 

H.  zapus  — 

Hemilriplems  americanus  D 

H.  villosus  D 

Iceliis  btcornis^  — 

Leptocottus  armatus'  D 

Myoxocephalus  aenaeus  D 

M.  octodecimspinosus  D 

M.  guadricornus  D 

M.  scorpius  D 


M.  thompsonP  — 

Naulichlhys  oculofascia-  D 
tus 

Oligocottus  maculosus  D 


O.  snyderi 
Onhonopias  triads' 
Paricehnus  hoplilicus^ 


Pseudobiennius  cotloides^        D 
Radulinus  asprellus  — 

R.  boleoides^  — 


D 


Scorpaenichlhys  marmor-       D 
atus 


D 


1.7-1.8 


1.5-1.8         several 
2.0  1 

1.6-1.7         few 


1.5-1.6         1 


2-2.5 


0.36 


0.38 


0.31-0.56     -5-6 


-5 


-9.1 


1  large 


0.8 


D 

1.3-1.5 

1  large 
many 
small 

D 

1.2-1.3 

- 

D 

0.9-1.0 

1  large 

2  small 

D 

3.2-3.3' 

— 

D 

2.8-3.0' 

_ 

D 

-4.5' 

_ 

> 19-23 


7.6-10.1  19 

-  7-12  - 

10-14  -14.5  >18.( 

10.9-11.6     S14.4  -20 


1.4-1.5 

present 

— 

3.9-4.8 

-8 

15-20 

1.5-1.7 

2-1- 

0.2 

4.7-6.3 

6.8 

- 

1.9-2.3 

1  or  more 

diameter 
varies 

6.3-7.3 

9-11 

-15 

1.5-2.2' 

_ 

_ 

-8 

ca.  10.5-1- 

_ 

1.8-2.5 

several 

0.4-0.5 

7.4-8.6 

9-15 

17-20 

8-10  -  - 

9  -9-11  -26 


4.2-4.5         7.2-7.6         7.5-10 


4.47 

6.2-8.4 

11-13 

2.9-3.8 

- 

- 

<5.6 

- 

-25 

12.5 





£4.7 

7.2-10.9 

al5 

- 

-8.7 

- 

6-7 

8.4 

-14-15 

2.0-2.2 


2.5-2.8 


1.4-1.9  1  large  0.27  5.8-6.0         7.5-8.7  14-15- 

several 

small 
-  -  -  7-8  12  - 

2.0  manv  _  _  _  _ 


Ehrenbaum,  1905-1909 

Hattori,  1964 

Richardson  and  Washing- 
ton, 1980 

Gorbunova,  1964a'' 

Richardson  and  Washing- 
ton, 1980 

Matarese  and  Vinter  (in 
prep.) 

Fahay,  1983;  Fuiman, 
1976 

Okiyama  and  Sando, 
1976 

Russell,  1976;  Ehren- 
baum, 1905-1909 

Richardson  and  Washing- 
ton, 1980;  Jones,  1962 

Fahay.  1983;  Lund  and 
Marcy,  1975 

Fahay,  1983;  Colton  and 
Marak,  1969 

Khan  and  Faber.  1974 

Russell.  1976;  Ehren- 
baum, 1905-1909; 
Mcintosh  and  Master- 
man,  1897 

Heufelder,  1982 

Richardson  and  Washing- 
ton, 1980;  Marliave, 
1975 

Washmgton,  1981;  Stein, 
1973 

Washington,  1981;  Stein, 

1972 
Bolin,  1941 

Richardson  and  Washing- 
ton, 1980 

Watanabe,  1976 

Richardson  and  Washing- 
ton, 1980 

Richardson  and  Washing- 
ton, 1980 

Richardson  and  Washing- 
ton. 1980;  Marliave, 
1975;  Blackburn,  1973 

Richardson  and  Washing- 
ton, 1980;  O'Connell, 
1953 

Fahay,  1983 

Bigelow  and  Schroeder, 
1953;  Rass,  1949 


Taliev,  1955 
Taliev,  1955 
Taliev,  1955 


410 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 

Table  107.    Continued. 


Type  of  egg 

pelagic  (P). 

Jody  length  (mm 

at 

demersal 

Number 

Largest  oil 

Family/subfamily 

(D)  or  vivip- 

Egg size 

of  oil 

globule  size 

Transfor- 

species 

arous  (V) 

(mm) 

globules 

(mm) 

Halching 

Flexion 

mation 

References 

A.  pallidus' 

D 

2.6-2.8' 

— 

_ 

-6 

— 

-16 

Taliev,  1955 

Asprocottus  gibbosus^ 

D 

3.3-3.4' 

— 

— 

— 

— 

— 

Taliev.  1955 

A.  herzensteiniP 

D 

3.0-3.2' 

— 

— 

— 

— 

a9 

Taliev,  1955 

A.  megalops' 

D 

3.5-3.7' 

— 

— 

— 

— 

— 

Taliev,  1955 

Batrachocottus  baicalensis 

D 

-3.0 

3-10 
small 

10.0 
post- 
flexion' 

N/A 

-16 

Chemyayev,  1981 

B.  multiradialus^ 

-4.0 

— 

— 

-6.0 

— 

— 

Taliev,  1955 

B.  nikolskii^ 

D 

2.9-3.1 

— 

— 

_ 

_ 

— 

Taliev,  1955 

B.  uschkanP 

— 

— 

— 

— 

— 

— 

_ 

Taliev,  1955 

Cottinella  boulengerp 

D 

2.8' 

— 

— 

— 

— 

— 

Taliev,  1955 

Cottocomephorus  gre- 

D 

1.2-1.8 

— 

— 

-6.8-7.0 

_ 

-19 

Taliev,  1955 

wingki 

C.  inermis^ 

D 

1.5-1.7 

— 

— 

— 

_ 

_ 

Taliev,  1955 

Paracottus  kessleri 

D 

1.0-1.45 

1  large 

0.3 

5.2-5.4 

-6.2 

-20 

Chemyayev,  1978 

P.  kneri 

D 

2.0-2.3 

_ 

_ 

6.8-7.1 

_ 

>10.8 

Taliev,  1955 

Procotlus  jeittelesP 

D 

2.5-3.3 

- 

- 

- 

- 

- 

Taliev,  1955 

Hexagrammidae 

Hexagrammos  agrammys 

D 

2.02-2.07 

many  co- 
alesce 
to  1 

8.15-8.61 

-11 

S 40-48 

Fukuhara,  1971 

H.  decagrammus 

D 

— 

— 

— 

ca.  8 

15-18 

-30" 

Kendall  and  Vinter,  1984 

H.  lagocephalus 

D 

2.0-2.6 

many 

-8-9 

12-15 

ca.  29 

Kendall  and  Vinter, 
1984;  Gorbunova, 
1964b  (as  H.  decagram- 
mas) 

H.  octogrammus 

D 

1.75-2.10 

many 

0.8 

6-7 

-12-15 

-30" 

Gorbunova,  1964b 

H.  otakii 

D 

2.3-2.7 

many 

— 

6.5-7.0 

-11 

— 

Gorbunova,  1964b;  Yusa, 
1960c 

H.  stelleri 

D 

— 

— 

— 

-7-9 

-12-15 

-30'' 

Kendall  and  Vinter,  1984 

Ophiodon  elongalus 

D 

2.9-3.2 

1 

— 

-9.0 

11-15 

-30- 

Kendall  and  Vinter,  1984 

Pleurogrammus  mono- 

D 

2.1-2.8 

many 

1.38-1.4 

10-11 

-14-19 

-30" 

Yusa,  1967;  Gorbunova, 

pterygius 

1964b 

Oxylebius  pictus 

D 

— 

— 

— 

4-5 

7-9 

-45 

Kendall  and  Vinter, 
1984;  DeMartini,  1976 

Zaniolepis  sp. 

— 

— 

— 

— 

-2.5 

~6 

?15 

Kendall  and  Vinter,  1 984 

Normanichthyidae 

Normanichthys  crockeri 

- 

- 

— 

— 

<4.4 

7-9 

>16 

Balbontin  and  Perez, 
1980 

Psychrolutidae 

Dasycoltus  seliger' 

— 

— 

— 

— 

— 

-10 

— 

Richardson,  1981a 

Gilbertidia  sigalutes- 

D 

2.3 

— 

— 

— 

-13-15 

-23 

Marliave,  1975 

Malacocottus  sp. 

— 

— 

— 

— 

— 

-7-9.8 

£24 

Richardson,  1981a 

Psychrolutes  paradoxus^ 

- 

- 

- 

- 

— 

-10.5 

-13-14 

Marliave,  1975 

'  Ovarian  or  newborn  larvae  of  30  species  of  Sebasres  not  listed  here  are  described  in  Efremenko  and  Lisovenko  (1970).  Westrheim  (1975),  and  Moser  el  al.  (1977). 

^  Incomplete  descnption  with  illustration. 

'  Rifle  ovarian  egg  diameter. 

•  Pelagic  juvenile  stage. 

'  Hatch  at  advanced  postflexion  stage. 

'  Confusion  exists  regarding  correct  identification  [Matarese  and  Vinter  (in  prep.)]. 


and  Moser  et  al.  (1977)  on  scorpaenids  and  Sparta  (1956b)  and 
Richards  (in  prep.)  on  triglids  and  peristediids. 

Scorpaenidae  (Figs.  220-223).— This  is  the  largest  and  most 
diverse  scorpaenoid  family  with  about  44  genera  and  more  than 
350  species.  The  classification  and  relationships  of  the  family 
are  in  controversy  (Washington  et  al.,  this  volume)  and  we 
follow  their  subfamily  groupings. 

Sebastinae.  —  Barsukov  (1981)  includes  3  genera  and  1 1 4  species 
in  this  temperate  and  boreal  group.  Sebastes  with  about  106 


species  accounts  for  almost  '/3  of  the  species  in  the  order.  At 
least  a  single  larval  stage  is  known  for  62  species  of  Sebastes 
and  flexion  or  postflexion  stages  have  been  described  for  about 
32  of  these  (Table  107).  Larval  stages  have  been  described  for 
one  of  the  6  species  oi  Helicolenus  and  are  unknown  for  the  two 
species  of  Hozukius. 

In  Sebastes  most  of  the  yolk  is  utilized  before  hatching  while 
the  eggs  lie  freely  within  the  ovary.  Hatching  precedes  extrusion 
and  newborn  larvae  range  from  3.8  to  7.5  mm  in  length  among 
the  various  species  and  have  functional  eyes,  jaws,  and  pectoral 
fins.  The  finfold  is  slightly  inflated  and  has  minute  cell-like 


WASHINGTON  ET  AL.:  SCORPAENIFORMES 


411 


LIO 


APO-4^ 

pp6-4 


PPO-3 


—  LOP 


B 


TM 


UIO-4 


PPO-5 


LIO-I 


LIO-2 


— PPO-3 


PPO-5 


APO-4 


PPO-4 


Fig.  220.  Head  spines  in  6.2  mm  (A),  8.2  mm  (B).  10.0  mm  (C)  and  16.0  mm  (D)  stained  larvae  of  Sehasles  melanostomus.  Abbreviations 
of  head  spines:  APO-2,  2nd  anterior  preopercular;  APO-3,  3rd  anterior  preopercular;  APO-4,  4th  anterior  preopercular;  CL,  cleithral;  lOP, 
interopercular;  LIO-1,  1st  lower  infraorbital;  LlO-2,  2nd  lower  infraorbital;  LOP,  lower  opercular;  LPST,  lower  posttemporal;  NA,  nasal;  NU, 
nuchal;  PA,  parietal;  PPO-1,  1st  posterior  preopercular;  PPO-2.  2nd  posterior  preopercular;  PPO-3,  3rd  posterior  preopercular;  PPO-4,  4th 
posterior  preopercular;  PPO-5,  5th  posterior  preopercular;  PRO,  preocular;  PSO,  postocular;  PT,  pterotic;  SC,  supracleithral;  SPO,  supraocular; 
TM,  tympanic;  UIO-1,  1st  upper  infraorbital;  UIO-2,  2nd  upper  infraorbital;  UIO-3.  3rd  upper  infraorbital;  UIO-4,  4th  upper  infraorbital;  UOP, 
upper  opercular;  UPST,  upper  posttemporal.  From  Moser  and  Ahlstrom,  1978. 


Structures  concentrated  along  the  dorsal  and  ventral  margins. 
Notochord  flexion  occurs  at  about  6-12  mm  and  transformation 
at  15-25  mm  (Table  107).  Many  species  have  a  distinct  pelagic 
juvenile  stage  which  can  reach  almost  60  mm  body  length. 

Preflexion  larvae  have  a  slender  body  (body  depth  1 3-23% 
of  body  length)  and  compact  gut;  snout-anus  distance  increases 
from  about  40-50%  of  body  length  to  over  60%  in  some  species 
during  the  larval  period.  The  caudal  and  pectoral  fins  begin 
forming  first,  followed  by  the  pelvics  and  then  the  dorsal  and 
anal  fins.  The  pectoral  fins  range  from  short  and  rounded  to 
elongate  and  fan-shaped,  reaching  almost  50%  of  body  length 
in  5.  levis  (Fig.  221).  The  pectoral  fin  base  is  shallow  (typically 
7-13%  of  body  length)  in  comparison  with  other  subfamilies. 
Ossification  of  skeletal  elements  begins  early  in  the  larval  period 


and  proceeds  rapidly  as  in  other  scorpaenoids;  vertebral  ossi- 
fication follows  the  pattern  of  other  scorpaeniforms,  with  the 
neural  arches  ossifying  before  the  centra  (Moser,  1972). 

Pigmentation  in  newborn  larvae  consists  of  a  melanistic  sheath 
over  the  gut  and  a  postanal  series  along  the  ventral  midline. 
Some  species  also  have  a  dorsal  midline  series  which  may  de- 
velop gradually.  Pigment  increases  with  development,  appear- 
ing on  the  head  (above  brain,  on  jaws  and  opercular  region), 
fins,  and  caudal  peduncle.  Often  the  pectoral  fins  (both  base  and 
blade)  have  diagnostic  pigment  patterns.  Several  of  the  western 
Pacific  species  are  heavily  pigmented  with  the  head  and  body 
covered  by  a  sheath  of  melanophores  (Fig.  221). 

Head  spines  are  a  prominent  feature  of  all  Sebastes  larvae. 
Pterotics,  parietals  (usually  serrated),  and  preopercular  spines 


412 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  221.  Larvae  of  Scorpaenidae.  (A)  Sehastes  oblongiis.  8.5  mm  TL  (from  Fujita,  1958);  (B)  5.  longispinis.  7.1  mm  TL  (from  Takai  and 
Fukunaga,  1971);  (C)  S.  huhbsi.  6.0  mm  TL  (from  Uchida  et  al.,  1958);  (D)  S.  zacentrus.  12.7  mm  SL  (from  Laroche  and  Richardson.  1981);  (E) 
5.  paucispims.  10.5  mm  SL  (from  Moser  et  al.,  1977);  (F)  5.  jordani.  15.5  mm  SL  (ibid.);  (G)  S.  levis.  10.4  mm  SL  (ibid.);  (H)  Hetwolenus 
daclyloplerus,  10.0  mm  (from  Tuning,  1961). 


form  during  the  preflexion  period  in  most  species,  and  other 
spines  appear  gradually  thereafter  (Fig.  220).  Although  there  is 
variation  in  larval  spine  complements  (Moser  and  Ahlstrom, 
1978;  Moser  and  Butler,  1981;  Richardson  and  Laroche,  1979; 
Laroche  and  Richardson.  1980,  1981).  it  is  apparent  that  1) 
the  adult  head  spine  complement  develops  during  the  larval 
period  and  2)  certain  spines  develop  during  the  larval  period 
but  are  not  present  in  adults.  Of  the  latter,  the  most  prominent 
are  the  pterotic,  anterior  preoperculars,  lower  posttemporal,  and 
upper  infraorbitals.'  The  fact  that  these  spines  do  occur  in  adults 
of  other  subfamilies  is  of  possible  phylogenetic  significance 
(Moser  and  Ahlstrom,  1978). 


Upper  infraorbitals  are  present  in  adults  of  a  few  species  ofSebastes. 


Helicolenus  is  viviparous,  the  fertilized  eggs  developing  in  a 
gelatinous  matrix  within  the  ovary  (Graham,  1939;  Krefft,  1961). 
Larvae  of  H.  dactyloptenis  have  been  described;  hatching  and 
birth  occur  at  a  smaller  size  (2.2  mm)  than  in  Sebastes.  although 
sizes  at  notochord  flexion  and  transformation  are  similar  (Table 
1 07).  Larvae  are  moderately  deep-bodied  (Fig.  221);  body  depth 
averages  29%,  33%,  and  49%  of  body  length  for  preflexion, 
flexion  and  postflexion  stages.  Head  and  gut  shape  are  similar 
to  that  of  Sebastes.  The  pectoral  fin  is  moderate  in  size  and 
rounded;  the  base  is  slightly  deeper  than  in  most  species  of 
Sebastes.  Sequence  of  fin  formation  is  similar  to  that  oi Sebastes. 
A  mass  of  spongy  tissue  develops  anteriorly  in  the  dorsal  finfold 
in  preflexion  larvae  and  persists  through  most  of  the  larval 
period;  the  structure  is  apparently  unique.  The  early  pigment 
pattern  consists  of  a  dorsolateral  gut  sheath,  melanophores  above 
the  brain,  on  the  lower  jaw,  in  a  short  median  ventral  series  just 


WASHINGTON  ET  AL.:  SCORPAENIFORMES 


413 


Fig.  222.     Larvae  of  Scorpaenidae.  (A)  Ponlmus  Type  A.  8.0  mm  SL  (from  Moser  et  al.,  1977);  (B)  Scorpaena  Type  A,  8.0  mm  SL  (ibid.);  (C) 
Scorpaenodes  xyris,  6.2  mm  SL  (ibid.);  (D)  Sebastotobus  sp.  7.7  mm  SL  (ibid.);  (E)  Ectreprosebasies  imus.  (s.l  mm  SL  (ibid.). 


anterior  to  the  caudal  fin,  and  on  the  distal  and  proximal  regions 
of  the  pectoral  fin  blade  (Fig.  221).  Head  spine  formation  is 
similar  to  that  of  Sebastes  species  which  have  full  larval  com- 
plements, except  that  spines  are  lacking  on  the  2nd  infraorbital 
bone  and  the  cleithrum. 

Scorpaeninae.  — Larval  stages  are  known  for  only  3  of  the  15 
genera  in  this  subfamily;  a  total  of  8  species  (or  generic  types) 


out  of  about  150  have  been  described  (Table  107;  see  Sparta, 
1956b  and  Moser  et  al.,  1977,  for  major  reviews).  Hatching 
occurs  at  about  2.0  mm  or  less;  newly-hatched  larvae  have  a 
large  elliptical  yolk  sac,  unpigmented  eyes,  pectoral  fin  buds, 
and  lack  a  mouth.  The  finfold  is  inflated  and,  along  with  the 
body  skin,  forms  a  balloon-like  envelope  that  is  attached  prin- 
cipally at  the  snout  and  pectoral  regions  (Orton,  1955d).  Cell- 
like granulations  cover  the  entire  envelope  but  are  concentrated 


414 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  223.  Larvae  of  Minoinae  (A).  Triglidae  (B,  C),  Congiopodidae  (D,  E),  Platycephalidae  (F),  Hoplichthyidae  (G,  H).  (A)  Minous  sp.?,  6.4 
mm  SL  (from  Leis  and  Rennis,  1983);  (B)  Prionolus  sp.,  6.4  mm  SL  (original);  (C)  Prionotus  slephanophrys.  8.8  mm  SL  (CalCOFI  7510  sla. 
1 1 7.70);  (D)  Congiopodus  spimfer.  1 0.8  mm  SL  (from  Brownell,  1 979);  (E)  Detail  of  pectoral  fin  of  Congiopodus  spmifer  (ibid.);  (F)  Platycephalidae, 
unidentified,  6.2  mm  SL  (from  Leis  and  Rennis,  1983);  (G)  Hoplichthys  sp.,  7.1  mm  SL  (original,  courtesy  M.  Okiyama);  (H)  Hoplichlhys  sp. 
17.2  mm  SL  (ibid.). 


at  the  median  edges  of  the  finfold.  Flexion  occurs  at  a  small  size 
(4-6  mm)  as  does  transformation  (10-17  mm).  Larvae  are  rel- 
atively deep-bodied  during  preflexion  and  flexion  and  more  so 
during  postflexion,  when  body  depth  averages  38-40%  of  body 
length  for  the  genera  listed  in  Table  107.  The  gut  is  compact 
and  the  head  becomes  massive.  Snout-anus  length  increases 
from  46-50%  of  body  length  in  preflexion  larvae  to  61-67%  in 
postflexion  larvae.  The  snout  has  a  steep  profile  (Fig.  222). 

The  pectoral  fins  are  well  developed  and  deep-based;  fin  base 
depth  is  13-15%  of  body  length  in  preflexion  larvae  and  14- 
1 8%  in  flexion  and  postflexion  larvae.  They  are  fan-shaped  and 
enlarged  in  Scorpaenodes;  fin  length  attains  41%  of  body  length 
during  the  postflexion  stage.  They  are  smaller  but  distinctively 
shaped  in  Scorpaena  (fan-shaped  with  scalloped  margin)  and 
Pontinus  (slightly  wing-shaped).  Ossification  of  fin  rays,  as  well 


as  skeletal  elements,  occurs  in  early  larvae  (4-5  mm).  The  cau- 
dal, pectoral,  and  pelvic  rays  begin  ossifying  almost  simulta- 
neously, followed  immediately  by  the  dorsal  and  anal  fins. 

Preflexion  larvae  have  a  postanal  ventral  midline  series  of 
melanophores  ranging  in  number  from  2-7  in  Scorpaena  guttata 
to  12-18  in  Scorpaenodes  xyris.  The  most  prominent  pigment 
is  on  the  pectoral  fins;  typical  patterns  are  a  concentration  at 
the  distal  margin  (Scorpaenodes,  some  Pontinus,  some  Scor- 
paena spp.),  a  solid  covering  over  most  of  the  fin  (some  Scor- 
paena spp.),  or  a  diagonal  bar  (some  Pontinus  spp.).  A  melanistic 
sheath  develops  over  the  dorsal  surface  of  the  gut  and  gas  bladder 
in  most  species  of  Scorpaena.  whereas  in  Scorpaenodes  and 
Pontinus  only  the  gas  bladder  is  pigmented.  Other  pigment  in 
Scorpaena  forms  at  the  cleithral  juncture  and  above  the  brain 
(Fig.  222). 


Fig.  224.  Larvae  of  the  Oxylebius  scorpaeniform  group  (A,  B)  and  the  hexagrammid  group  (C-F)  of  Washington  and  Richardson  (MS)  (see 
Washington  et  al.,  this  volume).  (A)  Oxylebius  pictus.  8.5  mm  SL  (from  Kendall  and  Vinter,  1984);  (B)  Zaniolepis  sp.,  7.7  mm  SL  (ibid.);  (C) 
Hexagrammos  oclogrammus,  15.2  mm  SL  (ibid.);  (D)  Pleurogrammus  monopterygius,  20.5  mm  SL  (ibid.);  (E)  Ophwdon  elongalus,  1 5.4  mm  SL 
(ibid.);  (F)  Anoplopoma  fimbria,  13.8  mm  SL  (Ahlstrom  and  Stevens.  1976). 


WASHINGTON  ET  AL.:  SCORPAENIFORMES 


415 


416 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


B 


:^\V\\\v'^\v\\\\\  \V\\  \  \ 


Fig.  225.  Larvae  of  Normanichthyidae  (A),  Cottocomephoridae  (B,  C),  Comephoridae  (D,  E).  (A)  Normanichlhys  crockeri.  8.5  mm  SL  (original); 
(B)  Coltocomephonis  grewingki.  7.4  mm  (from  Taliev,  1955);  (C)  Cottocomephorus  inenrus.  1 1.2  mm  (ibid.);  (D)  Comephorus  baicalensis.  6.9 
mm  (ibid.);  (E)  Comephorus  baicalensis.  21.3  mm  (ibid.). 


Cranial  spine  development  is  similar  to  that  in  sebastines. 
The  pterotic,  parietal,  postocular  (supraocular  crest),  posterior 
preoperculars  (2nd,  3rd,  and  4th)  anterior  preoperculars  (2nd 
and  4th)  and  lower  posttemporal  develop  during  the  prefiexion 
period.  The  lower  infraorbital  ( 1  st),  upper  infraorbitals  ( 1  st  and 
4th),  posterior  preoperculars  (1st  and  5th),  nuchal,  supraclei- 
thral,  cleithral,  upper  opercular,  and  lower  opercular  spines  ap- 
pear during  postflexion.  Late  in  the  postflexion  stage  the  lower 
infraorbital  (2nd),  nasal,  preocular,  and  supraocular  spines  ap- 
pear. Spines  which  do  not  develop  in  scorpaenine  larvae  but 
are  present  in  adults  of  most  genera  are  the  upper  infraorbitals 
(2nd  and  3rd),  upper  posttemporal,  tympanic,  and  sphenotic. 
In  Scorpaenodes  the  nuchal  spine  develops  during  the  prefiexion 
period  and  exceeds  the  parietal  spine  in  length,  giving  the  pa- 
rietal ridge  a  bifurcate  appearance.  In  other  scorpaenines  and 
all  other  scorpaenids  except  Sebastolobus.  the  nuchal  develops 
late  and  is  excluded  from  the  parietal  ridge. 


Pteroinae.— Early  prefiexion  larvae  have  been  described  for 
Pterois  lunulata  and  Dendrochirus  brachypterus  (Table  107). 
Newly-hatched  larvae  are  small  (1.1-1.6  mm)  and  similar  in 
morphology  to  those  of  Scorpaeninae.  The  pectoral  fins  are  large 
and  fan-shaped  with  pigment  at  the  distal  margin.  Postanal 
pigment  in  Pterois  consists  of  ventral  and  dorsal  midline  series. 
In  Dendrochirus  this  pigment  coalesces  to  form  a  band. 

Sebastolobinae.  — Life  history  series  have  been  described  for  Se- 
bastolobus alascanus  and  S.  altivelis  (Moser,  1974).  Larvae  are 
2.6  mm  at  hatching,  6.0-7.3  mm  at  notochord  flexion,  and  14- 
20  mm  at  transformation.  The  distinctive  pelagic  juveniles  (up 
to  56  mm  in  S.  altivelis)  have  a  prolonged  midwater  existence 
before  settling  to  the  deep  shelf  and  slope  habitat  of  the  adults. 
Larval  morphology  is  similar  to  that  of  scorpaenines.  The  pec- 
toral fins  are  large,  deep-based,  and  fan-shaped  (Fig.  222);  their 
rays  are  the  first  to  ossify,  followed  by  the  caudal  rays  and  then 


WASHINGTON  ET  AL.:  SCORPAENIFORMES 


417 


■r,i',h:  •. 


>j«-.i; 


dft^/ 


Fig.  226.  Larvae  of  the  Rhamphocottus  group  (A)  and  the  Scorpaenichthys  group  (B.  C)  of  cottids  of  Washington  and  Richardson  (MS)  (see 
Washington  et  al.,  this  volume).  (A)  Rhamphocottus  nchardsoni.  10.6  mm  SL  (from  Richardson  and  Washington,  1980);  (B)  Scorpaenichthys 
inarmoratus.  8.7  mm  SL  (ibid.);  (C)  Hemilepidotus  spinosus,  1 1.0  mm  SL  (ibid.). 


those  of  the  other  fins.  The  pectoral  fins  are  pigmented  at  the 
distal  margin;  other  pigment  includes  a  sheath  over  the  gut  and 
melanophores  above  the  brain.  Head  spination  is  highly  de- 
veloped (Fig.  222);  the  sequence  of  development  is  similar  to 
that  of  scorpaenines.  In  addition  to  the  spine  complement  of 
scorpaenines,  Sebastolobus  larvae  develop  the  2nd  and  3rd  up- 
per infraorbital  spines  and  the  1  st  anterior  preopercular  spine. 

Setarchinae.  —  Larvae  are  known  for  Ectreposebastes  imus  (Moser 
et  al.,  1977).  Hatching  and  notochord  flexion  occur  at  a  small 
size  as  in  the  scorpaenines;  however,  postflexion  larvae  attain 
a  large  size  (Table  107).  Larvae  have  the  deepest  body  of  known 


scorpaenids;  body  depth  reaches  55%  of  body  length  in  late 
postflexion  stage.  The  gut  is  compact  with  an  elongate  terminal 
section;  snout-anus  distance  averages  53%  of  body  length  in 
preflexion  larvae  and  76%  in  postflexion.  The  pectoral  fins  are 
deep-based,  fan-shaped,  and  large,  extending  to  the  caudal  pe- 
duncle (Fig.  222).  Fin  base  depth  and  fin  length  reach  22%  and 
57%  of  the  body  length  respectively.  The  pigment  pattern  con- 
sists of  a  postanal  ventral  series  of  1 1-14  melanophores  (not 
present  after  4.0  mm),  a  blotch  above  the  gas  bladder,  and  an 
almost  solid  sheath  over  the  pectoral  fin,  which  recedes  distally 
with  development.  Head  spine  development  is  similar  to  that 
of  scorpaenines. 


4^:^^€:^^^^^&>^ 


WASHINGTON  ET  AL.:  SCORPAENIFORMES 


419 


Fig.  227.  Larvae  of  the  Myoxocephalus  group  of  cottids  of  Washington  and  Richardson  (MS)  (see  Washington  et  al.,  this  volume).  (A) 
Paricelinus  hopliticus.  13.8  mm  SL  (from  Richardson  and  Washington,  1980);  (B)  Triglops  sp..  15.4  mm  SL  (ibid.);  (C)  Icelus  hicornis.  25  mm 
(from  Ehrenbaum,  1905-1909);  (D)  Chitonotus  pugetensis.  11.5  mm  SL  (from  Richardson  and  Washington,  1980);  (E)  Artednis  meanyi.  13.8 
mm  SL  (ibid.,  as  Iceiinus  sp.);  (F)  Icelinus  sp.,  1 1.9  mm  SL  (original);  (G)  Ascelichthvs  rhodorus.  1 1.0  mm  SL  (from  Matarese  and  Marliave, 
1982). 


Choridactylinae.— The  developmental  stages  of  Ininucus  ja- 
ponicus  have  been  described  by  Fujita  and  Nakahara  (1955)  and 
Sha  et  al.  (1981).  Larvae  are  3.2  mm  at  hatching,  6.4-8.2  mm 
at  flexion  and  about  10  mm  at  transformation.  Yolk-sac  larvae 
are  similar  to  those  of  Scorpaeninae.  Larvae  are  relatively  slen- 
der and  blunt-headed,  with  a  compact  short  gut  (Fig.  223).  The 
pectoral  fins  are  large  and  fan-shaped,  with  a  scalloped  margin; 
they  develop  a  series  of  large  blotches  distally.  One  to  several 
large  postanal  melanistic  blotches  form  on  the  postanal  ventral 
midline  and  the  gas  bladder  region  is  pigmented.  Sha  et  al.  ( 1 98 1 ) 
show  the  larvae  to  be  heavily  xanthic. 

Minoinae.  — Leis  and  Rennis  (1983)  described  a  larval  series 
tentatively  identified  as  Minous  sp.  It-  is  generally  similar  in 
morphology  and  pigmentation  to  Jnimicus;  however,  the  pec- 
toral fin  is  relatively  larger  and  has  a  different  pigment  pattern. 

Triglidae  (Fig.  223).  — E^s  are  only  known  for  3  of  the  8  genera 
of  triglids.  The  new  world  genus  Prionotus  has  multiple  oil 
globules  whereas  single  oil  globules  are  known  for  Chelidonich- 
thys  and  Lepidotrigla.  Larvae  are  poorly  known  with  complete 
series  having  been  described  for  4  species  in  3  genera  (Table 
107).  There  are  approximately  90  species  in  this  family  and 
many  are  very  difficult  to  identify  as  adults.  The  genus  Lepi- 
dotrigla has  40+  species  and  is  poorly  known  in  many  areas. 
Diagnostic  features  include  the  depressed  profile  of  the  head  and 
large  pectoral  fins  of  which  the  lowest  three  rays  become  de- 
tached during  transformation.  Meristics  are  very  similar  to 
platycephalids  and  caution  is  advised.  However,  most  triglids 
have  fewer  pectoral  rays  than  most  scorpaenoids.  Prionotus. 
including  Bellator,  has  1 3  to  15  plus  3  free  rays;  Trigla,  Chel- 
idonichthys.  Lepidotrigla,  and  Uradia  have  11  plus  3  free  rays; 
and  Pterygotrigla  and  Parapterygotrigla  have  11  to  13  plus  3. 


Peristediidae.  —  ELH  information  has  been  published  only  for 
Peristedion  cataphractum  of  the  eastern  Atlantic  (Table  107). 
Larvae  and  transforming  juveniles  have  elongated  upper  pec- 
toral rays  and  strong  head  spination  (see  plate  40  in  Padoa, 
1956e).  This  family  is  often  combined  with  the  Triglidae,  but 
differs  in  many  characters  such  as  the  presence  of  barbels,  2 
rather  than  3  free  pectoral  rays,  and  the  body  is  encased  in  bony 
scutes  rather  than  scales.  Three  genera  (Heminodus.  Parahem- 
inodus  and  Gargariscus)  have  jaw  teeth  and  two  genera  (Per- 
istedion and  Satyrichthys)  lack  jaw  teeth.  There  are  about  25 
species  found  in  the  tropics  of  all  oceans  in  deep  water  (>200 
m). 


Congiopodidae  (Fig.  223).  — Eggs  are  known  for  only  1  {Con- 
giopodus)  of  the  4  genera  of  Congiopodidae  (Brownell,  1978; 
Gilchrist,  1904;  Robertson,  1974).  The  pelagic  eggs  are  rela- 
tively large  (1.7-2.18  mm)  and  spherical,  with  a  narrow  peri- 
vitelline  space  and  no  oil  globules.  The  egg  surface  is  covered 
with  striations.  Early  life  history  stages  have  been  illustrated  for 
one  species,  Congiopodus  spimfer{Qvov<mt\\,  1979;  Gilchrist  and 
Hunter,  1919).  Robertson  (1975a),  illustrated  a  well-developed 
embryo  of  C.  leucopaecilus.  Larvae  hatch  at  about  5  to  6  mm 
NL  and  are  elongate  with  long  guts  reaching  50%  SL.  The  pec- 
toral fins  are  extremely  large  and  fan-shaped.  Melanistic  pig- 
ment is  present  on  the  head,  nape  and  on  the  dorsal  and  ventral 
surface  of  the  gut.  Two  large  blotches  of  pigment  on  the  dorsal 
and  ventral  midlines  form  a  band  midway  between  the  vent 
and  tail  tip.  The  large  pectoral  fins  have  a  distal  band  of  pigment 
which  gradually  expands  over  the  entire  fin  with  development. 
Larvae  develop  large  postocular  and  parietal  spines.  The  pres- 
ence of  preopercular  spines  can  not  be  determined  from  the 
description  by  Brownell  (1975). 


420 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


\ 


-i«75f*: 


WASHINGTON  ET  AL.:  SCORPAENIFORMES 


421 


Platycephalidae  (Fig.  22ij.  —  Platycephalids  spawn  small  spher- 
ical eggs  (<  1  mm)  with  a  single  oil  globule  (Chang  et  al..  1980; 
Uchida  et  al.,  1958).  Larvae  have  been  described  and  illustrated 
for  Platycephalus  indicus  (Ueno  and  Fujita,  1958)  and  for  a 
series  of  larvae  incorporating  seven  unidentified  species  (Leis 
and  Rennis,  1983).  Newly-hatched  platycephalids  are  relatively 
small  (1.7-2. 3  mm)  and  slender-bodied,  with  unformed  mouths, 
unpigmented  eyes,  and  large  yolk  sacs.  By  the  time  of  yolk 
absorption  larvae  have  large  heads  and  deep  bodies  which  taper 
toward  the  tail.  The  gut  is  quite  long  reaching  %  SL  during 
development.  The  pointed  snout  becomes  distinctively  long  and 
flattened.  Pigmentation  is  usually  present  on  the  head,  jaws, 
ventral  surface  of  the  gut  and  along  the  postanal  ventral  midline. 
Pigment  may  also  be  present  on  the  dorsolateral  surface  of  the 
tail  and  pectoral  fin.  Larvae  develop  4  to  9  preopercular  spines. 
Other  head  spines  include:  supraocular,  supracleithral,  parietal 
and  pterotic.  Unlike  most  other  scorpaeniforms,  head  spines 
persist  and  become  more  pronounced  in  juveniles.  Fin  devel- 
opment proceeds  as  follows:  pectoral,  caudal,  dorsal,  anal  and 
pelvic. 

Hoplichthyidae  (Fig.  22ij.  — The  pelagic  eggs  of  Hoplichthys 
haswelli  are  described  by  Robertson  (1975a)  as  small  and  spher- 
ical with  a  smooth  surface.  A  single  oil  globule  is  present.  De- 
scriptions of  hoplichthyid  larvae  have  not  been  published;  how- 
ever, based  on  Okiyama  (in  prep.)  larvae  are  quite  similar 
to  platycephalids.  Preflexion  larvae  (3.2  mm)  are  elongate  with 
large  heads  and  pointed  snouts.  The  gut  is  moderately  long 
(>50%  SL)  and  the  early-developing  pectoral  fins  are  large  and 
fan-shaped.  The  snout  becomes  increasingly  long  and  depressed 
during  development.  Pigmentation  is  limited  to  the  gut,  distal 
tip  of  the  pectoral  fin  and  a  band  on  the  ventral  finfold  midway 
between  the  vent  and  notochord  tip.  Numerous  clusters  of  small 
spines  develop  in  the  supraocular,  parietal  and  pterotic  regions. 
Seven  spines  form  on  the  posterior  margin  of  the  preopercle 
with  smaller  spines  at  their  base.  As  in  platycephalids,  head 
spines  persist  in  juveniles. 

Dadylopteridae  (Fig.  233).— The  pelagic  eggs  are  small  (0.8  mm) 
and  slightly  ovoid  with  a  single  oil  globule.  The  egg  surface  is 
smooth  and  unsculptured.  Larvae  hatch  at  about  1.8  mm  and 
undergo  flexion  of  the  notochord  between  3.9-6.5  mm.  Trans- 
formation to  the  juvenile  form  occurs  at  about  9  mm.  Larvae 
are  moderately  deep-bodied  with  a  distinctively  blunt  snout  and 
small  mouth.  The  gut  is  long,  reaching  about  75%  SL  in  post- 
flexion  larvae.  Pigmentation  occurs  over  the  head,  gut,  along 
the  postanal  ventral  midline  and  around  notochord  tip.  Pig- 
mentation increases  dramatically  over  most  of  the  body  in  post- 
flexion  larvae.  The  distinctive  head  armature  is  quite  different 
from  all  known  scorpaeniform  larvae  and  is  present  in  larvae 
as  small  as  2.3  mm  NL.  A  small  supraoccipital  spine  is  present 
only  during  the  larval  period.  The  extremely  long  posttemporal 
and  preopercular  spine  extend  posteriorly  to  the  middle  of  the 
anal  fin  in  larvae  by  about  6.5  mm  and  persist  in  juveniles  and 
adults. 


COTTOIDEI 

Eggs 

Eggs  are  known  from  representatives  of  six  of  the  nine  cottoid 
families  recognized  here  (Table  107).  Where  known,  most  cot- 
toids  spawn  demersal,  adhesive  eggs  which  often  form  clusters 
found  under  rocks.  Eggs  are  frequently  brightly  colored,  e.g., 
red,  blue,  green,  yellow.  The  eggs  of  Anoplopoma  fimbria  are 
pelagic.  The  Comephoridae  of  Lake  Baikal  are  reported  to  be 
viviparous. 

Most  eggs  are  spherical  and  average  1-2  mm  in  diameter, 
although  eggs  as  large  as  4  mm  have  been  reported  in  the  cottid 
Hemitripterus  and  some  of  the  cottocomephorids.  A  single  large 
oil  globule,  frequently  accompanied  by  several  small  ones,  oc- 
curs in  many  species.  The  surface  of  the  eggs  is  often  covered 
by  a  tough  adhesive  membrane,  and  may  be  smooth  as  in  An- 
oplopoma  and  Myo.xocephalns  aenaeus  (Fahay,  1983)  or  cov- 
ered by  tiny,  radiating  canals  as  in  Arledius  lateralis  and  Cli- 
nocottus  analis  (Budd,  1940). 

Larvae 

At  least  one  larval  stage  is  known  for  88  of  the  329+  species 
and  for  46  of  the  104  genera  of  cottoids.  Major  overviews  of 
larval  cottoid  taxonomy  include:  Richardson  and  Washington 
(1980)  on  cottids;  Kendall  and  Vinter  (1984)  on  hexagrammids; 
Taliev  (1955)  and  Chemyayev  (1971,  1975,  1978,  1981)  on 
comephorids  and  cottocomephorids;  and,  forthcoming  Laroche 
(in  prep.)  on  agonids. 

Larval  cottoids  exhibit  a  broad  diversity  of  form.  Size  at 
hatching  varies  from  2  to  12  mm.  Planktonic  life  may  be  quite 
brief  several  weeks  in  many  cottids,  or  may  be  extended  up  to 
a  year  with  a  special  pelagic  juvenile  stage  as  in  the  hexagram- 
mids. 

Cottoid  larvae  exhibit  such  a  diversity  of  form  and  devel- 
opment that  it  is  impossible  to  characterize  a  generalized  "cot- 
toid" larva. 

Hexagramrnidae  (Fig.  224).  — Larvae  are  known  for  10  of  the 
1 1  species  of  the  hexagrammid  genera  Hexagrammos.  Pleuro- 
grammus,  and  Ophiodon.  Major  works  presenting  descriptions 
and  illustrations  include  Kendall  and  Vinter  (1984)  and  Gor- 
bunova  (1964b).  Hexagrammids  hatch  at  a  relatively  large  size 
(6-1 1  mm  NL).  Development  is  gradual  from  hatching  to  the 
juvenile  stage  with  a  prolonged  epipelagic  prejuvenile  period 
(~  30-50  mm  SL).  Larvae  have  elongate,  slender  bodies  with 
large  eyes.  Larval  Hexagrammos  and  Pleurogrammus  have  blunt 
heads,  while  Ophiodon  larvae  have  pointed  snouts  and  large 
terminal  mouths. 

Larvae  are  heavily  pigmented  especially  dorsally.  Melano- 
phores  are  scattered  over  the  head,  gut  and  usually  on  the  dorsal 
and  ventral  midlines.  The  extent  of  postanal,  ventral  midline 
and  lateral  pigmentation  is  useful  in  specific  identification. 

Fin  formation  proceeds  in  the  following  sequence:  caudal, 
pectoral,  second  dorsal  and  anal,  first  dorsal  and  pelvic.  Larvae 
exhibit  delayed  ossification.  Vertebral  ossification  in  hexagram- 


Fig.  228.  Larvae  of  the  Myoxocephalus  cottid  group  of  Washington  and  Richardson  (MS)  (see  Washington  et  al.,  this  volume).  (A)  Onhonopias 
triacis.  7.0  mm  SL  (original);  (B)  Enophrys  bison.  7.0  mm  SL  (from  Richardson  and  Washington,  1980);  (C)  Myo.xocephalus  aenaeus,  7.0  mm 
SL  (from  Lund  and  Marcy,  1975);  (D)  Myo.xocephalus  polyacanlhocephalus.  12.0  mm  SL  (from  Richardson,  1981a);  (E)  Radulinus  asprellus, 
10.9  mm  SL  (from  Richardson  and  Washington,  1980);  (F)  Gymnocanthus  tncuspis.  13.0  mm  (from  Khan,  1972). 


422 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


mids  (and  Anoplopoma)  is  similar  to  that  in  Scorpaenoidei  with 
the  neural  and  hemal  arches  ossifying  before  the  associated  ver- 
tebral centra.  Vertebral  counts  are  notably  high  (47-63).  Head 
spines  are  absent  in  larval  Hexagrammos  and  Pleurogrammus 
and  extremely  reduced  in  Ophiodon.  with  late-stage  larvae  de- 
veloping 4  tiny  preopercular  spines. 

Anoplopomatidae  (Fig.  224.  Table  ;07j. -Larvae  of  only  Ano- 
plopoma have  been  described  and  illustrated  by  Kobayashi  (1957) 
and  Ahlstrom  and  Stevens  (1976).  Early  development  oi  Ano- 
plopoma is  similar  to  that  of  the  hexagrammids.  Larvae  hatch 
at  a  large  size  (~  9  mm  NL)  and  development  is  gradual  without 
great  changes  in  form. 

Larvae  are  slender  and  elongate  with  pointed  snouts  and  long 
guts.  The  distinctive  pectoral  fins  with  heavy  distal  pigmentation 
are  exceptionally  large  reaching  nearly  33%  SL  late  in  the  larval 
period.  Larvae  are  heavily  pigmented  with  melanophores  over 
most  of  the  head,  gut  and  lateral  surface  of  the  body. 

As  in  hexagrammid  larvae,  ossification  is  delayed  with  the 
neural  and  hemal  arches  ossifying  before  the  associated  vertebral 
centra.  Vertebral  counts  (61-66)  are  distinctively  high.  Pectoral 
fin  development  is  precocious.  Head  and  preopercular  spines 
are  absent. 

Oxylebius-Zaniolepis  (Fig.  224).- Oxylebius and  Zaniolepis are 
sometimes  included  in  the  Hexagrammidae,  but  are  herein  treat- 
ed separately  because  of  the  distinctiveness  of  their  larvae  from 
hexagrammids  (Washington  and  Richardson,  MS;  Kendall  and 
Vinter,  1984).  Larvae  of  Oxylebius  pict us  and  Zaniolepis  sp.  are 
illustrated  and  described  by  Kendall  and  Vinter  (1984).  Larvae 
hatch  at  a  small  size  (2.5-5  mm  NL),  undergo  notochord  flexion 
between  6  and  9  mm  NL,  and  transform  to  a  benthic  juvenile 
at  about  1 5  mm  SL. 

Oxylebius  and  Zaniolepis  are  relatively  short  and  deep-bodied 
with  large,  bulging  guts  and  rounded  snouts.  Pectoral  fins  de- 
velop early  and  are  distinctively  large  and  fan-shaped.  Pigmen- 
tation is  heavy  over  the  anterior  half  of  the  body  in  preflexion 
larvae  and  increases  over  the  postanal  lateral  body  witb  devel- 
opment. Zaniolepis  possesses  characteristic  snout  pigment  which 
is  absent  in  Oxylebius.  The  pectoral  fins  of  both  species  are 
densely  pigmented. 

Head  spination  is  well-developed  with  preopercular  (5  spines 
in  Oxylebius;  6-7  in  Zaniolepis),  posttemporal  and  supraclei- 
thral  spines  present.  Zaniolepis  larvae  develop  distinctive  prick- 
le-scales over  most  of  the  body  by  about  7  mm. 

Normanichthyidae  (Fig.  225^. -Larvae  of  the  monotypic  Nor- 
manichthys  crockeri  are  illustrated  and  described  by  Balbontin 
and  Perez  (1980).  Hatching  occurs  at  a  small  size  (4.4  mm  NL) 
and  flexion  of  the  notochord  occurs  at  7  to  9  mm.  Development 
from  hatching  to  the  juvenile  stage  is  gradual  without  great 
change. 

Larvae  are  elongate  and  slender  with  short,  coiled  guts  and 
distinctive  large  pectoral  fins.  Pigmentation  is  restricted  to  the 


pectoral  fins  and  the  ventral  midline  extending  from  the  isthmus 
to  the  tail.  In  small  larvae  several  large  melanophores  are  pre- 
sent on  the  dorsal  midline. 

Distinctive  features  of  larval  development  include:  the  ab- 
sence of  head  and  preopercular  spines,  delayed  ossification,  early 
development  of  the  pectoral  fin,  and  presence  of  only  5  bran- 
chiostegal  rays. 

Comephoridae  (Fig.  225).— The  endemic  comephorids  of  Lake 
Baikal  in  Russia  are  reported  to  be  viviparous  (Chemyayev, 
1971,  1975)  and  are  bom  at  a  relatively  large  size  (8.2-9.4  mm) 
but  are  not  well  developed.  Flexion  of  the  notochord  occurs  at 
about  8.2  to  13  mm.  Larvae  develop  very  slowly  with  trans- 
formation occurring  3  or  4  months  after  birth. 

Larvae  are  extremely  slender  and  elongate  with  small  heads 
and  very  short  coiled  guts.  Comephorids  are  quite  different  from 
other  cottoids  morphologically  and  are  blennioid  in  appearance. 
Pigmentation  is  usually  limited  to  the  gut  and  sometimes  in  a 
series  along  the  postanal  lateral  midline.  Four  small  preoper- 
cular spines  develop  in  late-stage  larvae;  other  head  spines  are 
absent. 

Cottocomephoridae  (Fig.  225).  — Larvae  of  seven  genera  of  Lake 
Baikal  cottocomephorids  have  been  described  and  illustrated 
(Chemyayev,  1971,  1975.  1978,  1981;  Taliev,  1955).  Larvae 
hatch  at  about  5  to  10  mm,  and  range  from  forms  with  large 
yolk  sacs  and  no  fin  development  (e.g..  Paracottus)  to  well  de- 
veloped, postflexion  forms  with  fins  well  developed  (e.g.,  Ba- 
trachocottus).  Size  at  transformation  varies  from  9  to  20  mm. 
Larvae  are  slender  with  moderately  short  guts  and  rounded 
snouts,  somewhat  similar  to  freshwater  cotlids  (Coitus)  in  form. 
Pigmentation  is  variable  with  melanistic  pigmentation  usually 
present  on  the  head,  nape,  gut  and  variously  on  the  dorsal  and 
ventral  midline.  Melanophores  are  frequently  present  in  a  row 
along  the  lateral  midline  near  the  tail  tip. 

Larvae  develop  4  small  preopercular  spines  accompanied  by 
two  spiny  projections  from  an  inner  preopercular  shelf  Other 
head  spines  are  lacking. 

Cottidae  (Figs.  226-231).  — The  taxonomic  status  of  the  family 
Cottidae  is  controversial  with  the  number  of  recognized  families 
ranging  from  1  to  17  (see  Washington  and  Richardson,  MS).  To 
minimize  confusion,  and  because  there  is  no  generally  agreed 
upon  classification  of  this  "family,"  we  use  the  generic  groupings 
identified  by  Washington  and  Richardson  (MS)  for  our  discus- 
sion of  early  life  history  information.  Larvae  are  known  for  28 
of  the  70+  cottid  genera.  A  general  overview  of  larval  cottid 
taxonomy  is  presented  in  Richardson  and  Washington  (1980), 
Richardson  (1981a),  Washington  (1981)  and  Fahay  (1983). 

Rhamphocottus  (Fig.  226).  — Larvae  of  this  distinctive,  mono- 
typic species  hatch  at  a  relatively  large  size  (6-7  mm  NL).  No- 
tochord flexion  occurs  at  7  to  8  mm  and  transformation  to  a 


Fig.  229.  Larvae  of  the  Artedius  Part  A  group  (A-C)  and  the  Couus  group  of  cottids  of  Washington  and  Richardson  (MS)  (see  Washington 
et  al.,  this  volume).  (A)  .Artedius  fenestralis.  9.9  mm  SL  (from  Richardson  and  Washington,  1980.  as  .Artedius  Type  2);  (B)  ClmocoUus  acuticeps. 
10.4  mm  SL  (from  Washington,  in  prep.);  (C)  Oligocottus  snyderi.  10.2  mm  SL  (from  Washington,  1981);  (D)  Leptocottus  armatus.  8.1  mm  SL 
(from  Richardson  and  Washington,  1980);  (E)  Cottus  asper.  8.2  mm  SL  (ibid.). 


WASHINGTON  ET  AL.:  SCORPAENIFORMES 


423 


-.•I'V-^ 


y 


%C:jk!ddi:::^:k^:::k:^::^^^ 


424 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  230.  Larvae  of  the  Psychro/ules  group  (A,  B)  and  the  Malacocollus  group  (C,  D)  of  cottids  of  Washington  and  Richardson  (MS)  (see 
Washington  et  al..  this  volume).  (A)  Psychrolutes paradoxus.  13.0  mm  SL  (from  Richardson.  1981a);  (B)  Gilbenidia  sigalutes,  13.0  mm  SL  (ibid.); 
(Q  Dasycottus  seliger,  10.3  mm  SL  (original);  (D)  Malacocollus  zonurus,  9.8  mm  SL  (original). 


WASHINGTON  ET  AL.:  SCORPAENIFORMES 


425 


S5^ 


Fig.  231.  Larvae  of  the  Hemitripterus  group  (A-C)  of  cottids  of  Washington  and  Richardson  (MS)  (see  Washington  et  al..  this  volume)  and 
Agonidae.  (A)  Hemitripterus  vitlosus.  ca.  15.5  mm  SL  (from  Kyushm,  1968);  (B)  Blepsias  arrlwsus.  1  1.0  mm  SL  (from  Richardson,  1981a);  (C) 
Nautichthys  oculofasciatus.  1  1.7  mm  SL  (from  Richardson  and  Washington.  1980);  (D)  Agonomalus  or  Hypsagonus  sp.,  8.2  mm  SL  (original, 
courtesy  B.  Vinter). 


benthic  juvenile  occurs  at  about  14  to  15  mm  SL.  Rhampho- 
cottus  larvae  are  extremely  deep-bodied  with  a  very  long  snout- 
anus  length. 

Larvae  are  uniformly  covered  with  melanophores  except  for 
the  caudal  peduncle  and  ventral  surface  of  the  gut.  Rhampho- 
cottus  develop  small  prickle-scales  over  most  of  the  body  by  9 
or  10  mm.  Larvae  develop  only  one  preopercular  spine  in  con- 


trast to  the  usual  four  possessed  by  most  cottid  larvae.  Parietal, 
nuchal,  supracleithral,  posttemporal  and  postocular  spines  occur 
during  the  larval  period. 

Hemilepidotus-Scorpaenichthys  (Fig.  226).— Larvae  of  this  group 
hatch  at  4  to  6  mm  NL.  Transformation  to  the  neustonic  or 
pelagic  juvenile  phase  occurs  at  about  1 3  to  20  mm.  Larvae  are 


Fig.  232.  Larvae  of  Agonidae  (all  original).  (A)  Hypsagonus  quadricornis.  11.5  mm  SL;  (B)  Bolhragonus  swani.  6.3  mm  SL;  (C)  Xeneretmus 
latifrons.  9.6  mm  SL;  (D)  Slellerina  xyosterna.  10.2  mm  SL;  (E)  Ocella  verrucosa,  10.1  mm  SL;  (F)  Aspidophoroides  monopterygius.  14.3  mm 
SL. 


WASHINGTON  ET  AL.:  SCORPAENIFORMES 


427 


Fig.  233.     Larvae  of  Dactylopteridae.  (A)  Daayloplerus  voliians.  2.4  mm  (from  Padoa,  1956e);  (B)  Dactylopterus  volitans.  7.5  mm  (ibid.). 


long  and  slender  at  hatching  with  moderately  long  guts  (44  to 
60%)  and  rounded  snouts.  They  become  increasingly  deep-bod- 
ied with  development. 

Larvae  are  relatively  heavily  pigmented  with  melanophores 
over  the  head  and  gut.  Scorpaenichthys  larvae  have  dense  pig- 
ment covering  the  body  except  for  the  caudal  peduncle  while 
Hemilepidotus  spp.  have  postanal  pigment  concentrated  on  the 
dorsal  and  ventral  body  midlines.  Lateral  melanophores  de- 
velop above  and  below  the  notochord  in  Hemilepidotus. 

Hemilepidotus  and  Scorpaenichthys  larvae  develop  four 
prominent  preopercular  spines.  Hemilepidotus  possess  numer- 
ous head  spines  while  Scorpaenichthys  develop  bony  bumps  in 
corresponding  areas.  Larvae  of  this  group  develop  unique  pitted 
dermal  bones  on  the  head.  In  addition,  the  uppermost  pectoral 
radial  is  tiny  and  fuses  to  the  scapula  in  larval  Scorpaenichthys 
and  nearly  so  in  Hemilepidotus. 

Myoxocephalus  group  (Figs.  227  and  228).  — This  is  the  least 
well-defined  and  most  diverse  cottid  group  containing  1 3  genera. 
Where  known,  size  at  hatching  varies  from  2.9  to  10  mm.  Trans- 
formation to  the  benthic  juvenile  stage  vanes  from  7.6  to  20 
mm. 

Members  of  this  group  are  generally  slender-bodied  with 
pointed  snouts;  however,  Enophrys  is  stout-bodied,  and  Or- 
thonopias  has  a  blunt,  rounded  snout. 

Pigmentation  is  variable.  Heavy  pigment  on  the  dorsal  surface 
of  the  gut,  on  the  nape  and  along  postanal  ventral  midline  is 
characteristic  of  many  members  of  this  group.  Several  genera 
possess  heavy  melanistic  pigmentation  on  the  lateral  body  sur- 
face (e.g.  Radulinus,  some  Myoxocephalus).  Head  pigment  may 
be  present. 

Larvae  of  this  group  develop  four  preopercular  spines  and  a 


distinctive  bony  preopercular  shelf.  Parietal,  nuchal,  supra- 
cleithral,  posttemporal  and  occasionally,  postocular  spines  de- 
velop in  late-stage  larvae. 

Artedius  group  (Fig.  229,  Table  107).— This  group  contains  3 
genera,  Artedius  (in  part),  Clinocottus  and  Oligocottus  and  the 
larvae  have  been  described  by  Washington  (1981).  Larvae  hatch 
at  3  to  5  mm  and  transform  to  benthic  juveniles  at  approxi- 
mately 10  to  13  mm.  Larvae  are  stubby-bodied  with  a  slightly 
humped  appearance  at  the  nape.  Snouts  are  rounded  and  guts 
trail  distinctively  below  the  ventral  body  midline.  Several  species 
oi Artedius  develop  dorsal  gut  diverticula  while  Clinocottus  acu- 
ticeps  develops  long  hindgut  diverticula. 

Larvae  are  relatively  lightly  pigmented  and  characterized  by 
pigment  on  the  nape,  over  the  gut  and  along  the  postanal  ventral 
midline.  Head  pigment  is  present  in  some  species. 

Larvae  develop  a  unique  preopercular  spine  pattern  with  6 
to  24  spines.  Parietal  and  supracleithral  spines  are  variable  in 
this  group  and  may  form  in  clusters,  individually  or  not  at  all. 

Leptocottus  group  (Fig.  229).— This  group  includes  the  genera 
Leptocottus  and  Coitus.  Hatching  occurs  at  4  to  5  mm  and 
transformation  ranges  from  8  to  12  mm.  Larvae  are  relatively 
slender-bodied  with  rounded  snouts  and  moderately  short  guts. 
Pigmentation  is  usually  light  with  melanophores  on  the  nape, 
over  the  gut  and  widely  spaced  along  the  postanal  ventral  mid- 
line. Head  pigment  may  be  present. 

Where  known,  these  larvae  develop  four  weak  preopercular 
spines;  however,  other  head  spines  are  lacking. 

Psychrolutes  group  (Fig.  230).— This  group  includes  two  genera 
Psychrolutes  and  Gilbertidia.  Larvae  hatch  at  a  relatively  large 


428 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


size,  about  6  to  7  mm.  They  transform  and  settle  from  the 
plankton  at  about  1 8  to  20  mm  SL.  Larvae  are  generally  tadpole 
shaped  with  large  rounded  heads  tapering  toward  the  tail.  Larvae 
possess  an  outer  layer  of  loose  flabby  skin. 

Melanistic  pigment  occurs  on  the  head,  nape,  gut  and  char- 
acteristically on  the  pectoral  fins.  Postanal  ventral  midline  me- 
lanophores  are  absent;  however,  pigment  is  added  laterally  with 
development. 

Head  and  preopercular  spines  are  absent. 

Malacocottus  group  (Fig.  230).— This  group  includes  Malaco- 
cottus  and  Dasycottus.  Size  at  hatching  is  not  known.  Larvae  of 
this  group  are  similar  to  those  of  the  Psychwlules  group  with 
large,  blunt  heads  tapering  to  the  tail.  An  outer  bubble  or  layer 
of  skin  is  present  in  both  genera  and  is  particularly  pronounced 
in  Malacocottus. 

Pigmentation  is  present  on  the  head,  nape  and  over  the  entire 
gut.  Pigment  occurs  laterally  on  the  anterior  third  of  the  tail  in 
Malacocottus  larvae.  As  in  the  Psychrolutes  group,  the  pectoral 
fins  are  characteristically  pigmented. 

Larvae  develop  four  preopercular  spines  with  a  fifth  accessory 
spine  present  in  Malacocottus. 

Hemitripterus  group  (Fig.  23 1 ). — This  group  includes  the  genera: 
Hemiiripterus,  Blepsias  and  Nautichthys.  Hatching  occurs  at  a 
relatively  large  size,  7  to  1 3  mm  NL.  Newly-hatched  larvae  have 
elongate,  slender  bodies  which  become  deeper  with  develop- 
ment. Nautichthys  larvae  have  distinctively  long,  pigmented 
pectoral  fins. 

Pigmentation  is  relatively  heavy  with  melanophores  covering 
the  head,  dorsal  surface  of  the  gut  and  over  the  lateral  body 
surface  except  for  the  caudal  peduncle.  Nautichthys  and  Hem- 
itripterus larvae  possess  distinctive  pigment  bands  extending 
onto  the  dorsal  and  ventral  finfolds  that  are  not  found  in  other 
cottid  larvae. 

Larvae  develop  four  prominent  preopercular  spines  and  a 
strong  frontoparietal  spiny  ridge.  This  group  is  characterized  by 


delayed  ossification  in  the  larval  period  and  a  unique  "honey- 
comb" pattern  of  ossification  on  the  head.  Hemitripterus  larvae 
develop  large  bony  prickles,  similar  to  the  prickle-scales  found 
in  agonids. 

Agonidae  (Figs.  231  and  232).  — hi  least  one  early  life  history 
stage  of  9  of  the  49  nominal  species  is  known.  Agonids  hatch 
at  5.5  to  8.0  mm  NL.  Development  is  a  gradual  transformation 
to  the  juvenile  form  attained  at  20  to  30  mm. 

Agonid  larvae  are  generally  long  and  slender  with  relatively 
long  guts.  Extremes  of  form  range  from  short  stout  genera  such 
as  Agonomalus  and  Bothragomis  to  the  extremely  attenuated 
forms  such  as  Ocella  and  Aspidophoroides.  Larvae  have  dis- 
tinctively large,  fan-shaped  pectoral  fins. 

Pigmentation  varies  in  the  family.  Melanistic  pigment  may 
be  present  on  the  head,  nape,  scattered  over  the  gut  and  fre- 
quently in  bands  on  the  postanal  lateral  surface  of  the  body. 
The  pectoral  fins  are  distinctively  pigmented  often  with  distal 
bands  of  melanistic  pigment.  In  some  species  (e.g.  .-igonomalus, 
Hypsagonus)  pigmentation  extends  onto  the  dorsal  and  ventral 
finfolds. 

Larvae  are  characterized  by  spiny  heads  with  large  fronto- 
parietal spiny  ridges,  postocular  spines,  and  usually  four  large 
preopercular  spines.  Tiny  rows  of  spines  form  in  small  larvae 
and  help  distinguish  agonid  larvae.  These  rows  correspond  to 
the  plates  (scales)  of  adults. 

(B.B.W.)  Gulf  Coast  Research  Laboratory,  East  Beach 
Drive,  Ocean  Springs,  Mississippi  39564;  (W.A.L.)  School 
OF  Natural  Resources,  Department  of  Fisheries, 
Humboldt  State  University,  Arcata,  California  95521; 
(H.G.M.)  National  Marine  Fisheries  Service,  Southwest 
Fisheries  Center,  Post  Ofhce  Box  271.  La  Jolla,  Cal- 
ifornia 92038;  (W.J.R.)  National  Marine  Fisheries  Ser- 
vice, Southeast  Fisheries  Center,  75  Virginia  Beach 
Drive,  Miami,  Florida  33149. 


Cyclopteridae:  Development 
K.  W.  Able,  D.  F.  Markle  and  M.  P.  Fahay 


THE  scorpaeniform  family  Cyclopteridae  is  composed  of  two 
subfamilies  (Nelson,  1976),  the  Cyclopterinae  (lumpfishes) 
with  7  nominal  genera  and  28  species,  and  the  Liparidinae 
(snailfishes)  with  18  nominal  genera  and  150+  species  (Table 
108).  Some  authors  have  considered  the  subfamilies  as  separate 
families  (Gill,  1891;  Garman,  1892;  Jordan  and  Evermann. 
1896-1900;  Regan,  1929;  Burke,  1930;  Matsubara,  1955;  Ueno, 
1970),  while  others  have  treated  them  together  (Boulenger,  1910; 
Berg,  1940;  Greenwood  et  al..  1966).  We  follow  Nelson  (1976) 
without  prejudice;  both  groups  appear  distinct  yet  are  clearly 
sister  taxa.  The  most  compelling  synapomorphy  is  a  ventral 
sucking  disk  (secondarily  lost  in  some  liparidines)  formed  from 
pelvic  fin  rays.  The  cyclopterid  disk  differs  structurally  from 


analogous  structures  in  Gobiesociformes  and  Gobiidae  (see  for 
example,  Briggs,  1955;  Ueno,  1970).  Certain  osteological  (Ueno, 
1970)  and  menstic  differences  (Table  108)  between  the  subfam- 
ilies are  marked.  The  lumpfishes  have  two  dorsal  fins  (the  first 
dorsal  may  be  embedded  in  the  skin  and  not  externally  visible 
in  some  genera)  with  few  total  elements  (4-8  spines  and  8-12 
rays),  few  anal  rays  (6-13)  and  vertebrae  (23-29).  The  snailfishes 
have  a  single  dorsal  fin  with  numerous  elements  (28-82),  and 
more  anal  rays  (24-76)  and  vertebrae  (38-86)  (Table  108). 

Representatives  of  the  Liparidinae  have  been  collected  in  all 
oceans  from  the  Arctic  to  the  Antarctic.  They  are  found  from 
intertidal  depths  to  greater  than  7  km  (Andriashev.  1954;  1975). 
However,  their  distribution  over  shallow  continental  shelves  is 


ABLE  ET  AL.:  CYCLOPTERIDAE 


429 


Table  108.    Nominal  Cyclopterid  Genera.  Nlimber  of  Species,  and  Range  of  Meristic  Characters  for  Each.  Based  primarily  on  data 
from  Burke  (1930),  Schmidt  (1950),  Ueno  (1970),  Andriashev  (1975),  Andriashev  and  Neelov  (1976),  Stein  (1978),  and  Kjdo  (1983).  Dorsal  fin 

counts  are  given  as  dorsal  spines  and  dorsal  rays  for  Cyclopterinae. 


No.  of 
species 

Fin  rays 

Pylonc  caecae 

Genus 

Dorsal 

Anal 

Pectoral 

Caudal 

Vertebrae 

Cyclopterinae 

Aptocyclus  De  La  Pylaie 

1 

V,  8- 

•11 

6-9 

19-22 

9-11 

15-43 

27-29 

Cyclopsis  Popov 

1 

VI,  11 

-12 

10 

23-24 

10-11 

9 

7 

Cyclopleropsis  Soldatov  and  Popov 

8 

VI-VII.   IC 

1-12 

9-13 

25-28 

9-11 

9-10 

25-26 

Cyclopterus  Linnaeus 

1 

VI-VIII,  9- 

•11 

9-10 

19-20 

11-12 

36-79 

28-29 

Eumicrolremus  Gill 

14 

V-VIII,  9- 

•13 

9-13 

19-29 

9-12 

8-12 

26-29 

Letholremus  Gilbert 

2 

VI-VII,  8- 

•11 

7-10 

20-23 

10-11 

4 

23-24 

Pelag(xychis  Lmdberg  and  Legeza 

1 

IV-V,  9- 

■10 

8-9 

19-21 

10 

7 

7 

Liparidinae 

Acanthotipans  Gilbert  and  Burke 

2 

45-52 

38-47 

20-26 

8-10 

0-6 

50-54 

Careproclus  Kroyer 

47  + 

40-67 

32-60 

17-37 

6-12 

0-60 

47-71 

Crystallias  Jordan  and  Snyder 

1 

56 

53 

31 

10 

71 

7 

Cryslallichthys  Jordan  and  Gilbert 

2 

48-53 

42-44 

30-35 

10-12 

36-40 

7 

Elassodiscus  Gilbert  and  Burke 

2 

49-68 

45-60 

27-32 

8-9 

ca.  14-16 

60 

Genioliparis  Andriashev  and  Neelov 

1 

53 

49 

19 

6 

7 

62 

Gynnichlhys  Gilbert 

1 

— 

_ 

25 

14 

— 

•y 

Liparis  Scopoli 

50-60 

28-49 

24-45 

28-41 

10-12 

10-90 

38-53 

Lipanscus  Gilbert 

I 

50-52 

47-49 

13-15 

4 

ca.  6 

7 

Nectolipans  Gilbert  and  Burke 

1 

50-55 

45-50 

19-25 

6 

6-9 

7 

Notolipans  Andriashev 

3 

41-57 

38-53 

31 

10 

_ 

50-65 

Odontolipans  Stein 

1 

51 

46 

17 

6 

7 

59 

Osteodiscits  Stein 

1 

47-52 

40-44 

20-25 

6-7 

0 

51-56 

Paralipans  Collett 

27  + 

48-82 

42-76 

14-39 

3-8 

5-41 

57-86 

Polypera  Burke 

3 

37-44 

31-34 

33-37 

_ 

200-300 

7 

Rhinn/ipans  Gilbert 

2 

ca.  68 

ca.  60 

20-23 

1-3 

7-12 

7 

Rhodkhlhys  Collett 

1 

56-60 

54-57 

16-17 

10 

? 

ca.  65 

Temnocora  Burke 

1 

45-48 

39 

33-37 

- 

20 

9 

limited  to  the  cooler  waters  of  the  arctic,  antarctic  and  temperate 
regions  with  the  possible  exception  of  L.  ftshclsoni  from  the  Red 
Sea  (Smith,  1968).  The  Cyclopterinae  are  more  restricted  in 
their  distribution,  occurring  exclusively  in  the  northern  hemi- 
sphere's boreal  and  arctic  waters  (Ueno.  1970)  where  they  are 
usually  limited  to  continental  shelves.  Although  most  cyclop- 
teridsare  benihic  the  cycloptenne  Pelagocyclus  v/7/ar/(Lindberg 
and  Legeza,  1955)  and  the  liparidines  Nectolipans pclagiciis  and 
Lipanscus  nanus  (Stein,  1978)  are  pelagic.  Lipans  fahricii  is 
considered  cryopelagic  in  the  high  Arctic  (Tsinovsky  and 
Mernikov,  1980).  The  cycloptenne  Cyclopterus  lumpus  is  ben- 
thic  during  the  reproductive  season  and  pelagic  at  other  limes 
(Thorsteinsson,  1981;  Able,  in  prep.). 

Development 

The  available  information  on  early  life  history  stages  is  in- 
adequate to  allow  confident  generalizations  about  the  biology 
or  systematics  for  most  members  of  the  family.  This  is  due  to 
rarity  of  material  (adults  and  especially  larvae)  and  the  incom- 
plete understanding  of  cyclopterid  taxonomy. 

Eggs 

Cyclopterid  eggs  are  moderate  to  large  (0.8  to  8.0  mm),  de- 
mersal and  adhesive  (Table  109).  Variation  in  fecundity  is  gen- 
erally related  to  female  length  (Stein,  1980a;  Lisovenko  and 
Svetlov,  1981;  Matarese  and  Borton,  in  prep.)  but  appears  to 
be  a  complex  function  of  egg  diameter  as  well  (Table  109).  Much 
of  the  available  information  on  cyclopterid  eggs,  summarized 


in  Table  109,  is  based  on  observations  of  ovarian  eggs  or  oth- 
erwise incomplete  descriptions.  It  is  possible,  for  example,  that 
one  or  more  oil  globules  may  be  characteristic  of  all  cyclopterid 
eggs.  Sculpturing  of  the  chorion  surface  has  been  reported  for 
Lipans  montagui  (Mcintosh  and  Prince,  1890),  L.  tanakae 
(Aoyama,  1959)  and  L.  atlanticus  (Detwyler,  1963).  Pores  in 
the  chorion  have  been  reported  for  Cycloptents  lumpus  and  L. 
montagui  (Mcintosh  and  Prince,  1890).  We  have  found  that 
sculpturing  of  the  chorion  occurs  in  L.  liparis  (Fig.  234A,  B,  C) 
and  Paraliparis  calidus  and  possibly  Eumicrolremus  orbis  (Fig. 
234D,  E,  ¥).  Pores  in  the  chorion  are  quite  numerous  in  L. 
liparis  (Fig.  234B,  C).  Pits  are  present  on  some  portions  of  the 
egg  surface  of  £.  orbis  (Fig.  234D,  E,  F). 

Incubation  is  moderately  long  (5  to  10  weeks)  in  the  few 
reported  cases  (Russell,  1976;  Andriashev,  1954;  Matarese  and 
Borton,  in  prep.).  The  combination  of  relatively  large  eggs  and 
long  incubation  times  results  in  an  advanced  state  of  develop- 
ment at  hatching  for  some  members  of  each  subfamily.  In  these 
instances  fin  rays  and  disk  are  formed  and  notochord  flexion  is 
underway  prior  to  hatching  (Fig.  235A,  Table  110).  Hatching 
at  an  advanced  state  of  development  is  characteristic  for  all 
deep-water  Liparidinae  that  have  been  relatively  well  studied 
(Andriashev  et  al.,  1977;  Stein,  1978).  Hatching  may  be  cued 
to  wind  induced  temperature  changes  for  some  inshore  Liparis 
(Frank  and  Leggett,  1983). 

Some  form  of  parental  protection,  either  egg  hiding,  paternal 
guarding,  or  both  may  also  be  characteristic  (Table  109).  Some 
Pacific  Careproctus  deposit  eggs  within  the  gill  cavities  of  lith- 


430 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  109.    Summary  of  Egg  Characteristics  of  Cyclopteridae. 


Egg  or 

maximum  ovar- 

ian 

Clutch 

egg  diameter 

Oil 

count  or  npe 

Paternal  care  or 

Species 

(mm) 

globule(s) 

egg  fecundity 

egg  deposition  sites 

Source(s) 

Cyclopterinae 

Aptocyclus  ventricosus 

2.3-2.4 

present 

3,800 

paternal  guarding 

Kyushin,  1975;  Schmidt, 
1950;  Kobayashi,  1962 

Cyclopsis  lentacularis 

2.0 

7 

1,540 

7 

Lindberg  and  Legeza,  1955 

Cyclopieropsis  macalpini 

5.0 

? 

60-70 

paternal  guarding, 
hidden  (mollusc  shells) 

Parr,  1926 

Cyctopterus  lumpus 

2.2-2.7 

present 

15,000-200,000 

paternal  guarding 

Zhitenev,  1970;  Russell, 
1976;  Andriashev,  1954 

Eumkrotremus  birulai 

3.9-4.0 

present 

1,230 

hidden  (mollusc  shells) 

Honma,  1956;  Ueno,  1970 

Eumicrotremus  derjugini 

4.0-5.0 

7 

7 

7 

Andriashev,  1954 

Eumicrotremus  orbis 

2.2 

present 

325-477 

paternal  guarding 

Matarese  and  Borton,  MS 

Eumicrotremus  soldatori 

3.1 

7 

4,049 

•> 

Ueno,  1970 

Eumkrotremus  spinosus 

3.2-4.5 

7 

7 

paternal  guarding, 
hidden  (mollusc  shells) 

Andriashev.  1954 

Lethotremus  awae 

1.4 

7 

232 

7 

Ueno,  1970 

Liparidinae 

Acantholiparis  opercularis 

4.8 

7 

1-6 

7 

Stein,  1980a 

Careproctus  sp. 

5.0 

7 

7 

hidden  (lithodid  crab 
gill  cavity) 

Hunter,  1969;  Vinogradov. 
1950 

Careproctus  sp. 

3.0-3.5 

7 

100 

hidden  (lithodid  crab 
gill  cavity) 

Anderson  and  Cailliet, 
1974 

Careproctus  falklandwa 

? 

7 

? 

hidden  (lithodid  crab 
gill  cavity) 

Balbontin  et  al.,  1979 

Careproctus  longifilis 

7.1 

7 

16 

7 

Stein,  1980a 

Careproctus  melanurus 

4.2 

7 

534 

hidden  (lithodid  crab 
gill  cavity) 

Peden  and  Corbett,  1973 

Careproctus  mkrostomus 

7.6 

present 

15 

7 

Stein,  1980a 

Careproctus  oregonensis 

5.6 

7 

3-5 

7 

Stein,  1980a 

Careproctus  ovigerum 

7.8 

7 

756 

7 

Stein,  1980a 

Careproctus  raslrinoides 

4.5 

7 

7 

7 

Schmidt,  1950 

Careproctus  reinhardti 

4.5 

7 

300 

7 

Collett.  1905; 
Andriashev,  1954 

Careproctus  sinensis 

5.0 

7 

7 

hidden  (lithodid 
crab  gill  cavity) 

Rass.  1950 

Liparis  atlanticus 

0.8-1.4 

present 

1 ,400-3,000 

paternal  guarding, 
hidden  (algae) 

Detwyler,  1963 

Liparis  fabricii 

2.1-2.7 

7 

485-735 

7 

Andriashev,  1954 

Liparis  fucensis 

1.0 

7 

7 

paternal  guarding, 
hidden  (mollusc 
shell,  tubeworms) 

DeMartini,  1978; 
Marliave.  1976 

Liparis  inquitinus 

1.0-1.3 

present 

231-563 

hidden  (hydroids) 

Able  and  Musick,  1976 

Liparis  liparis 

1.4-1.7 

present 

7 

hidden  (hydroids) 

Russell,  1976 

Liparis  montagui 

1.0-1.2 

present 

700 

hidden  (red  algae) 

Russell.  1976; 
Andriashev.  1954 

Liparis  pukhellus 

1.5 

7 

941-996 

? 

Johnson,  1969 

Liparis  tanakae 

1.7-1.8 

present 

7 

hidden  (sea  weed) 

Aoyama,  1959 

Notoliparis  kermadecensis 

8.0 

7 

16 

7 

Neilsen,  1964 

Osteodiscus  cascadiae 

5.3 

7 

1-5 

7 

Stein,  1980a 

Paraliparis  bathybius 

4.5 

7 

422-434 

? 

Collett,  1905; 
Andriashev,  1954 

Paraliparis  calidus 

2.6-2.9? 

7 

7 

? 

Wenner,  1979 

Paraliparis  copei 

2.0 

7 

45-86 

7 

Wenner,  1979 

Paraliparis  deani 

2.0 

7 

0 

7 

Hart,  1973 

Paraliparis  garmani 

3.5 

7 

190-317 

oral  brooding?/ 
paternal  guarding? 

Wenner,  1979;  Stein. 
1980a 

Paraliparis  gracilis 

2.6-2.9 

7 

7 

7 

Marshall,  1953 

Paraliparis  latifrons 

4.5 

7 

2-8' 

? 

Stein,  1980a 

Paraliparis  megalopus 

4.3 

7 

32 

7 

Stein,  1980a 

Paraliparis  mento 

2.5 

7 

101 

7 

Stein,  1980a 

Paraliparis  rosaceus 

3.6 

7 

1,277 

? 

Stein,  1980a 

Rhinoliparis  barbulifer 

2.5 

7 

7 

7 

Schmidt,  1950 

Rhodichthys  regina 

3.2-4.0 

7 

70 

7 

Johnsen,  1921 

ABLE  ET  AL.:  CYCLOPTERIDAE 


431 


Fig.  234.  Scanning  electron  micrographs  of  Lipans  liparis  egg  (A,  B,  C,  Zoologisch  Museum  Amsterdam  1  14.522.  North  Sea)  and  Eumicro- 
iremus  orhis  egg  (D,  E,  F)  from  the  study  by  Matarcse  and  Borton  (in  prep.).  The  depression  in  A  and  B  is  the  micropyle.  Scale  bar  equals  200 
n  (A),  19  M  (B),  4.9  M  (C),  280  m  (D).  28  m  (E),  3.3  m  (F). 


odid  crabs;  a  site  which  may  provide  both  protection  and  water 
circulation. 


Larvae 

In  the  Cyclopterinae  development  has  only  been  described 
for  4  of  7  nominal  genera  and  4  species  (Table  110).  Other 
partial  descriptions  are  for  Aptocyclus  venlricosus  (Kobayashi, 
1962)  and  Eumicrotremus  spinosus  (Ehrenbaum,  1905-1909; 
Koefoed,  1909).  In  the  Liparidinae,  larvae  of  3  of  18  nominal 
genera  and  10  species  have  been  described  (Table  1 10).  Besides 
those  listed,  partial  descriptions  have  been  published  for  Car- 
eproclus  georgianus  (Efremenko,  1983a),  Careproctus  falklan- 
dica  and  Careproctus  sp.  (Balbontin  et  al.,  1979)  and  several 
Liparis:  L.  atlanticus  (Detwyler.  1963),  L./a/>na/ (Ehrenbaum, 
1905-1909;  Koefoed,  1909;  Johansen,  1912;  Dunbar,  1947),  L. 
fuscensis  (Marliave,  1976),  L.  lipans  (Ehrenbaum  1904,  1905- 


1909;  Ehrenbaum  and  Strodtman,  1904;  Page,  1918),  L.  mon- 
?a.^/ (Mcintosh  and  Prince,  1890;  Mcintosh  and  Mastermann, 
1897;  Ehrenbaum  and  Strodtman,  1904;  Ehrenbaum,  1905- 
1909;  Page,  1918;  Arbault  and  Boutin,  1968b),  L. /a«aA:ae(Aoy- 
ama,  1959;  Kim  etal.,  1981),  and  L. /;/«/<:a/!/5( Johansen,  1912). 

Morphological  characters.  — CycXoplend  larvae  typically  have 
flaccid  skin  enveloping  the  entire  body,  a  short  bulbous  head 
usually  without  spines,  large  eyes,  and  a  trilobed  lower  lip.  The 
sucking  disk  forms  early  in  development  and  may  be  present  at 
hatching  in  some  forms  (Fig.  235-238).  The  preanal  length  is 
short  and  the  gut  is  coiled.  Cyclopterines  may  have  both  dorsal 
fins  at  hatching  (Fig.  235B),  typically  have  larger  disks  at  hatch- 
ing, and  usually  have  more  pigmentation  at  hatching  (Pig.  235- 
238)  than  liparidine  larvae.  Some  cyclopterine  larvae  develop 
dermal  spines  that  become  pronounced  tubercles  in  adults  (Ueno, 
1970).  In  many  liparidine  larvae  the  medial  surface  of  the  pec- 


432 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  235.     Egg  (A)  and  larvae  (B  — 4.5  mm  SL,  C— 6.3  mm  SL,  ventral  view)  of  Eumicrotremiis  orbis  from  Matarese  and  Borton  (in  prep.)  and 
larvae  of  Cycloplerus  lumpus  (D— 5.0  mm  SL.  Damariscotta  River,  Maine,  HML  H-24029). 


toral  fin  has  numerous  melanophores  (Fig.  237)  and  during  de- 
velopment the  fin  may  become  bilobed  (Fig.  238).  The  gill  open- 
ing decreases  in  size  during  development. 

Disk  size  varies  within  each  subfamily  and  may  be  related  to 
habitat.  Pelagic  forms  such  as  the  cyclopterine  Pelagocychis  vi- 
tiazidind  the  liparidine  L.  fabricii (¥i%.  237)  have  small  orgreatly 
reduced  disks.  Some  pelagic  forms,  such  as  Nectoliparis  pclag- 
icus  and  Lipanscus  nanus  lack  disks  entirely. 

The  arrangement  of  the  cranium  may  offer  useful  insights  into 
cyclopterid  phylogeny.  Svetovidov  ( 1 948)  noted  that  the  cranial 
cavity  extends  into  the  interorbital  space  in  Liparis  but  only 
reaches  the  hind  margin  of  the  orbit  in  Cycloplerus.  Our  material 
indicates  that  this  character  state  changes  ontogenetically  in 
Cycloplerus  with  the  earliest  stages  showing  the  liparidine  state. 


Able  and  McAllister  (1980)  suggested  that  tooth  shape  ex- 
hibits polarity,  with  trilobed  teeth  with  equal  lobes  representing 
the  primitive  condition,  trilobed  teeth  with  a  larger  central  lobe 
an  intermediate  condition,  and  simple  teeth  the  derived  con- 
dition. The  ontogeny  of  teeth  in  Liparis  supports  this  statement. 
All  Liparis  examined  to  date  possess  trilobed  teeth  early  in 
development.  With  continued  growth  the  oldest  teeth  may  be- 
come simple,  as  in  L.  fabricii  (Able  and  McAllister.  1980). 

Caudal  morphology  and  ontogeny  show  variation  that  may 
prove  useful  for  identification  and  phylogenetic  studies.  Within 
liparidines  caudal  elements  vary.  For  example.  Stein  (1978) 
noted  a  lack  of  epurals  in  Careproclus  longifilis,  whose  caudal 
structure  he  considered  typical  of  deepwater  eastern  Pacific  li- 
paridines he  examined,  while  we  note  the  typical  presence  of 


Fig.  236.     Larvae  of  Lipans  (from  top  to  bottom).  Liparis  allanUciis  (7.9  mm  NL.  47°37'N,  62°02'W,  HML  H-2140);  ventral  view  of  above; 
L.  cohem{\i.6  mm  NL,  Damariscotta  River,  Mame,  HML  H-24030);  and  an  unidentified  cyclopterid  (5.8  mm,  CALCOH  6401  Sta.  70.52). 


434 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  237.     Liparis  fabricii  larvae  (from  top  to  bottom):  16.7  mm  NL,  NZ  4,  74°06'N,  81°30'W,  NMC  83-1  135;  ventral  view  of  above;  32. i 
mm  NL,  NZ  292,  74°27'N,  82°03'W,  NMC  83-1 136,  from  Arctic  Canada. 


two  epurals  in  some  western  North  Atlantic  Liparis  (Fig.  239) 
from  shallow  water.  Caudal  development  also  varies.  In  Cy- 
cloplerus  (Fig.  239),  Eumicrotremus  (Matarese  and  Borton,  in 
prep.),  and  deepwater  southern  hemisphere  liparidines  (An- 
driashev  et  al.,  1977)  the  notochord  is  resorbed  and  flexion  is 
complete  at  hatching.  In  western  Atlantic  Liparis.  especially  L. 
fabricii.  notochord  resorption  and  flexion  are  delayed  as  late  as 
50  mm  SL  (Fig.  237,  Table  1 10). 
Body  proportions  are  also  useful  taxonomic  characters  for 


larval  identification.  Within  Liparis.  larval  L.  fabricii  are  sep- 
arable from  other  western  North  Atlantic  Liparis  by  a  relatively 
shorter  head  length,  smaller  eye  diameter,  shallower  body  depth 
and  shorter  preanal  distance.  The  disk  size  relative  to  eye  length 
has  also  proven  effective  in  distinguishing  between  all  species 
of  western  North  Atlantic  Liparis  (Able  et  al.,  MS).  The  size  of 
the  gill  opening  is  difficult  to  measure  consistently  but  it  de- 
creases as  development  proceeds  in  Liparis.  suggesting  that  a 
reduced  gill  opening  is  a  derived  character  state. 


Fig.  238.  Larvae  of  Careproclus  and  Paraliparis  (from  top  to  bottom).  Careproclus  reinhardti.  with  yolk  sac,  1 2.6  mm  SL,  Chaleur  Bay,  Gulf 
of  St.  Lawrence,  Canada,  HML  H-24031;  ventral  view  of  above;  Paraliparis  copei.  24.0  mm  SL,  St.  Lawrence  River  estuary,  Canada,  HML 
H-24032;  and  P.  calidus  12.9  mm  SL,  St.  Lawrence  River  estuary,  Canada,  HML  H-24033;  ventral  view  of  above. 


ABLE  ET  AL.:  CYCLOPTERIDAE 


435 


tsoU 


436 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  239.  Caudal  development  of  Cyclopterus  tuinpus  (A)  6.0  mm  NL,  (B)  12.5  mm  SL.  (C)  18.0  mm  SL,  HML  H-3093,  43°12'N,  66°00'W; 
and  Liparis  fahricii  (D)  ca.  20  mm  NL,  72°30.4'N,  76°46.2'W,  NMC  83-1 137;  (E)  ca.  34  mm  NL,  74°27'N,  82°03'W,  NMC  83-1 138;  (F)  ca.  145 
mm  SL,  70°07'15"N,  60°44'15"W,  NMC  83-1 139).  Scale  bars  equal  one  mm. 


The  arrangement  and  degree  of  adherence  of  the  soft  flaccid 
skin  of  cyclopterid  larvae  may  be  of  taxonomic  value.  In  Cy- 
clopterus and  all  western  North  Atlantic  forms  examined  (Able 
et  al.,  MS)  the  skin  conforms  loosely  to  the  entire  surface  of  the 
body.  In  an  unidentified  cyclopterid  from  the  eastern  Pacific  the 
skin  forms  a  distinct  bubble  over  the  anterior  portion  of  the 
body  and  then  adheres  tightly  over  the  posterior  portion  (Fig. 
236). 


Pi,^n7enl.  —  Cycloptennes  are  usually  more  heavily  pigmented 
at  hatching  (Kyushin,  1975;  Matarese  and  Borton,  in  prep.;  Fig. 
235)  than  liparidines  (Figs.  236,  238).  An  exception  is  L.  fahricii 
(Fig.  237)  which  has  well-developed  pigment.  Throughout  de- 
velopment, all  Liparis  we  examined  from  the  western  North 
Atlantic  possess  melanophores  on  the  medial  surface  of  the 
pectoral  fin,  on  the  abdomen,  and  a  line  of  melanophores  at  the 
base  of  the  anal  finfold  and  fin.  The  abdominal  melanophores 


ABLE  ET  AL.:  CYCLOPTERIDAE 


437 


Table  1 10.    Ontogeny  of  Character  Development  for  Cvclopterids  Based  on  Available  Literatlire.  Stage  of  development  at  hatching 
indicated  by  pre  (preflexion).  and  flex  (flexion).  X  indicates  event  takes  place  before  hatching. 


Stage 

Length  (min)  at  development  of  character 

al 

Disk 

Nostnl 

Species 

Source 

hatching 

Hatching 

formation 

Rexion 

splitting 

Poslflexion 

Demersal  phase 

Cyclopterinae 

Aptocyclus 

Kyushin, 

1975 

flex 

6.5-7.0  TL 

X 

X 

X 

7 

at  hatching 

ventricosus 

Cyclopteropsis 

Parr,  1926 

flex 

9 

X 

X 

? 

7 

7 

macalpim 

Cyclopterus 

Fritzsche, 

1978 

flex 

4.0-7.4  TL 

X 

X 

? 

8.0-10.0  TL? 

7 

lumpus 

Eumicrolremus 

Matarese  and 

flex 

4.5-4.7  SL 

X 

X 

X 

55  SL 

at  hatching 

orbis 

Borton, 

in  prep. 

Liparidinae 

Careproctus 

Peden  and  Corbett, 

flex? 

? 

X 

X 







metanuius(1) 

1973 

Careproctus 

Able  et  al. 

,  MS 

flex 

ca.  9.8  NL 

X 

X 

— 

17.2-21.1  SL 

at  hatching? 

reinhardli 

Liparis 

Able  et  al. 

,  MS 

pre 

ca.  3.1  NL 

3.3  NL 

5.8-6.9  NL 

5.4-6.3  NL 

12.1-17.1  SL 

7 

atlanticus 

Liparis 

Able  et  al. 

,  MS 

pre 

ca.  3.4  NL 

3.7-6.0  NL 

8.1-8.5  NL 

5. 

1-9.4  NL 

14.9-19.0  SL 

14  TL 

inquilinus 

Liparis  coheni 

Able  et  al. 

,  MS 

pre 

ca.  5  NL 

ca.  5.0  NL 

8.5-9.3  NL 

7.8-9.6  NL 

19.9-20.7  SL 

29-36  TL 

Liparis  gihbi/s 

Able  et  al. 

,  MS 

pre 

4.8  NL 

7.4  NL 

7.4-10.3  NL 

12. 

7-15.4  NL 

20.0-41.7  SL 

9 

Liparis 

Able  et  al. 

,  MS 

pre 

ca.  8  NL 

8.6-11.5  NL 

11.9-13.4  NL 

14. 

1-17.2  NL 

48.2-52.1  SL 

7 

fabncii 

Liparis 

Able  et  al. 

,  MS 

pre 

7 

9 

? 

7 

20.8-27.0  SL 

7 

tunicatus 

Paraliparis 

Able  et  al. 

,  MS 

flex? 

? 

— 

X? 



18.6-20.7  SL 

at  hatching? 

calidus 

Paraliparis 

Able  et  al. 

,  MS 

flex? 

? 

— 

X'' 

— 

ca.  18.0  SL 

at  hatching? 

copei 

are  variable,  with  some  species  lacking  meianophores  on  the 
ventral  surface  behind  the  disk  (Fig.  236)  while  in  L.  fabricii 
they  are  prominent  (Fig.  237).  A  second  row  of  meianophores 
occurs  on  the  edge  of  the  anal  finfold  in  preflexion  L.  fucensis 
(Marliave,  1976).  The  early  appearance  of  meianophores  on  the 
lateral  surface  of  the  tail  pnor  to  beginning  of  notochord  flexion 
is  diagnostic  for  L.  atlanticus  (Fig.  236)  among  western  North 
Atlantic  Liparis  with  the  exception  of  L.  fabricii.  Liparis  fabricii. 
unlike  other  Liparis  examined,  has  numerous,  stellate  meia- 
nophores over  most  of  the  body  and  these  become  increasingly 
numerous  with  development  (Fig.  237).  By  late  flexion  the  pe- 
ritoneum is  completely  black,  the  pectoral  fins  and  head  are 
very  dark,  and  oblique  patches  of  meianophores  are  apparent 
on  the  dorsal  and  anal  fins  (Fig.  237).  All  of  these  patterns  are 
unique  to  this  species  and  suggest  that  it  may  be  relatively  iso- 
lated within  the  genus.  Generally,  pigmentation  patterns  should 
be  used  with  caution  since  geographical  vanation  does  occur, 
as  for  L.  gibbus  (Able  et  al.,  MS). 

Ontogenetic  schedule.— On  the  basis  of  current  information,  it 
appears  that  certain  developmental  landmarks  are  useful  for 
distinguishing  between  groups  of  cycloptends  and  may,  in  some 
instances,  reflect  relationships.  The  degree  of  development  at 
hatching  is  variable  both  between  and  within  subfamilies  (Table 
110).  All  cyclopterines  studied  hatch  late  in  development,  at 
relatively  large  sizes,  when  many  developmental  characters  are 
nearly  complete  (see  Fig.  235).  Embryonic  development  is  more 


variable  within  the  liparidines  (Table  1 10);  some  Careproctus, 
Paraliparis  {Fig,.  238)  and  other  deepwater  forms  from  the  south- 
em  hemisphere  (Marshall,  1953;  Andriashev  et  al.,  1977)  ap- 
parently hatch  late  in  development,  at  large  sizes  while  shallow 
water  Liparis  studied  to  date  hatch  as  preflexion  larvae  (Able 
et  al.,  MS). 

Within  Liparis,  the  development  of  several  characters  occurs 
over  a  wide  size  range  (Table  1 10).  For  example,  in  L.  atlanticus 
and  L.  inquilinus  hatching,  disk  formation,  nostril  splitting,  flex- 
ion and  postflexion  and  assumption  of  demersal  habitat  occur 
at  relatively  small  sizes,  while  in  L.  fabricii  all  of  these  events 
are  delayed  until  larger  sizes.  Other  species  (L.  coheni,  L.  gibbus) 
are  intermediate.  Liparis  fabricii.  which  shows  the  most  delayed 
development,  may  remain  pelagic  throughout  its  life  (Able  and 
McAllister,  1980;  Tsinovsky  and  Mel'nikov,  1980).  While  some 
of  this  variation  may  be  explained  by  the  variation  in  egg  size 
it  can  not  account  for  the  great  differences  observed.  We  suggest 
that  delayed  development  is  associated  with  delayed  assumption 
of  the  demersal  habitat  and  that  this  represents  neoteny. 

(K.W.A.)  Biological  Sciences  and  Center  for  Coastal  and 
Environmental  Studies,  Rutgers  University,  New 
Brunswick,  New  Jersey  08903;  (D.F.M.)  Huntsman 
Marine  Laboratory,  St.  Andrews,  New  Brunswick  EOG 
2X0,  Canada;  (M.P.F.)  National  Marine  Fisheries  Ser- 
vice, Northeast  Fisheries  Center,  Sandy  Hook 
Laboratory,  Highlands,  New  Jersey  07732. 


Scorpaeniformes:  Relationships 
B.  B.  Washington,  W.  N.  Eschmeyer  and  K.  M.  Howe 


THE  order  Scorpaeniformes  is  a  large,  morphologically  di- 
verse group  containing  about  20  families  (depending  on 
classification  used),  250  genera,  and  over  1,000  species.  The 
order  is  defined  by  the  presence  of  a  suborbital  stay,  a  posterior 
extension  of  the  third  infraorbital  bone  which  in  nearly  all  species 
is  firmly  attached  to  the  preopercle.  Infraorbital  bones  for  many 
scorpaeniform  groups  were  discussed  most  recently  by  Poss 
(1975).  Many  workers  have  suggested  that  the  stay  may  have 
evolved  independently  (Matsubara,  1943;Quast,  1965;  Green- 
wood et  al.,  1966;  Poss,  1975;  and  Nelson,  1976). 


Relationships 

The  higher  classification  of  the  Scorpaeniformes  remains  con- 
troversial and  uncertain,  both  in  terms  of  monophyly  and  in 
the  definition  of  families  and  their  relationships.  Confusion  ex- 
ists not  only  at  the  subordinal  levels,  but  also  at  lower  taxonomic 
levels.  For  example,  between  1  and  17  families  of  cottids  have 
been  recognized  by  previous  workers. 

Two  workers  presented  hypotheses  of  relationships  within  the 
Scorpaeniformes.  Matsubara  (1943),  in  a  detailed  study  of  Jap- 
anese scorpaenoids  based  on  osteological  and  anatomical  char- 
acters, briefly  treated  relationships  of  scorpaenoids  to  other  scor- 
paeniforms.  His  graphic  presentation  of  relationships  is  shown 
in  Figure  240.  Several  lineages  are  recognizable:  1)  the  Hexa- 
grammidae,  Anoplopomatidae,  and  "generalized"  scorpaenids; 
2)  Peristediidae,  Triglidae,  and  Dactylopteridae;  3)  "special- 
ized" scorpaenids,  Bembridae,  Platycephalidae,  and  Hoplich- 
thyidae;  4)  Cottidae  and  Agonidae;  and,  5)  Cyclopteridae  and 
Liparididae.  In  1955,  Matsubara  refined  his  hypothesis  of  re- 
lationships and  presented  a  classification  with  categories  equiv- 
alent to  three  suborders,  several  superfamiliesand  included  fam- 
ilies as  follows: 


Cottida 
Cottina 

Scorpaenicae 

Scorpaenidae,  Synanceiidae,  Congiopodidae 
Hexagrammicae 

Anoplopomatidae,  Hexagrammidae 
Plalycephalicae 
Parabembridae,    Bembridae,    Platycephalidae,    Hop- 
lichthyidae 
Cotticae 

Cottidae,  Psychrolutidae 
Agonicae 

Agonidae,  Aspidophoridae 
Triglicae 
Triglidae,  Peristediidae 
Dactylopterina 

Cephalacanthidae 
Cyclopteridae 

Cyclopteridae,  Liparididae 

Quast  (1965)  presented  a  notably  different  hypothesis  of  re- 


lationships of  the  scorpaeniforms.  His  work  was  based  on  char- 
acters which  were  useful  in  comparisons  with  the  hexagrammids 
and  included  many  characters  taken  from  the  earlier  works  of 
Gill  (1888),  Regan  (1913a)  and  Berg  (1940).  Quast  proposed  that 
the  Scorpaeniformes  included  three  basic  lineages:  1 )  the  cottid- 
hexagrammid  (including  the  Cyclopteridae  and  Agonidae);  2) 
the  anoplopomatid;  and,  3)  the  scorpaenoid  (including  all  other 
families).  Quast  (1965)  did  not  incorporate  his  recommended 
revisions  in  his  formal  synopsis  of  scorpaeniforms  because  he 
believed  that  the  cottoids  and  anoplopomatids  were  still  in  need 
of  intensive  study. 

Several  studies  of  particular  character  complexes  have  also 
contributed  to  understanding  of  relationships  within  the  order. 
Freihofer  (1963),  in  a  study  of  patterns  of  the  ramus  lateralis 
accessorius  and  associated  nerves  in  teleosts,  found  three  pat- 
terns of  nerves  in  scorpaeniforms  which  suggested  three  groups: 
1)  the  Scorpaenidae  and  Synanceidae;  2)  the  Hexagrammidae, 
Cottidae,  and  Liparididae;  and,  3)  the  Anoplopomatidae.  These 
groupings  seem  to  support  Quast's  hypothesis  of  relationships 
but  many  families  were  not  examined  by  Freihofer.  Hallacher 
(1974)  provided  a  summary  of  gasbladder  muscles  in  the  scor- 
pionfish  genus  Sebastes  and  included  observations  on  other 
scorpaeniforms.  Matsubara  (1943)  treated  this  feature  for  Jap- 
anese scorpaenoids.  Hallacher  recognized  four  states  of  the  ex- 
trinsic muscle  in  scorpaeniforms.  His  characters  were  based  on 
the  connections,  or  lack  of  connections,  of  this  muscle  between 
the  cranium,  pectoral  girdle,  vertebral  column,  and  the  gas- 
bladder.  His  observations  partially  supported  Matsubara's  hy- 
pothesis of  scorpaeniform  lineages. 

Scorpaeniform  fishes  have  been  considered  as  pre-perciforms 
or  as  perciform  derivatives  but  their  relationship  to  other  fishes 
remains  uncertain.  Many  workers  have  argued  that  the  Scor- 
paeniformes evolved  from  a  "generalized"  perciform  ancestor 
because  of  striking  similarities  in  general  body  form,  and  ana- 
tomical and  osteological  characters  of  generalized  scorpaenids 
and  perciforms  (Gill,  1888;  Regan,  1913a;  Taranets,  1941;  Mat- 
subara, 1943;  Gregory,  1959;  Quast,  1965;  Gosline,  1971;  Lau- 
der and  Liem,  1983).  Others  (Greenwood  et  al.,  1966;  Nelson, 
1976)  have  tentatively  placed  the  Scorpaeniformes  as  a  distinct 
pre-perciform  group  of  the  Acanthopterygians. 

As  previously  mentioned,  several  authors  have  suggested  that 
the  Scorpaeniformes  may  be  polyphyletic  and  hence,  derived 
from  several  groups.  Greenwood  et  al.  (1966)  noted  that  some 
scorpaeniforms  share  similarities  of  the  parietals  and  cheek 
muscles  with  cods,  while  others  share  similarities  with  toad- 
fishes,  and  still  others  with  perciformes.  Freihofer  (1970),  on 
the  basis  of  nerve  evidence,  suggested  that  gobiesocids  were 
related  to  cottoids,  especially  liparidids.  Although  several  au- 
thors have  suggested  that  the  suborbital  stay  may  have  evolved 
more  than  once  in  the  Scorpaeniformes,  little  consideration  has 
been  given  to  the  hypothesis  that  other  groups  of  fishes  may 
have  lost  the  suborbital  stay.  Within  the  Scorpaeniformes,  sev- 
eral groups  show  a  reduction  or  loss  of  the  suborbital  stay. 
Groups  which  have  lost  the  circumorbital  bones,  and  possibly 
a  suborbital  stay  (e.g.  gobiesocids,  callionymids,  lophiiforms 


438 


WASHINGTON  ET  AL.:  SCORPAENIFORMES 


439 


and  gobioids)  should  not  be  excluded  from  consideration  of 
relationships  to  some  scorpaeniform  groups. 

In  summary,  the  limits  of  the  order,  suborders,  families  and 
distribution  of  families  in  the  suborders  are  the  subject  of  con- 
siderable disagreement  among  current  workers.  These  problems 
will  not  be  resolved  without  a  worldwide  revision  of  the  order. 
At  this  point,  we  assume  that  the  order  is  monophyletic.  For 
the  purposes  of  summarizing  information  on  this  order,  we  treat 
two  broad  suborders:  the  Scorpaenoidei  and  the  Cottoidei.  We 
consider  these  groups  as  a  convenient  way  to  discuss  disagree- 
ments in  classification  of  specific  groups  and  hypotheses  of  re- 
lationships; we  do  not  propose  that  they  are  monophyletic  groups. 

Suborder  Scorpaenoidei 

For  this  paper,  we  recognize  the  Scorpaenoidei  to  include  the 
following  families:  Scorpaenidae  (broad  sense  of  Matsubara, 
1943),  Triglidae,  Peristediidae,  Bembridae,  Platycephalidae, 
Hoplichthyidae,  and  Dactylopteridae.  Some  of  these  families 
have  been  assigned  to  separate  suborders  or  superfamilies  and 
the  dactylopterids  have  often  been  placed  in  a  separate  order 
(Quasi,  1965;  Nelson,  1976;  Lauder  and  Liem,  1983). 

Meristic  features  and  approximate  number  of  species  for  in- 
cluded groups  are  provided  in  Table  111.  Data  have  been  drawn 
from  many  sources  and  may  not  be  complete  for  some  genera 
or  may  omit  extremes  found  in  abnormal  individuals. 

Matsubara's  work  (1943)  is  the  most  thorough  study  of  scor- 
paenoids  to  date.  His  hypothesis  of  relationships  (Figure  240) 
is  based  on  a  wide  variety  of  characters  including  those  of  the 
infraorbital  bones,  suspensorium,  hyoid  apparatus,  cranium, 
pectoral  girdle  and  gasbladder.  Matsubara  included  14  subfam- 
ilies in  his  family  Scorpaenidae.  He  recognized  three  large  ge- 
neric groups  or  lineages  within  the  scorpaenoids  which  he  la- 
beled: Sebasles-slem,  Scorpaena-siem  and  Cocotropus-stem.  His 
Sehasles-stem  contains  two  subfamilies,  the  Sebastinae  and 
Neosebastinae  which  were  viewed  as  the  most  primitive  or  "gen- 
eralized" of  the  scorpaenoids.  His  second  group,  the  Scorpaena- 
stem,  includes  five  subfamilies:  Scorpaeninae,  Pteroinae,  Setar- 
chinae,  Sebaslolobinae,  and  Plectrogeninae.  The  third  group, 
the  Cocotropus-ilem,  includes  six  subfamilies:  Apistinae,  Con- 
giopinae,  Aploactinae,  Minoinae.  Pelorinae,  and  Erosinae.  The 
latter  two  groups  were  considered  "specialized"  or  derived  rel- 
ative to  the  Sebastes-%\em.  Other  worJcers  (Greenwood  et  al., 
1966;  Nelson,  1976;  Poss  and  Eschmeyer,  1978)  have  departed 
from  Matsubara's  classification  of  the  Scorpaenidae  by  elevating 
some  subfamilies  of  Matsubara  to  family  status.  In  addition, 
the  Congiopodidae  [but  not  Matsubara's  Congiopinae  (sic)]  has 
been  recognized  as  a  separate  family  in  a  monotypic  suborder 
by  Greenwood  et  al.  (1966)  and  Nelson  (1976)  and  as  a  super- 
family  by  Quast  (1965).  Other  scorpaenoid  groups  not  treated 
by  Matsubara  (1943)  have  been  given  separate  status  within  the 
Scorpaenoidei  by  the  aforementioned  workers,  and  include  the 
Caracanthidae  and  Pataecidae.  In  his  later  work  on  fish  hier- 
archy, Matsubara  (1955)  recognized  three  families  of  scorpae- 
nids  which  basically  correspond  to  his  earlier  three  "stem"  groups. 

We  basically  follow  the  phylogenetic  hypotheses  of  Matsubara 
(1943,  1955)  in  presenting  general  trends  in  relationships  within 
the  suborder.  The  following  discussion  highlights  groups  where 
problems  or  disagreements  about  relationships  are  persistent. 
A  phylogenetic  approach  based  on  information  presented  here 
would  result  in  family  and  subfamily  lines  being  interpreted 
quite  differently.  However,  we  believe  presentation  of  a  new 


Fig.  240.     Schematic  representation  of  scorpaeniform  relationships 
from  Matsubara  ( 1 943). 


classification  is  premature;  a  thorough  study  of  the  scorpaenoids 
is  required  on  a  worldwide  basis. 

The  Sebastinae  is  currently  considered  to  be  the  most  prim- 
itive or  generalized  group  of  scorpaenoids  because  of  the  in- 
complete suborbital  stay  in  Sebastes,  weak  head  spination,  and 
general  body  plan  similar  to  the  percoids  (Matsubara,  1943  and 
others).  Matsubara  (1943)  proposed  that  Sebastes  was  the  most 
generalized  genus  within  the  subfamily  with  a  transition  series 
to  Helicolenus.  Eschmeyer  and  Hureau  (1971)  and  Barsukov 
(1973)  believed  that  Matsubara's  transition  series  is  reversed 
with  Helicolenus  the  most  generalized  genus  and  Sebastes  being 
a  relatively  derived  form. 

The  subfamily  Scorpaeninae  with  its  1 50  genera  is  considered 
a  "catch-basket"  subfamily,  and  there  is  no  certainty  that  it  is 
monophyletic. 

Matsubara  (1943)  noted  that  the  Setarchinae  lack  a  basi- 
sphenoid  as  do  cottoids  and  that  the  second  and  third  actinosts 
intervened  between  the  hypercoracoid  and  hypocoracoid.  He 
concluded  from  these  observations  that  the  Setarchinae  and 
cottoids  shared  a  common  ancestor.  However,  Eschmeyer  and 
Collette  (1966)  disagree.  In  their  review  of  the  Setarchinae,  a 
small  basisphenoid,  connected  only  by  cartilage,  was  found  in 
cleared  and  stained  specimens;  they  stated  that  Matsubara's 
conclusion  was  untenable. 

Matsubara  (1943)  suggested  that  the  genus  Sehastolobus  was 
closely  related  to  the  genus  Plectrogenium  (subfamily  Plectro- 
geninae) because  of  their  shared  lack  of  gasbladders,  notched 
pectoral  fins  and  prominent  rows  of  spines  along  the  sides  of 
their  head.  He  further  noted  (1943:160):  thai  "Plectrogenium 


440 


ONTOGENfY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  111.    Meristic  Features  for  Suborder  Scorpaenoide 

.  Parentheses  show 

rarer  counts; 

abnormal  specimens  not  i 

ncluded. 

Dorsal  fin  rays 

a 

nal  fin  rays 

[Genera 

Pectoral 

Pelvic 

Taxon 

Spe 

les 

Spines 

Sofl  rays 

Tolal 

Spines 

Soft  rays 

rays 

rays 

Vertebrae 

Scorpaenidae 

Sebastinae 

Helicolenus 

12 

(11)  12(13) 

11-14 

23-26 

3 

5-6 

(17)  18-20 

1  +  5 

25 

Hozukius 

2 

12 

11-12 

23-24 

3 

6 

18 

I  +  5 

(25)26 

Sehastiscus 

4 

(11)  12(13) 

11-13 

22-25 

3 

(4) 5 (6) 

17-19 

1  +  5 

25(26) 

Sebastes 

110 

(12) 13-16 

11-17 

24-30 

3 

5-10 

15-21 

I  +  5 

(25)26-31 

Scorpaeninae 

[15]  150 

12-13 

8-10 

21-23 

3 

5 

14-21 

I  +  5 

24 

Sebastolobinae 

Adelosebastes 

1 

13 

13 

26 

3 

5 

21 

I  +  5 

26 

Trachyscorpia 

2 

12-13 

8-9 

20-22 

3 

5 

20-24 

I  +  5 

25-26 

Sebastolobus 

2 

15-17 

8-10 

23-27 

3 

(4)  5  (6) 

20-24 

I  +  5 

27-30 

Plectrogeninae 

[1] 

2 

12 

(6)7 

(18)  19 

3 

5 

20-24 

I  +  5 

26 

Pteroinae 

[5] 

17 

12-13 

9-12 

22-24 

2-3 

5-9 

13-21 

I  +  5 

24 

Setarchinae 

[3] 

5 

(11)  12(13) 

9-10(11) 

21-23 

2-3 

5-6  (7) 

19-25 

I  +  5 

24 

Neosebastinae 

[2] 

12 

13 

(6)  7-8  (9) 

19-22 

3 

(5)6 

19-23 

I  +  5 

25-26 

Apistinae 

Apistus 

1 

-2 

14-16 

8-10 

23-25 

3 

6-8 

10-12  +  1 

I  +  5 

(25)26 

Cheroscorpaena 

1 

13 

7-9 

20-22 

3 

6-7 

9  +  3 

I  +  5 

26 

Minoinae 

[1] 

10 

8-11  (12) 

(8)9-14 

19-24 

2 

7-11 

11  +  1 

I  +  5 

(24) 25-27 

Choridactylinae 

Chohdactylus 

2 

12-15 

8-9(10) 

21-23(24) 

2 

8-10 

9  +  3 

I  +  5 

26-28 

Inimicus 

8 

15-18 

(5)  6-9 

23-26  (27) 

2 

8-13 

10  +  2 

I  +  5 

27-30 

Synanceiinae 

Synanceia 

4 

-5 

(12)  13-14(15) 

4-7 

18-20 

3 

4  (5*)-6  (7)* 

11-19 

1  +  4-5 

24 

Erosa 

1 

14 

5-6 

19-20 

3 

5-6 

14-16 

1  +  4 

(24)  25  (26) 

Dampierosa 

1 

12(?13) 

(8)9 

21 

2 

6 

12 

I  +  4 

— 

Pseudosyna  nceia 

1 

(15) 16-17 

4*-6* 

19*-21* 

3 

7*-8* 

14-15(16) 

I  +  3 

26-27 

Leptosynanceia 

1 

16 

5* 

21 

3-4 

5*-6* 

13-15 

I  +  4 

28 

Trachwephatus 

1 

(11)  12(13) 

12*- 14* 

24»-26* 

2 

12*-14* 

14-15 

I  +  5 

(28)29(30) 

Tetrarogidae' 

Ablabys 

3 

15-17 

7-11 

23-27 

3 

5-9 

11-13 

I  +  5 

26-28 

Centropogon 

3 

15-16 

7-9 

23-24 

(2)3 

5(6) 

13-15 

I  +  5 

26-27 

Coccolropsis 

1 

14-16 

5-6 

19-22 

3 

(3)4(5) 

11 

I  +  3 

25-27 

Cottapistus 

1 

13-15 

5-7 

18-21 

3 

5-6 

(13)  14(15) 

I  +  4 

(24)25 

Gtyptauchen 

1 

16-18 

6-7 

23-25 

3 

5 

13-15 

I  +  5 

26-28 

Gymnapisles 

1 

13-14 

7-9 

20-22 

3 

5-6 

11-12 

I  +  5 

27-29 

Liocramum 

1 

13-14 

6-9 

20-22 

3 

5-6 

13-15 

I  +  4 

24-25 

Neocentropogon 

4 

13-15 

7-8 

20-22 

3 

5-7 

13-16 

I  +  5 

25 

Notestes 

1 

14-16 

8-10 

22-25 

3 

5 

11-14 

I  +  5 

27-28 

Ocosia 

6 

14-17 

7-9 

23-24 

3 

5 

12-13 

I  +  5 

26-30 

Paracentropogon 

2 

-4 

13-15 

6-8 

20-22 

3 

(3)4(5) 

9-12 

I  +  4 

25-27 

Richardsonichthys 

1 

12-13 

5-8 

18-20 

3 

(5)  6-7 

14-16 

I  +  5 

24-25 

Sny derma 

2 

12-14 

10-11 

22-24 

3 

5-6 

13-15 

I  +  5 

24-28 

Tetraroge 

2 

13(14) 

6-9 

20-22 

3 

(4)5(6) 

11-12 

1  +  5 

24-26 

Vespicula 

ca 

5 

3  +  8-13 

3-8 

18-21 

3 

3-5 

10-14 

I  +  5 

24-26 

Paetaecidae' 

Aelapciis 

3 

19-22 

11-13 

30-34 

7-9 

4-5 

8 

0 

35-37 

Neopalaecus 

1 

19-23 

7-10 

29-31 

5-7 

3-4 

8 

0 

34-37 

Pataecus 

1 

22-25 

14-17 

38-40 

9-11 

4-7 

8 

0 

41-44 

Ganthanacanthidae' 

Gnathanacanthus 

1 

11-13 

9-11 

20-23 

3 

8-9 

10-12 

I  +  5 

28-30 

Congiopodidae' 

Alerlichthys 

1 

14-16 

10-13 

26-27 

2 

7-9 

9 

1  +  5 

30-31 

Congiopodus 

4? 

16-20 

11-14 

28-33 

0-2 

7-10 

(8)9 

I  +  5 

36-38 

Perryena 

1 

14-15 

8-9 

23-24 

3 

5-6 

12 

I  +  5 

28 

Zanchlorynchus 

1 

7-9 

12-14 

19-22 

0 

8-10 

8-9 

I  +  5 

36 

Aploactinidae- 

Acanthosphex 

1  + 

11-13 

7-11 

19-22 

1(2) 

6-8 

9-10 

I  +  2 

24-26 

Adventor 

1 

3+10 

7-9 

21-23 

1 

8-10 

12-14 

1  +  2 

27  (28) 

Aploactis 

1 

12-15 

11-15 

24-28 

1-3 

10-12 

11-14 

I  +  2 

28-30 

Aploactisoma 

1 

13-15 

12-16 

26-29 

1 

9-13 

10-11 

I  +  2 

30-33 

Bathyaploactis 

2 

14-15 

7-9 

21-23 

3-4 

5-9 

10-12 

I  +  2 

25-28 

Cocolropus 

10 

12-15 

7-12 

19-24 

1-2 

6-9 

11-14 

I  +  3 

25-28 

WASHINGTON  ET  AL.:  SCORPAENIFORMES 


441 


Table  111.    Continued. 


[Genera] 
Spec 

les 

Dorsal  tin  rays 

Anal  fin  rays 

Pecioral 
rays 

Pelvic 
rays 

Taxon 

Spines 

Soft  rays 

Total 

Spines 

Soft  rays 

Vertebrae 

Ensphe.x 

3  + 

10-12(13) 

9-16 

21-28 

1(2) 

9-15 

11-15 

1  +  (1)2 

27-31 

Eschmeyer*** 

1 

8 

13 

21 

3 

8 

19-20 

I  +  3 

24 

Kanekonia 

3 

11-13 

7-10 

20-22 

1(2) 

7-9 

13-16 

I  +  (1)2 

25-26 

Neaploactis 

1 

4  +  7+1 

9-10 

21-22 

1(2) 

7-9 

12 

I  +  3 

26 

Paraploaclis 

7 

12-15 

8-11 

22-24 

1(2) 

7-10 

13-15 

I  +  3 

26-28 

Perislrominous 

1 

12-13 

10-11 

22-24 

0-2 

7-10 

14-15 

I  +  3 

26-27 

Prosoproaus 

1 

12 

8 

20 

2 

7 

13 

I  +  3 

26 

Ptarmus 

2 

13-16 

7-10 

21-23 

2 

4-7 

9-10 

I  +  2 

25-30 

Sihenopus 

1 

3  +  9 

8-10 

20-22 

1 

7-9 

14-15 

I  +  2 

26 

Xenaploaclis 

3 

3  +  9-10 

8-9 

21-22 

1 

9-10 

14-15 

I  +  3 

27-28 

Caracanthidae 

[1] 

4 

6-8 

11-14 

18-23 

2 

11 

12-15 

I  +  2-3 

24 

Triglidae 

[ca.  10]  80 

7-11 

10-19 

18-26 

0-1 

11-18 

11-16  +  3 

I  +  5 

34-38 

Penstediidae 

[3]  40 

7-9 

16-23 

24-31 

0 

16-23 

+  2 

I  +  5 

Bembridae 

Parabembras 

1 

9;  11-12 

8-9 

18;  20 

3 

5 

21 

I  +  5 

26 

Bembradon 

1 

6 

14 

20 

0 

14*-15* 

23 

I  +  5 

Bembras 

1 

10-11 

12 

21 

0 

14*-15* 

I  +  5 

Bembradium 

1 

8-9 

12 

20-21 

0 

10*-11* 

24-27 

I  +  5 

26 

Platycephalidae 

[18]  60 

6-9(10) 

11-15 

18-23 

O-I 

10-14 

16-22 

I  +  5 

27 

Hoplichthyidae 

[1] 

10 

5-6 

14-16 

19-21 

0 

16-18 

13-14  +  3-4 

I  +  3-5 

26 

Dactylopteridae 

Dactylopterus 

1 

7 

8 

15 

0 

6* 

34-37 

I  +  4 

22 

Dactyloptena 

6 

7*-8** 

8(9) 

15(16) 

0 

6(7)* 

28-35 

I  +  4 

22 

•  Last  ray  single  (usual  condition  is  a  double  ray). 
"  I  +  0  +  V  +  I  or  I  +  I  +  V  +  I  =  7-8  spines. 
**"  Placement  uncertain. 
'  Data  supplied  by  Poss. 
-  From  Poss  (1982). 


nanuni  is  closely  related  to  the  bembrids,"  and  that  "it  is  very 
probable,  therefore,  that  the  platycephalids,  bembrids  and  hop- 
lichthyids  arose  from  an  ancestor  not  very  unlike  the  scorpaenid. 
Plntrogeniuin  naiiuin."  Matsubara  and  Ochiai  (1955)  present 
additional  characters  which  support  this  view.  Other  characters 
observed  by  one  of  us  (WNE)  which  support  this  conclusion 
include,  "similar  caudal  skeletons  and  scales"  [comparison  of 
bembrid  Parabembras  ciirtis  (SU  49456,  cleared  and  stained) 
and  Pleclrogeiiiu?}!].  At  present,  this  available  evidence  suggests 
that  the  scorpaenid  subfamilies  Sebastolobinae  and  Plectrogen- 
inae  and  the  families  Bembridae,  Platycephalidae  and  Hoplich- 
thyidae may  form  a  monophyletic  assemblage. 

Another  scorpaenid  subfamily,  the  Apistinae,  also  has  ques- 
tionable relationships  within  the  family.  Matsubara  ( 1 943)  placed 
Apislus  at  the  base  of  his  Cocolropiis  stem  which  led  to  a  number 
of  specialized  scorpaenid  groups.  However,  the  Apistinae  have 
a  "unique",  bilobed  gasbladder  with  an  intrinsic  muscle,  unlike 
other  scorpaenids  (Matsubara,  1943;  Hallacher,  1974).  The  trig- 
lids  and  peristediids  possess  a  similar  gasbladder.  Other  char- 
acters which  appear  to  unite  the  Apistinae,  Triglidae,  and  Per- 
istediidae  include  elongate  pectoral  fin  rays,  1  to  3  lower  free 
pectoral  fin  rays(l  in  Apistus.  3  in  Chcroscorpaena.  3  in  triglids, 
and  2  in  peristediids)  and  shape  and  expansion  of  the  head  bones 
(especially  the  infraorbidal  bones).  These  characters  suggest  that 
the  scaled,  less  bony-headed  Apistinae  may  be  the  primitive 
sister  group  of  the  Triglidae  and  Penstediidae.  This  would  in- 
volve the  independent  development  of  a  moveable,  preorbital 
bone  with  long  spine  in  the  Apistinae.  If  in  fact,  the  Apistinae 
forms  part  of  a  monophyletic  assemblage  with  the  triglids  and 
peristediids,  a  change  in  classification  would  be  warranted. 


Matsubara  (1943)  recognized  five  Japanese  genera  within  his 
subfamily  Congiopinae  (sic)  which  more  recently  have  been 
placed  in  the  tentative  scorpaenid  subfamily,  Tetraroginae  (Poss 
and  Eschmeyer.  1975;  see  also  Smith,  1958b).  The  presently 
recognized  Congiopodidae  is  considered  to  contain  7  to  9  species, 
all  of  which  are  confined  to  the  Southern  Hemisphere  (More- 
land,  1960;  Hureau,  1971).  Moreland  (1960:241)  slated:  "the 
Congiopodidae  show  relationship  with  the  Scorpaenidae,  par- 
ticularly with  Snyderina  and  Ocosia  from  Japan  [studied  by 
Matsubara  (1943)],  and  are  clearly  derived  from  a  scorpaenid 
stock  of  perhaps  Indo-Pacific  origin." 

We  tentatively  include  the  Dactylopteridae  in  our  discussion 
of  the  Scorpaenoidei,  however  relationships  of  these  fishes  re- 
main uncertain.  They  have  been  variously  placed  in  their  own 
order  (Regan,  1913a;  Berg,  1940;  Greenwood  et  al.,  1966;  Lau- 
der and  Liem,  1983)  and  as  a  suborder  of  the  Scorpaeniformes 
(Gill,  1888;  Nelson,  1976).  Many  workers  have  noted  that  the 
dactylopterids  differ  markedly  from  scorpaeniforms  in  a  number 
of  osteological  characters  such  as:  1)  nasals  fused  into  a  median 
plate;  2)  very  large  extrascapulars;  and,  3)  mesethmoid  and  in- 
tercalar  absent. 

Matsubara  (1943)  suggested  that  despite  these  notable  differ- 
ences, the  dactylopterids  possess  the  characteristic  suborbital 
arrangement  of  bones  of  the  generalized  scorpaenids  and  triglids. 
.•Accordingly,  Matsubara  (1943)  placed  them  near  the  triglids 
and  peristediids,  evolving  from  a  generalized  scorpaenid  ances- 
tor. One  of  us  (WNE)  has  noted  that  the  gasbladders  of  the 
triglids  and  dactylopterids  appear  to  be  similar,  with  anterior 
and  posterior  lobes  and  very  large  intrinsic  muscles  (sec  Evans, 
1973,  for  information  on  triglids).  In  dactylopterids  the  gas- 


442 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 
Table  1 12.    Meristic  Features  for  Suborder  Cottoidei.  Dashes  [— ]  indicate  data  not  available. 


Dorsal  fin 

A 

lal  tin 

No.  of 
species 

Pectoral 
fin 

Pelvic 
fin 

Vertebrae 

Genera 

Spines 

Rays 

Spines 

Rays 

References 

Agonidae 

Agonomalus 

4 

8-10 

5-8 

0 

11-12 

11-12 

1,2 

— 

Howe' 

Agonopsis 

3 

6-11 

6-9 

0 

7-12 

12-15 

1,2 

39-42 

Howe' 

Agonus 

7 

5-11 

5-14 

0 

5-17 

13-19 

1,2 

36 

Howe' 

Anoplagonus 

2 

0 

4-6 

0 

4-5 

8-12 

1,2 

41-45 

Howe' 

Aspidophoroides 

3 

0 

4-7 

0 

4-7 

9-16 

1,2 

38-40 

Howe' 

Balhyagonus 

4 

5-8 

5-8 

0 

5-9 

14-16 

1,2 

40-46 

Howe' 

Bolhragonus 

2 

2-5 

4-6 

0 

4-6 

10-12 

1,2 

31-38 

Howe' 

Brachyopsis 

2 

7-9 

7-9 

0 

10-14 

14-15 

1,2 

— 

Freeman,  1951 

Hypsagonus 

2 

7-11 

5-7 

0 

9-11 

12-14 

1,2 

36 

Howe' 

Ocella 

6 

7-13 

6-9 

0 

7-18 

14-18 

1,2 

33-39 

Howe' 

Odontopyxis 

1 

3-6 

5-7 

0 

5-7 

12-15 

1,2 

37-42 

Howe' 

Pallasina 

3 

5-9 

6-9 

0 

9-14 

10-13 

1,2 

45-47 

Howe' 

Percis 

2 

5-7 

5-9 

0 

7-9 

11-12 

1,2 

42 

Howe' 

Sarrilor 

2 

6-9 

5-8 

0 

6-8 

13-17 

1,2 

— 

Howe' 

Stellenna 

1 

6-8 

5-8 

0 

7-9 

16-19 

1,2 

34-37 

Howe' 

Tilesina 

2 

19-21 

7-10 

0 

25-28 

15-16 

1,2 

— 

Freeman,  1951 

Xeneretmus 

4 

5-9 

6-9 

0 

5-9 

12-17 

1,2 

39-43 

Howe' 

Anoplopomalidae 

.-1  noplopoma 

1 

17-30 

16-21 

III 

15-19 

I.  5 

61-66 

Richardson  and  Washington,  1980; 
Andriashev,  1954;  Miller  and  Lea, 
1972;  Howe' 

Erilepis 

1 

12-14 

16-20 

III 

11-14 

16-19 

1,5 

45-46 

Andriashev.  1955a 

Comephoindae 

Comephorus 

2 

6-9 

28-34 

0 

27-36 

10-15 

- 

48-50 

Taliev,  1955 

Cottidae 

Alcichthys 

1 

9-10 

14-17 

0 

13-16 

15-16 

1,2-3 

33-36 

Watanabe.  1960 

Antipodocottus 

2 

8 

14-15 

0 

11-12 

18-19 

1,2 

— 

Nelson,  1975 

*Archaulus 

1 

9-10 

28-29 

0 

22-23 

16 

1,3 

— 

Howe' 

Archistes 

1 

10 

23 

0 

18 

15-16 

1,3 

— 

Jordan  and  Gilbert,  1899 

Argyrocollus 

1 

8-9 

14-19 

0 

11-16 

13-14 

1.2-3 

35-36 

Watanabe,  1960 

*Anedieltichthys 

1 

7-9 

12-13 

0 

9-11 

21-23 

1,3 

— 

Howe' 

*Artedieltina 

1 

_ 

— 

— 

— 

— 

— 

— 

Artedielliscus 

1 

— 

— 

— 

— 

— 

— 

— 

Artediellus 

15 

6-9 

11-14 

0 

10-14 

20-24 

1,3 

28-30 

Howe';  Leim  and  Scott,  1966 

Arledius 

7 

7-10 

12-18 

0 

9-14 

13-17 

1.2-3 

30-35 

Howe  and  Richardson.  1978;  Wash- 
ington. 1981 

Ascelichthys 

1 

7-10 

17-19 

0 

13-16 

16-18 

0 

33-36 

Howe  and  Richardson.  1978 

Asemichthys 

1 

9-11 

14-16 

0 

15-16 

16-18 

1,3 

33-35 

Howe  and  Richardson.  1978 

Astrocollus 

2 

7-10 

12-14 

0 

10-12 

15-17 

1,2-4 

28-29 

Watanabe.  1960.  1976 

Bero 

1 

9-10 

15-16 

0 

13-15 

15-16 

1,2 

32-35 

Watanabe.  1960 

Blepsias 

2 

6-10 

20-26 

0 

18-22 

11-17 

1,3 

37-39 

Howe  and  Richardson,  1978; 
Watanabe,  1960 

Chitonolus 

1 

8-11 

14-17 

0 

14-17 

16-18 

1,2-3 

35-36 

Howe  and  Richardson,  1978 

Ctinocottus 

5 

7-10 

13-17 

0 

9-14 

12-15 

1,3 

31-35 

Howe  and  Richardson,  1978;  Wash- 
ington, 1981 

Cottiusculus 

3 

7-10 

11-15 

0 

9-15 

19-22 

1,3 

24-29 

Watanabe,  1960 

Coitus 

35± 

4-10 

14-23 

0 

10-18 

10-19 

1,2-5 

31-39 

Howe- 

*Crossias 

1 

8-10 

17-20 

0 

10-16 

14-16 

1,3 

— 

Soldatovand  Lindberg.  1930; 
Watanabe.  1960 

*Damma 

1 

8-10 

13 

0 

12 

16 

1,2 

_ 

Watanabe,  1960 

*Enophrys 

6 

7-9 

9-14 

0 

6-13 

15-19 

1.  3-4 

29-33 

Sandercock  and  Wilimovsky,  1968; 
Howe  and  Richardson,  1978 

Furcina 

2 

8-11 

15-20 

0 

13-18 

13-15 

1,2 

32-37 

Watanabe,  1960 

Gymnocanthus 

6 

9-12 

13-18 

0 

14-20 

15-21 

1,3 

33-40 

Howe  and  Richardson,  1978;  Leim 
and  Scott.  1966;  Watanabe.  1960; 
Wilson.  1973 

HemUepidotiis 

6 

8-11 

18-20 

0 

13-16 

14-17 

1,4 

35-37 

Howe  and  Richardson,  1978 

Hemitripierus 

3 

11-19 

11-14 

0 

12-15 

18-22 

1.3 

37-41 

Howe  and  Richardson,  1978; 
Leim  and  Scott,  1966 

Icelinus 

9 

8-12 

12-18 

0 

10-17 

15-19 

1,2 

33-39 

Howe  and  Richardson,  1978;  Yabe 
etal.  1980;  Peden.  1981 

*Icelm 

13 

7-10 

17-24 

0 

13-20 

17-20 

1,3 

37-44 

Howe' 

Jordania 

1 

17-18 

15-18 

0 

22-24 

13-15 

1,4-5 

46-48 

Howe  and  Richardson,  1978 

WASHINGTON  ET  AL.:  SCORPAENIFORMES 


443 


Table  112.    Continued. 


Dorsal  fin 

Anal  fin 

No.  or 

species 

Pectoral 
fin 

Pelvic 
fin 

Vertebrae 

Genera 

Spines 

Rays 

Spines 

Rays 

References 

Leiocottus 

1 

9-10 

16-17 

0 

15-20 

18 

1,3 

35-36 

Howe  and  Richardson,  1978 

Leptocollus 

1 

6-8 

15-20 

0 

15-20 

17-20 

1.4 

35-39 

Howe  and  Richardson,  1978 

Megalocottus 

2 

8-10 

12-15 

0 

11-13 

16-18 

1,3 

— 

Howe';  Soldatov  and  Lindberg. 
1930 

Mesocotlus 

1 

8-9 

14-15 

0 

10-12 

— 

1,3-4 

— 

Soldatov  and  Lindberg,  1930 

Microcottus 

1 

7-9 

12-14 

0 

10-12 

14-17 

1,3 

32-34 

Howe' 

Myoxocephalus 

18 

8-12 

10-20 

0 

8-16 

14-19 

1,3 

34-46 

Howe-;  Andriashev.  1954 

Naulichlhys 

3 

7-10 

19-30 

0 

14-21 

13-17 

1.3 

35-41 

Peden,  1970 

Ocynectes 

2 

9-13 

12-17 

0 

6-11 

13-15 

1,2 

29-31 

Watanabe,  1960 

Oligocollus 

4 

7-10 

15-20 

0 

9-15 

12-15 

1,3 

33-37 

Washington,  1981;  Howe  and  Rich- 
ardson, 1978 

Orthonopias 

1 

8-9 

15-18 

0 

12-15 

13-15 

1,3 

33-35 

Howe  and  Richardson,  1978 

Pancelinus 

1 

12-13 

19-20 

0 

23-24 

14-15 

1,5 

42 

Howe  and  Richardson,  1978 

Phallocollus 

1 

10-12 

22-24 

0 

22-25 

14-16 

1,3 

— 

Howe' 

Porocotlus 

6 

8-10 

13-18 

0 

11-18 

13-19 

1.3 

34-38 

Howe';  Andriashev.  1954; 
Watanabe.  1960 

Pseudoblennius 

6 

8-11 

15-21 

0 

12-18 

13-16 

1,2 

32-38 

Watanabe.  1960 

Radulinopsis 

7 

9-10 

14-15 

0 

14-15 

16-17 

1,3 

— 

Soldatov  and  Lindberg.  1930 

Radulinus 

5 

8-11 

20-23 

0 

21-25 

17-20 

1,3 

38-40 

Howe  and  Richardson,  1978 

Rhamphocotlus 

1 

7-9 

12-14 

0 

6-8 

14-16 

1.3-4 

26-28 

Howe  and  Richardson.  1978 

Ricuzenius 

2 

8-11 

14-20 

0 

10-19 

15-19 

1.2-3 

28-32 

Watanabe.  I960,  1976;  Jordan 
andStarks.  1904 

Scorpaenichthys 

1 

8-12 

15-19 

0 

11-14 

14-16 

1,4-5 

35-37 

Howe  and  Richardson,  1978 

Sigmisies 

2 

8-10 

19-26 

0 

14-20 

13-15 

1,3 

34-36 

Howe  and  Richardson,  1978 

Stelgislrum 

T 

8-9 

17-19 

0 

12-14 

14-16 

1.3 

36 

Howe' 

Slernias 

1 

10-11 

22-24 

0 

22-24 

16-18 

1.3 

44-46 

Howe  and  Richardson,  1978 

Sllegicottus 

1 

9 

19 

0 

17 

18 

1.3 

— 

Howe  and  Richardson,  1978 

Sllengis 

3 

7-11 

13-16 

0 

11-15 

11-20 

1,2 

29-35 

Watanabe.  1960 

Synchirus 

I 

8-10 

19-21 

0 

18-21 

21-24 

1,3 

38-39 

Howe  and  Richardson.  1978 

*Taurocollus 

-> 

11 

15-16 

0 

12-13 

19 

1,3 

_ 

Howe';  Taranetz.  1935 

Thecopterus 

1 

10 

14 

0 

11 

20 

1,2 

— 

Howe' 

Thyriscus 

1 

10 

21 

0 

17 

15 

1,3 

38-39 

Howe' 

Trachydermus 

1 

8 

18-19 

0 

16-17 

16-17 

1,4 

34-36 

Watanabe.  1960 

Tnglops 

9 

9-13 

20-32 

0 

19-32 

15-24 

1,3 

44-54 

Howe';  Andriashev.  1954; 
Watanabe,  1960;  Leim  and 
Scott,  1966 

Vellilor 

1 

10 

18-20 

0 

17-20 

13-15 

1,2 

36-39 

Watanabe,  1960 

Zeslicelus 

1-2 

5-7 

10-13 

0 

8-11 

19-21 

1,2-3 

25-26 

Howe  and  Richardson,  1978 

Cottocomephoridae 

Asprocoltus 

5 

5-8 

12-17 

0 

11-16 

13-17 

1,3-4 

30-34 

Taliev,  1955 

Ahyssocottus 

5 

3-7 

10-16 

0 

8-15 

12-18 

1,2-3 

31-34 

Taliev,  1955 

Bat  rachoco!  Ills 

4 

5-8 

14-19 

0 

10-15 

14-19 

1,3 

32-37 

Taliev,  1955 

Conmclla 

-) 

5-7 

13-17 

0 

11-13 

15-17 

1,3 

33-34 

Taliev,  1955 

Collocomcphorus 

2 

7-10 

17-21 

0 

19-22 

17-21 

1,4 

37-42 

Taliev,  1955 

Mclacotlus 

1 

7 

19 

0 

13 

16 

1,3 

— 

Tahev.  1955 

Paracottus 

4 

6-9 

15-20 

0 

12-22 

16-19 

1.4 

33-39 

Taliev,  1955 

Procottus 

1 

6-10 

18-21 

0 

12-16 

16-19 

1.3 

35-37 

Taliev,  1955 

Erenuniidae 

Ereunias 

1 

9-11 

12-15 

0 

11-13 

14-15 

0 

36-39 

Yabe.  1981;  Watanabe.  1960.  1976 

Marukawickthys 

1 

10 

12-15 

0 

11-12 

15 

1,4 

34-39 

Yabe,  1981;  Watanabe,  1960,  1976 

Hexagrammidae 

Hexagrammos 

6 

16-25 

18-26 

0-1" 

19-26 

17-21 

1,  5 

47-57 

Kendall  and  Vinter,  1984;  Washington 
and  Richardson,  MS 

Ophiodon 

1 

25-28 

19-21 

3** 

21-25 

16-18 

1,5 

56-59 

Kendall  and  Vinter,  1984;  Washington 
and  Richardson,  MS 

Pleurogrammus 

1 

21-24 

24-30 

0-1" 

23-28 

23-28 

1,5 

58-63 

Kendall  and  Vinter,  1984;  Washington 
and  Richardson,  MS 

Oxylehius 

1 

15-17 

13-16 

3-4** 

14-17 

14-17 

1,5 

36-40 

Kendall  and  Vinter,  1984;  Washington 
and  Richardson,  MS 

Zaniolepis 

2 

21-22 

11-12 

3" 

18-20 

14 

1,5 

40-43 

Kendall  and  Vinter,  1984;  Washington 

Normanichthyidae 
Normanichthys 


10-11 


11-12 


14-15        17-19 


I,  5 


and  Richardson,  MS 


36-37       Balbontin  and  Perez.  1980 


444 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  112.    Continued. 


No.  of 
species 

Dorsal  fin 

Ana!  fin 

Pecioral 
fin 

Pelvic 

lin 

Vertebrae 

Genera 

Spines 

Rays 

Spines 

Rays 

References 

Psychrolutidae 

Coltunculus 

4 

6-9 

13-17 

0 

10-14 

17-23 

1.2-3 

28-29 

Howe-;  Nelson,  1982 

Dasycollus 

1-2 

8-11 

13-16 

0 

12-16 

22-26 

1,3 

34-35 

Howe  and  Richardson.  1978; 
Nelson,  1982 

Ebinania 

5 

6-8 

15-18 

0 

11-14 

17-24 

1.3 

— 

Nelson,  1982 

Eurvmen 

1 

8 

21-23 

0 

15-17 

25-26 

1.3 

38 

Howe  and  Richardson.  1978 

*Maiacocoltus 

4(?) 

8-10 

12-15 

0 

9-13 

19-23 

1.3 

30-33 

Howe  and  Richardson.  1978;  Nelson, 
1982 

Neophrinichlhvs 

2 

7-12 

14-18 

0 

11-14 

23-26 

1.3 

31-34 

Nelson.  1977 

Psychrolutes 

6 

6-12 

13-20 

0 

12-15 

15-26 

1.3 

33-35 

Nelson.  1982;  Stein  and  Bond.  1978 

*  Taxonomic  status  is  not  agreed  upon  by  current  workers 
•*  No.  of  anal  spines  recognized  vanes  among  workers. 

'  Howe.  K.  Compilation  of  menstic  data  from  published  and  unpublished  sources,  available  from  Northwest  and  Alaska  Fishenes  Center.  NOAA-NMFS,  Seattle,  WA. 
-  Howe.  K..  unpublished  data. 


bladder  is  huge,  occupying  much  of  the  body  cavity  with  anterior 
lobes  reaching  near  the  rear  of  the  cranium.  A  more  thorough 
study  of  these  gasbladders  is  needed.  Dactylopterids.  in  relation 
to  triglids,  have:  1)  a  hinged  bony  connection  with  the  pre- 
opercle;  2)  much  heavier  and  more  elaborate  ossification  of  the 
cranium;  3)  first  three  vertebrae  elongate  and  modified;  and,  4) 
reduced  opercular  and  gill  openings.  Given  the  extreme  osteo- 
logical  modifications  of  these  fishes,  a  current  working  hypoth- 
esis (WNE)  is  that  the  Apistinae.  triglids,  peristediids  and  dac- 
tylopterids share  a  common  ancestry.  However,  more 
information  is  needed  before  any  formal  changes  are  proposed. 

Suborder  Cottoidei 

We  include  the  following  families  in  this  group:  Hexagram- 
midae,  Zaniolepidae.  Anoplopomatidae.  Cottidae  (broad  sense). 
Agonidae.  Cyclopteridae,  and  Liparididae.  The  limits  of  these 
families  and  subfamilies  are  not  well-defined  and  there  is  con- 
siderable lack  of  understanding  among  workers  in  defining  both 
family  limits  and  those  of  higher  categories  (see  Washington 
and  Richardson,  MS  for  review).  We  treat  these  diverse  groups 
together  in  order  to  facilitate  discussion  of  past  classifications, 
not  because  we  believe  they  necessarily  form  a  monophyletic 
assemblage. 

Meristic  features  and  approximate  number  of  species  for  in- 
cluded groups  are  provided  in  Table  1 1  2.  Data  have  been  com- 
piled from  many  sources  and  may  not  be  complete  for  some 
groups  and  may  omit  extremes  found  in  abnormal  individuals. 

Matsubara  ( 1 955).  in  a  thorough  treatment  of  Japanese  species, 
recognized:  1)  a  superfamily  "Hexagrammicae"  (including  An- 
oplopomatidae and  Hexagrammidae);  2)  a  superfamily  "Cot- 
ticae"  (including  Cottidae  with  subfamilies  and  Psychrolutidae); 
and,  3)  a  superfamily  "Agonicae"  (including  Agonidae  and  As- 
pidophoridae).  He  placed  the  cyclopterids  and  liparidids  in  a 
larger  division,  Cyclopterina. 

Quast  (1965),  in  a  treatment  which  focused  on  relationships 
of  hexagrammid  fishes,  followed  Regan  (1913a)  and  Berg  ( 1 940) 
in  recognizing  a  superfamily  Hexagrammoidae  and  a  superfam- 
ily Cottoidae.  He  separated  the  Cottoidae  from  the  Hexagram- 
moidae on  the  basis  of  four  characters:  1 )  lack  of  a  basisphenoid; 

2)  dentigerous  upper  pharyngeals  restricted  to  one  or  two  pairs; 

3)  pleural  ribs  absent  or  developed  on  only  a  few  posterior 
abdominal  vertebrae;  and,  4)  pectoral  interradial  foramina  small 
or  absent.  However,  Quast  proposed  that  the  hexagrammids 


and  cottoids  form  a  single  evolutionary  lineage  within  the  Scor- 
paeniformes  and  that  the  Anoplopomatidae  are  significantly 
distinct  from  both  the  hexagrammid-cottid  lineage  and  the  scor- 
paenid  lineage  to  warrant  separate  superfamily  status.  He  further 
suggested  that  the  zaniolepids  are  intermediate  between  the  hex- 
agrammids and  cottids. 

Other  workers  (Greenwood  et  al,,  1966;  Nelson,  1976)  have 
placed  the  hexagrammids.  anoplopomatids.  and  zaniolepids  to- 
gether in  the  suborder  Hexagrammoidei.  and  the  cottids  (broad 
sense  of  Washington  and  Richardson,  MS),  agonids,  and  cy- 
clopterids in  the  suborder  Cottoidei.  Hallacher  (1974)  found  a 
cranioclavical  (gasbladder)  muscle  present  in  the  zaniolepids, 
cottids  (broad  sense),  agonids,  and  cyclopterids.  In  contrast, 
Hexagrammos  was  found  to  have  the  scorpaenoid  condition. 

In  the  following  discussion,  we  present  information  about 
recent  studies  which  have  helped  resolve  relationships  within 
cottoid  subgroups  and  outline  groups  where  problems  remain. 

The  systematic  status  of  the  Hexagrammidae  is  the  subject 
of  disagreement  at  the  specific  through  family  levels.  Quast  (1965) 
and  Nelson  (1976)  include  four  genera  in  the  Hexagrammidae— 
Oxylehiits.  Ophiodon,  Hexagrammos  (inc\u(i\n%  Agramnms),  and 
Pleurogramfuiis.  Quast  considered  Oxylehiits  to  be  the  most 
primitive  genus  because  of  low  numbers  of  meristic  elements 
and  the  "lack  of  specializations."  Hexagrammos  and  Pleiiro- 
grammus  were  considered  to  be  closely  related,  relatively  spe- 
cialized genera  because  of  the  reduction  in  head  spination.  dorsal 
and  anal  fin  spines,  etc. 

Quast  (1965)  and  Nelson  (1976)  included  the  two  species  of 
the  genus  Zaniolepis  in  the  family  Zaniolepidae.  Other  workers 
(Rutenberg,  1962)  have  included  Zaniolepis  in  the  family  Hex- 
agrammidae. while  others  (Hart,  1973)  have  combined  Zani- 
olepis and  Oxylehius  in  the  family  Zaniolepidae. 

The  Anoplopomatidae  contains  two  monotypic  genera,  An- 
oplopoma  and  Erilepis  (Quast,  1965;  Nelson.  1976),  however 
some  workers  have  placed  Erilepis  in  its  own  family,  the  Eri- 
lepidae. 

Those  families  that  traditionally  have  been  placed  in  or  near 
the  Cottidae  are  not  clearly  defined.  Previous  workers  have 
proposed  between  1  and  1 7  families  of  "cottids."  Greenwood 
et  al.  (1966)  and  Nelson  (1976)  recognize  7  cottid  families: 
Cottidae,  Icelidae.  Cottocomephoridae.  Comephoridae.  Cot- 
tunculidae,  Psychrolutidae  and  Normanichthy  idae.  Other  work- 
ers have  chosen  to  combine  these  7  families  in  the  single  family 


WASHINGTON  ET  AL.:  SCORPAENIFORMES 


445 


Cottidae,  until  further  study  can  define  the  phylogenetic  rela- 
tionships or  monophyletic  nature  of  these  groups  (Howe  and 
Richardson,  1978;  Washington  and  Richardson,  MS). 

Yabe  (1981)  recognized  the  family  Ereuniidae  for  the  Jap- 
anese "cottid"  genera  Ereunias  and  Marukawichthys.  He  used 
derived  characters  such  as  free  pectoral  fin  rays  and  associated 
pectoral  giidle  modifications  to  define  the  family.  Yabe  con- 
cluded that  the  genus  hclus  belonged  in  the  Cottidae.  Previous 
workers  (Matsubara,  1936;  Berg.  1940;  Nelson,  1976)  have  placed 
Marukawichihys  and  Ereunias  in  the  family  Icelidae  with  mem- 
bers of  the  genus  Icelus. 

Nelson  (1982)  has  revised  the  "family'  Psychrolutidae  which 
includes  two  subfamilies  (Psychrolutinae  and  Cottunculinae). 
Nelson  could  not  define  the  family  as  monophyletic  on  the  basis 
of  unique,  derived  characters  and  stated  that  the  question  of 
whether  to  include  the  psychrolutids  in  the  Cottidae  was  sub- 
jective at  this  time.  He  rejected  a  close  affinity  between  the 
psychrolutids  and  liparidids  as  suggested  by  early  workers. 

The  families  Comephoridae  and  Cottocomephoridae  are  en- 
demic to  the  Lake  Baikal  basin  (U.S.S.R.).  Berg  (1940)  recog- 
nized each  as  separate  families  within  the  superfamily  Cotto- 
idae.  Taliev  (1955),  after  detailed  study  of  the  two  groups, 
suggested  that  they  had  originated  from  cottid  ancestors  and 
cited  as  evidence  their  similarities  to  two  cottid  genera,  Meso- 
cottus  and  Trachydenmis.  Both  Taliev  (1955)  and  Kozhov  ( 1 963) 
placed  the  cottocomephorids  in  the  Cottidae  while  the  vivipa- 
rous comephorids  were  recognized  as  a  separate  family. 

The  family  Agonidae  has  been  reviewed  only  by  Freeman 
(1951)  who  suggested  that  the  agonids  were  most  closely  related 
to  the  cottids.  The  family  is  distinct  in  having  fused,  bony  plates 
covering  the  body. 

Nonnanichthys  crocken.  the  sole  member  of  the  Normanich- 
thyidae,  occurs  off  the  coasts  of  Peru  and  Chile.  Its  relationships 
are  obscure.  Norman  (1938b)  considered  it  to  be  a  primitive 
cottid,  while  others  (Berg,  1940;  Quast,  1965)  have  placed  it  in 
its  own  family,  in  the  superfamily  Cottoidae.  In  addition  to  a 
different  body  plan,  the  suborbital  stay  of  Nonnanichthys  is 
quite  distinct  from  other  scorpaeniform  fishes  (Poss,  1975).  Its 
relationships  to  cottoids  have  yet  to  be  established. 

relatronships  based  on 
Larval  Characters 

Larvae  of  only  about  20%  of  the  sc&rpaeniform  genera  are 
known,  and  only  recently  have  larvae  been  used  in  systematic 
studies  (see  Richardson,  1981a;  Washington,  1981;  Kendall  and 
Vinter,  1984;  Washington  and  Richardson,  MS).  The  most  ex- 
tensive information  dealing  with  systematic  characters  of  scor- 
paeniform larvae  is  presented  in  a  recent  study  by  Washington 
and  Richardson  (MS).  This  work  dealt  with  over  100  osteolog- 
ical  characters  of  larval  and  juvenile  cottids  and  their  allies. 
About  half  of  the  70  characters  used  in  their  analysis  were  re- 
stricted to  the  larval  period.  In  general,  larval  characters  were 
most  useful  in  defining  groups  below  the  subordinal  level. 

Larvae  of  many  scorpaenoid  families  are  not  yet  known.  Char- 
acters such  as  head  and  preopercular  spination  and  pectoral  fin 
length  and  pigmentation  may  be  useful  in  future  systematic 
analyses;  however,  at  present,  larvae  of  too  few  taxa  are  known 
to  suggest  relationships  within  the  suborder  Scorpaenoidei. 

The  results  of  Washington  and  Richardson's  (MS)  study,  agree 
with  those  of  past  studies  which  propose  that  a  scorpaenid-like 
stock  was  ancestral  to  the  Scorpaeniformes  and  was  derived 


from  a  "generalized"  perciform.  Larvae  ofthescorpaenid  genera 
Sebastes.  Sebastolobus,  and  Scorpaena  possess  some  characters 
which  are  but  slight  modifications  of  those  possessed  by  some 
generalized  percoids.  In  contrast,  other  scorpaeniform  larvae 
examined  possessed  considerable  modifications  of  these  char- 
acters. These  generalized  scorpaenid  characters  include  among 
others:  presence  of  predorsal  bones;  large,  fused  first  anal  pte- 
rygiophore  with  three,  stout  anal  spines;  pleural  ribs  on  abdom- 
inal vertebrae;  epipleurals  attached  to  pleural  ribs;  hypurals  1  -I- 
2  partially  fused;  hypurals  3-1-4  partially  fused;  presence  of  a 
fifth  hypural  and  parhypural;  all  hypural  elements  autogenous 
and  a  specialized  neural  spine  on  preural  centrum  2.  Without 
the  suborbital  stay,  larvae  of  a  scorpaenid  such  as  Sebastes  cou\d. 
easily  be  mistaken  for  those  of  a  generalized  percoid.  We  con- 
sider these  character  states  to  represent  the  plesiomorphic  con- 
dition in  the  Scorpaeniformes. 

Washington  and  Richardson's  study  focused  in  detail  on  cot- 
toid  and  hexagrammoid  fishes  where  larvae  of  many  taxa  are 
fairly  well  known.  They  found  that  the  hexagrammoids  exhibit 
many  character  complexes  which  are  derived  relative  to  the 
scorpaenids.  These  include:  1 )  reduced  anal  spines  and  first  anal 
pterygiophore;  2)  the  pleural  and  epipleural  ribs  inserted  to- 
gether on  the  vertebral  parapophyses;  and,  3)  the  pectoral  radials 
broadened  and  anvil-shaped,  but  with  distinct  foramina  be- 
tween them.  None  of  these  characters  is  unique  to  the  larval 
period. 

Within  the  taxa  traditionally  assigned  to  the  Hexagrammoidei 
(Nelson,  1976),  two  monophyletic  groups  are  recognized  by 
Washington  and  Richardson  (MS).  The  first  includes  the  hex- 
agrammid  genera  Hexagrammos,  Pleurogrammus.  and  Ophio- 
don  and  the  anoplopomatid  genus  Anoplopoma.  This  group  is 
defined  by  seven  autopomorphies:  1)  reduced  head  spination; 
2)  prolonged  chondrification;  3)  a  unique  (within  Scorpaeni- 
formes) sequence  of  ossification  of  the  vertebral  centra;  4)  paired 
first  dorsal  fin  elements;  5)  five  preural  centra  involved  in  caudal 
fin  support;  6)  anterior  insertion  of  principal  caudal  rays;  and, 
7)  a  high  number  of  vertebrae  and  ribs.  The  first  four  characters 
are  restricted  to  the  larval  period. 

In  contrast,  larvae  of  the  second  group,  Oxylcbius  and  Zan- 
iolcpis,  do  not  possess  any  of  the  synapomorphies  of  the  first 
group.  They  do  share  one  derived  character— an  unfused  neural 
arch  and  spine  of  the  first  vertebral  centrum.  The  arms  of  the 
first  neural  "arch"  and  spine  remain  unfused  for  a  brief  time 
during  larval  development,  a  unique  condition  among  known 
scorpaeniform  larvae.  Other  larval  characters  support  the  sep- 
aration of  these  groups,  but  we  are  cautious  in  the  interpretation 
of  these  characters.  They  include:  1)  large  versus  small  size  at 
hatching;  2)  neustonic  versus  planktonic  larvae;  and  3)  long, 
slender  versus  deep  body  shape. 

Washington  and  Richardson  (MS)  concluded  that  the  first 
group  of  hexagrammoids  is  very  distinctive  and  differs  from  all 
other  scorpaeniforms  so  far  examined,  particularly  in  the  mode 
of  ossification  of  the  vertebral  column  and  in  the  number  of 
preural  centra  involved  in  the  caudal  fin  support.  Because  of 
the  uniqueness  of  these  characters,  Washington  and  Richardson 
(MS)  suggest  that  members  of  this  hexagrammoid  group  prob- 
ably comprise  a  separate  lineage  within  the  order,  distinct  from 
Oxylebius  and  Zaniolepis  and  the  other  cottoids.  The  second 
group,  Oxylebius  and  Zaniolepis.  is  distinctive  but  appears  to 
be  closer  in  many  characters  to  the  scorpaenids  than  to  other 
hexagrammoids. 


446 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


<:^^ 


Fig.  241.     Hypothesis  of  cottoid  relationships  modified  from  Washington  and  Richardson  (MS). 


Washington  and  Richardson's  (MS)  hypothesis  of  relation- 
ships among  the  other  cottoids  studied  is  shown  in  Fig.  241. 
Characters  observed  in  the  cottoid  families,  Cottidae  (broad 
sense  of  Washington  and  Richardson,  MS),  Agonidae,  and  Cy- 
clopteridae,  are  derived  relative  to  both  the  scorpaenids  and 
hexagrammids.  The  cottids,  agonids,  and  cyclopterids  share  four 
apomorphic  characters,  none  of  which  is  restricted  to  the  larval 
period.  These  include:  1)  pleural  ribs  absent  or  restricted  to  the 
posterior  three  abdominal  vertebrae;  2)  epipleurals  independent 
or  sessile;  3)  small  first  anal  pterygiophore;  and.  4)  no  anal 
spines. 

The  cyclopterids  (including  lipandids)  appear  to  be  a  distinct 
family  defined  by  a  modified  ventral  sucking  disc  and  are  the 
sister  group  of  the  cottids  and  agonids.  (See  Able,  Markle  and 
Fahay,  this  volume,  for  discussion  of  cyclopterid  relationships). 

The  cottids  and  agonids  share  three  derived  characters:  1 )  the 
first  anal  pterygiophore  is  simple;  2)  there  are  no  supernumerary 
anal  elements;  and  3)  the  haemal  spine  of  preural  centrum  2  is 
enlarged.  Again,  none  of  these  characters  is  unique  to  the  larval 
period. 

Among  the  28  genera  of  cottids  examined,  Washington  and 
Richardson  (MS)  recognized  eight  monophyletic  groups  which 
are  defined  by  one  or  more  apomorphic  characters.  Rhampho- 
cottiis,  a  monotypic  genus,  is  characterized  by  four  distinctive 
autapomorphies,  two  of  which  are  larval  characters.  Rhanipho- 
cottus  larvae  possess  a  unique  body  shape  with  an  extremely 
long  snout  to  anus  length  (>60%  SL)  and  deep  body  shape  (29- 
40%  SL).  Rhamphocottus  larvae  also  possess  only  one  preoper- 
cular  spine.  Other  workers  have  also  found  Rhamphocoitits  to 


deviate  from  other  cottids  and  have  placed  it  in  its  own  family 
(Gill,  1888;  Johnson,  19 18;  Jordan,  1923;Bohn,  1934;Taranets, 
1941). 

Hemtlepidotus  and  Scorpaenichthys  form  another  cottid  group 
and  are  defined  by  five  autapomorphies,  three  of  which  are 
unique  to  the  larval  period.  First,  members  of  both  genera  de- 
velop heavy,  pitted  dermal  bone  on  the  cranium  which  forms 
early  in  larval  development.  As  the  bone  develops,  ossification 
proceeds  unevenly  with  small  pockets  of  bone  apparently  re- 
sorbed  forming  pitted  areas,  while  surrounding  areas  are  thick- 
ened. Second,  larvae  develop  broad  supraocular  bony  shelves 
which  project  laterally  over  the  orbit.  Third,  the  dorsalmost 
radial  of  the  pectoral  fin  is  reduced  in  size  and  becomes  fused 
or  nearly  fused  to  the  scapula  during  larval  development. 

These  three  characters  are  not  present  in  any  other  cottids 
examined.  Although  both  Scorpaenichthys  and  Hcimlepidotus 
have  been  postulated  as  "primitive"  cottids  by  workers  studying 
adults,  they  have  not  previously  been  considered  closely  related 
to  each  other. 

The  remaining  cottids  and  agonids  share  four  additional  de- 
rived characters:  1)  neural  spine  of  PU  2  elongate;  2)  neural 
spine  of  first  vertebra  absent;  3)  upper  and  lower  hypural  plates 
fused  to  each  other  and  fused  to  the  urostyle;  and.  4)  the  first 
neural  arch  is  unfused,  rather  it  forms  in  a  broad  U-shape.  The 
last  character  is  a  larval  feature  found  only  in  these  taxa. 

Five  additional  generic  groups  are  defined  by  one  to  six  au- 
tapomorphies. Although  these  five  groups  contain  the  majority 
of  cottid  genera,  no  synapomorphies  were  found  which  united 
these  groups  and  yet  separated  them  from  the  agonids.  The 


WASHINGTON  ET  AL.:  SCORPAENIFORMES 


447 


Myoxoccpha/us  group  includes  13  genera  defined  by  the  unique 
larval  character  of  a  bony  shelf  on  the  anterior  portion  of  the 
preopercle.  The  Artedius  group  includes  Clinocoiiiis.  Oligocot- 
liis.  and  Artedius  Group  A  (see  Washington,  1981).  This  group 
is  defined  by  six  autapomorphic  characters  including  three  unique 
larval  features:  1)  multiple  preopercular  spines;  2)  enlargement 
and  expansion  of  the  anterior  neural  arches;  and,  3)  first  three 
neural  arches  unfused.  The  Psychrolutes  group  includes  Gilhert- 
idia  and  Psychrolutes  and  is  defined  by  six  apomorphic  char- 
acters. Only  one,  the  absence  of  head  and  preopercular  spines, 
is  unique  to  the  larval  period.  The  Malacocottus  group  includes 
Dasycottiis  and  Malacocottus  and  is  defined  by  heavy,  bony 
arches  on  the  cranium  which  form  late  in  larval  development. 
Members  of  the  last  two  groups  were  recently  combined  in  the 
family  Psychrolutidae  (Nelson,  1982)  and  correspond  to  his 
subfamilies  Psychrolutinae  and  Cottunculinae,  respectively.  The 
Coitus  group,  including  Cottus  and  Leptocottus.  is  defined  by 
four  aulapomorphies,  two  of  which  are  larval  characteristics: 
the  first  proximal  dorsal  pterygiophore  is  simple  and  slender  in 
contrast  to  all  other  cottid  larvae  and  the  parhypural  is  absent 
in  larvae  of  these  genera.  Further,  larvae  of  these  genera  exhibit 
a  delay  in  ossification  of  the  cranium  and  reduced  head  spi- 
nation. 

The  last  two  cottoid  groups  are:  the  Hemitripterus  group  in- 
cluding the  "cottids"  Hemitripterus,  Nautichthys.  and  Blepsias. 
and  the  agonids.  These  share  three  derived  characters:  1 )  mod- 
ified prickle-scales;  2)  a  knobby  fronto-parietal  ridge;  and,  3) 


broad  plate-like  epurals.  The  first  two  characters  are  unique 
larval  features  of  this  group. 

These  characters  provide  evidence  that  the  Hemitripterus 
group,  traditionally  placed  in  the  Cottidae,  may  be  the  sister 
group  of  the  Agonidae.  Several  agonid  genera,  such  as  Hypsa- 
gomts  and  Agonomelas  are  very  similar  to  members  of  the 
Hemitripterus  group  both  as  larvae  and  adults.  In  addition, 
larvae  of  these  genera  share  several  apparently  derived  char- 
acters. However,  the  agonids,  including  Hypsagonus  and  Ago- 
nomalus  share  several  autapomorphies  unique  to  the  agonids 
mcluding  one  or  two  plate-like  epurals,  and  extreme  modifi- 
cations of  the  pectoral  girdle. 

The  implications  of  these  findings  are  that  the  agonids  are 
derived  from  the  cottids  and  according  to  cladistic  methodology 
should  be  relegated  to  a  sub-unit  of  the  Cottidae.  However, 
Washington  and  Richardson  (MS)  do  not  propose  any  formal 
changes  in  the  cottids  and  agonids  at  this  time.  Larvae  of  only 
about  a  third  of  the  cottid  genera  have  been  studied.  In  addition, 
the  family  or  families  of  cottids  have  not  been  clearly  defined 
on  the  basis  of  derived  characters,  and  until  such  time,  we  cannot 
hope  to  fully  understand  the  cottid-agonid  interrelationships. 

(B.B.W.,  K.M.H.)  Gulf  Coast  Research  Laboratory,  East 
Beach  Drive,  Ocean  Springs,  Mississippi  39564;  (W.N.E.) 
Department  of  Ichthyology,  California  Academy  of 
Sciences,  Golden  Gate  Park,  San  Francisco,  Califor- 
nia 94118. 


Tetraodontoidei:  Development 
J.  M.  Leis 


THE  tetraodontoid  fishes  (Gymnodontes)  are  a  diverse  sub- 
order of  one  large  and  three  small  families  and  about  150 
recent  species  (Winterbottom,  1974a;  Tyler,  1980).  The  four 
families  (Table  1 13)  are  largely  tropical,  but  many  species  are 
temperate.  Most  species  are  marine  and  bottom-associated  in 
shallow  waters,  but  the  Molidae  is  entirely  pelagic  and  both  the 
Diodontidae  and  Tetraodontidae  have  fully  pelagic  species.  The 
Tetraodonlidae  also  includes  a  number  of  fully  freshwater  species. 
Many  tetraodontoids  have  a  pelagic,  often  oceanic,  juvenile 
stage. 

Development 

Development  of  tetraodontoid  fishes  is  not  particularly  well- 
known.  Previous  reviews  of  the  early  development  of  the  group 
are  by  Breder  and  Clark  (1947),  Tortonese  (1956)  and  Martin 
and  Drewry  ( 1978).  Early  development  of  triodontids  is  entirely 
unknown,  and,  overall,  information  is  available  for  only  36 
species.  The  information  available  for  particular  species  of  these 
36  is  often  scanty.  However,  for  the  Molidae,  information  is 
available  for  all  three  species.  Complete  (i.e.,  egg  to  juvenile) 
information  is  available  for  about  10  species  (Table  1 14).  In  the 
following  sections,  I  assume  that  the  few  taxa  for  which  infor- 


mation is  available  are  representative  (these  taxa  and  the  de- 
velopmental stages  concerned  are  listed  in  Table  1 14).  The  fol- 
lowing sections  should  be  read  in  conjunction  with  Table  1 14; 
citations  listed  in  Table  114  are  not  repeated  in  the  text.  In 
parentheses  after  the  family  heading  I  give  the  number  of  species 
for  which  some  information  is  available. 

On  the  basis  of  early  life  history  characters,  the  tetraodontoid 
fishes  are  a  more  coherent  group  than  the  balistoid  fishes. 


Table  113.     Merlstic  Characters  of  Tetraodontoid  Fishes 

Principally  after  Tyler  (1980).  N  is  the  approximate  number  of 

recent  species  largely  after  Nelson  (1976).  Pelvic  fins  are  lacking  in  this 

suborder. 


Family 


N 


C 


Vene- 
brae 


Triodontidae 
Tetraodontidae 
Diodontidae 
Molidae 


I 

130 

15 

3 


0-n,  1 1 

7-34 
10-18 
15-20 


10 

7-27 
10-18 
14-18 


15-16 
12-20 
18-25 

7-13 


12 
II 

9-10 
12-26* 


20 
16-30 
18-23 
16-18 


*  Not  a  true  caudal  fin.  but  a  clavus  or  pseudocaudal. 


448 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  1 14,    Tetraodontoid  Taxa  for  whk  h  Information  is  Available  on  Egg  and  Larval  Stages.  YS  — yolk-sac  stage;  pre  — preflexion 

stage;  flex  — flexion  stage;  post  — postflexion  stage;  U  — unstated;  den- demersal;  pel  — pelagic;  PS— examined  for  the  present  study.  Numbers  in 

parentheses  after  each  genus  refer  to  the  number  of  species  represented.  A  blank  means  no  information  available  on  that  stage. 


Eggs 


Larvae 
developmental  stage 


Type 


Oil 
droplets 


Tetraodontidae 

Canthigasler  ( >  1 ) 

dem 

0.68-0.72 

cluster 

x 

Carmoletraodon  ( 1 ) 

U 

0.78 

U 

Chelonodon  ( 1 ) 

dem 

U 

U 

Chonerhinos  (5) 

U 

1.1-2.3 

U 

Fugu  (7) 

dem 

0.85-1.32 

cluster 

x 

Lagocephalus  ( 1 ) 

dem 

0.61-0.70 

cluster 

X 

Sphoeroidcs  ( 1 ) 

dem 

0.85-0.91 

cluster 

X 

Tetraodon  ( 1 ) 

dem 

1.0 

U 

Torquigener  (2) 

dem? 

0.94 

U 

X 

Unidentified  (7) 

X 

Diodontidae 

Allomvcterus 

pel 

2.0-2.2 

20-25 

Chilomycterus  (3) 

pel 

U-1.8 

U 

Diodon  (3) 

Tragiilichlhys  ( 1 ) 
Molidae- 
Masliirus  ( 1 ) 
Mola  ( 1 ) 
Ranzania  ( 1 ) 


pel 


1.62-2.1 


cluster 


pel  1.8  cluster 

pel  1.42-1.68  cluster 


X  Fujita,  1962;  Stroud  et  al.,  MS;  PS 

Breder  and  Rosen,  1966 

Breder  and  Rosen,  1966 

Roberts,  1982a 
X  Uchida  et  al.,  1958;  Mito,  1966;  Masuda  et 

al.,  1975;  Fujita.  1962 
X  Uchida  et  al.,  1958;  Fujita,  1962.  1966 

X  Welsh  and  Breder.  1921 

Breder  and  Rosen,  1 966 
X  Munro,  1945 

X  Miller  et  al.,  1979;  Leis  and  Rcnnis.  1983; 

PS 


Robertson,  1975a 

Evermann  and  Kendall,  1898;  Breder,  1927; 
Nichols  and  Breder,  1927;  Fowler,  1945; 
Breder  and  Clark,  1947;  Heck  and  Wein- 
stein,  1978;  Moyer,  pers.  comm.;  Fujita. 
1962;  PS 

Sanzo,  1930d';  Mito.  1966;  Leis,  1978;  Sak- 
amoto and  Suzuki,  1978;  Fujita,  1962;  PS 

PS 


X  Schmidt,  1921;  Martin  and  Drewry,  1978 

X  Martin  and  Drewry,  1978 

X  Schmidt,  1921;  Leis.  1977 


'  Misideniified  as  Crayracion  sp.  (Tetraodontidae). 

■  No  caudal  fin  forms:  pre,  fiex  and  post  in  this  case  refer  to  clavus  formation,  not  notochord  lleMon, 


£gg.j.— Tetraodontoid  fishes  are  oviparous.  Pelagic  and  demer- 
sal eggs  are  known;  the  chorion  is  smooth;  clusters  of  oil  droplets 
are  present;  eggs  range  in  size  from  large  (2. 1  mm)  to  small  (0.6 
mm)  and  are  spherical;  incubation  times  are  long  and  range 
from  3  to  20  days;  development  at  hatching  varies;  the  peri- 
vitelline  space  is  narrow;  the  yolk  is  unsegmented;  and  embryos 
may  be  heavily  pigmented.  Parental  care  of  eggs  is  present  only 
in  some  tetraodontids. 

Larvae.  — M\  tetraodontoid  larvae  are  pelagic.  Development  in 
most  tetraodontids  is  direct;  in  molids  and  diodontids  special- 
ized ontogenetic  stages  may  exist.  There  are  few  larval  special- 
izations except  in  the  Molidae,  and  development  is  usually  com- 
pleted at  a  small  size.  There  is  often  an  apparently  unspecialized 
pelagic  juvenile  stage,  which  may  be  very  large  at  settlement. 
Larvae  are  enclosed  in  a  more  or  less  inflated  vesicular  sac. 
Larvae  are  deep  and  wide  in  head  and  trunk,  and  the  tail  is 
comparatively  small  and  compressed.  The  head  is  large  and 
rounded  and  the  gut  is  coiled  and  massive.  The  eye  is  partic- 
ularly large.  The  specialized  adult  scales  form  directly  (i.e.,  do 
not  pass  through  an  unspecialized  spinule  stage).  In  molids  spe- 
cialized larval  spines  are  formed.  The  pectoral  fin  is  the  first  to 
form,  and  the  caudal  fin  is  last.  Except  for  the  tail  of  molids, 
structures  are  not  formed  and  subsequently  lost  — they  never 
form.  The  specialized  dentition  develops  during  the  larval  stage 
directly,  without  any  intervening  generalized  teeth.  However, 
diodontids  and  tetraodontids  may  have  small,  raised  points 


along  the  cutting  edges  of  their  beak-like  teeth.  Meristic  char- 
acters are  summarized  in  Table  1 1 3  [see  Tyler  ( 1 980)  for  further 
information].  The  number  of  vertebrae  is  low  ( 16-30),  as  is  the 
number  of  caudal  fin  rays  (0-12).  Pelvic  fins  arc  lacking  and 
except  for  some  triodontids.  the  fins  lack  spines.  Larvae  are 
heavily  pigmented.  The  few  larval  specializations  which  do  oc- 
cur are  the  vesicular  dermal  sac  of  all  species  and  the  huge 
dermal  spines  of  molids. 

Only  two  groups  have  specialized  ontogenetic  stages  between 
larvae  and  juveniles.  In  the  Diodontidae,  some  Atlantic  species 
of  Chilomycterus  (sensu  lato)  have  a  postflexion  stage  {"Lyo- 
sphaera")  that  lacks  dermal  spines,  but  has  fleshy  protuberences 
in  the  locations  the  spines  will  occupy  and  other  enlarged  pro- 
tuberences unassociated  with  spines  (Evermann  and  Kendall, 
1898;  Breder,  1927;  Heck  and  Weinstein,  1978).  In  the  Molidae, 
Mola  and  Mastwus  have  a  deep-bodied,  compressed  stage 
("Molacanlhus")  that  has  reduced  larval  spines,  and  a  distinctly 
non-adult  shape  (Martin  and  Drewry,  1978). 

Family  Accounts 
Triodontidae. —  'Ho\.hm%  is  known  of  triodontid  eggs  or  larvae. 

Tetraodontidae.  — Tetraodontid  eggs  are  demersal,  small  to  me- 
dium-sized, have  multiple  oil  droplets  (Table  1  14)  and  hatch 
in  3-20  days.  The  very  large  ovarian  eggs  of  Chonerhinos  (Table 
1 14),  a  highly  specialized  freshwater  genus  (Tyler,  1980;  Rob- 
erts, 1982a),  are  here  regarded  as  a  specialization  for  freshwater 


LEIS:  TETRAODONTOIDEI 


449 


Fig.  242.  Tetraodontoid  yolk-sac  larvae.  All  specimens  are  enclosed 
in  a  more  or  less  well-developed  vesicular  dermal  sac.  The  vesicles  are 
omitted  in  the  drawmgs.  From  top  to  bottom:  Lagocephalus  tunaris 
(Tetraodontidae)  1.7  mm  (1.9  mm  TL)  (after  Fujita,  1966);  Fugii  par- 
dalis  (Tetraodontidae)  2.6  mm  (2.84  mm  TL)  (after  Uchida  et  al.,  1 958); 
Diodon  (hystnx^)  (Diodontidae)  2.6  mm  (after  Leis,  1978);  and  Ran- 
-aiiia  laevis  (Molidae)  1.8  mm  (after  Lets,  1977). 


conditions  [freshwater  species  commonly  have  larger  eggs  than 
their  marine  confamiliais  (Roberts,  pers.  comm.)].  The  chorion 
is  adhesive.  Parental  care  of  eggs  is  known,  but  not  universal. 
Development  of  larvae  at  hatching  varies  with  species:  jaws 
totally  unformed  to  partially  formed;  the  eye  ranges  from  un- 
pigmented  to  completely  pigmented;  the  pectoral  bud  may  be 
present  or  absent;  a  moderately  developed  vesicular  dermal  sac 
encloses  head  and  trunk;  much  yolk  remains;  and  pigment  ranges 
from  moderate  to  heavy  (Fig.  242).  If  the  often  huge  yolk  sac 
is  ignored,  larvae  are  initially  cylindrical,  but  become  progres- 
sively deeper  and  wider-bodied  with  growth  (Fig.  243).  Larvae 


Fig.  243.  Tetraodontoid  larvae.  From  top  to  bottom:  Unidentified 
tctraodontid  larva  (possibly  Canlhigaslcr),  3.6  mm,  from  the  Great 
Barner  Reef.  Note  small  spines  in  skin;  Tragnlichlhys  jacitliferus  (Dio- 
dontidae), 4.2  mm.  from  the  Great  Barrier  Reef  (small  circles  in  the 
dermal  sac  represent  incipient  spines  and  arc  ossified);  and  Ranzania 
lac'vis  (Molidae),  3.9  mm  (after  Leis,  1977). 


remain  deeper  than  broad  until  they  acquire  the  ability  to  inflate. 
Until  mid-preflexion  stage  the  body  remains  relatively  fusiform 
with  a  well-developed  tail  (relative  to  other  tetraodontoids  and 
ostraciids).  The  moderately-developed  vesicular  sac  often  dis- 
appears during  the  pretlexion  stage,  but  may  be  retained  in  some 
species  until  after  flexion.  This  sac  does  not  correspond  to  the 
inflatable  belly  found  in  this  family.  The  gill  opening  closes  to 
a  pore  shortly  after  the  yolk  is  absorbed,  but  the  membranes 
are  thin  and  transparent  and  thus  easily  missed.  Sequence  of  fin 


450 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


formation  is:  P.-D  =  A-C'.  The  long  notochord  tip  persists  for 
a  time  following  flexion.  The  vesicles  of  the  dermal  sac  are  said 
to  be  the  source  of  the  small  dermal  ossifications  (Welsh  and 
Breder,  1921).  The  dermal  ossifications  (=scales)  develop  di- 
rectly into  small,  often  embedded,  spines.  The  dermal  spines 
seem  to  first  appear  on  the  belly,  usually  in  the  preflexion  stage. 
Depending  on  species,  the  spines  may  appear  on  the  rest  of  the 
body  shortly  thereafter,  slowly  and  gradually,  or  not  at  all.  Pig- 
ment is  mitially  heavy  over  the  gut,  brain  and  yolk  sac,  and 
usually  spreads  to  cover  much  of  the  head  and  trunk  before 
flexion.  Welsh  and  Breder  (1921)  and  Munro  (1945)  report  the 
presence  of  a  single  opercular  spine  in  preflexion  larvae  of 
Sphoeroides  maculalus  and  Torquigener  pleurogramina,  re- 
spectively. None  of  the  larvae  examined  for  the  present  study 
has  such  a  spine,  but  these  two  species  have  not  been  examined. 

Diodontidae.  —  Diodontid  eggs  are  pelagic,  large,  have  multiple 
oil  droplets  (Table  114)  and  hatch  in  3  to  5  days.  Larvae  are 
moderately  to  well  developed  at  hatching,  but  development  var- 
ies between  species  and  possibly  between  populations  of  the 
same  species:  jaws  range  from  totally  unformed  to  formed  and 
apparently  functional;  eyes  are  partially  to  fully  pigmented;  the 
gill  opening  is  reduced  to  a  pore;  moderate  to  heavy  pigment 
(including  yellow,  red  and  orange)  is  present;  much  yolk  re- 
mains; and  a  well-developed,  inflated,  vesicular  dermal  sac  en- 
closes head  and  trunk  (Fig.  242).  Larvae  are  deep-bodied  and 
broader  than  deep  (Fig.  243).  At  hatching  or  very  shortly  there- 
after, diodontid  larvae  are  extremely  rotund  with  head  and  trunk 
a  single  ball-like  unit.  The  tail  is  small  and  becomes  relatively 
smaller  still  with  age.  It  becomes  nearly  vestigial  during  flexion, 
but  thereafter  starts  to  increase  in  size.  Body  shape  changes  little 
during  development.  The  fins  form  P.-D  =  A-C.  The  mouth 
is  large  compared  with  other  tetraodontiform  larvae.  Shortly 
before  flexion,  lens-like  thickenings  form  in  the  dermal  sac,  and 
(depending  on  species)  small  swellings  or  elongate  papillae  form 
over  these.  The  large  spines  (=scales)  subsequently  form  inside 
these  structures  without  an  intermediate  stage.  In  most  species, 
spines  are  present  around  the  time  flexion  is  completed,  but  in 
Chilomyclerus  antennatus  and  C.  schoepfi  (but  not  C.  affinis  or 
C.  orbicularis)  there  is  a  specialized  pelagic  stage  which  lacks 


'  Sequence  of  ossification  of  first  element  in  each  fin,  except  that  the 
symbol  for  caudal  fin  (C)  refers  to  completion  of  notochord  flexion.  Fin 
preceding  dash  forms  prior  to  fin  following  dash. 


spines  and  may  have  some  of  the  elongate  papillae  enormously 
enlarged  (the  genus  Lyosphaera  was  described  from  such  a  stage). 
The  spines  in  the  "Lyosphaera"  stage  form  after  settlement. 
Nostrils  of  diodontids  form  in  a  conventional  manner.  Only 
following  development  of  a  short  tentacle  with  two  openings  do 
the  split  nasal  flaps  of  Dicotylichthys  or  the  open  reticulated 
nasal  cups  of  Ctiilomyclerus  ajfinis  form  during  the  late  juvenile 
stage.  Pigment  is  moderate  to  heavy  and  in  preflexion  larvae 
much  heavier  dorsally  than  ventrally.  Following  flexion,  there 
is  a  tendency  for  the  belly  to  become  more  heavily  pigmented 
than  the  dorsum. 

A/olidae.  —  Mo\id  eggs  are  pelagic,  large,  have  multiple  oil  drop- 
lets (Table  114).  and  hatch  in  7  to  8  days.  Larvae  are  devel- 
opmentally  very  advanced  at  hatching  with:  jaws  formed;  eyes 
pigmented;  gill  opening  reduced  to  a  pore;  a  well-developed 
vesicular  dermal  sac  enclosing  head  and  trunk;  the  cleithrum 
and  several  pectoral  fin  rays  ossified;  a  dorsal  fin  anlage;  heavy 
pigment;  and  an  unknown  amount  of  yolk  (Fig.  242).  The  body 
is  deep  (Fig.  243)  and  wide  but  not  as  wide  as  deep.  At  hatching 
molid  larvae  are  extremely  rotund  with  head  and  trunk  a  single 
ball-like  unit.  The  compressed  tail  becomes  progressively  small- 
er. With  growth  and  body  spine  development  the  body  even- 
tually becomes  more  compressed  and  a  ventral  keel  forms.  The 
fins  form  Pi-D  =  A-Clavus.  The  P,  forms  very  early  and  be- 
comes large.  The  tail  of  young  larvae  is  normal,  but  soon  begins 
to  atrophy,  and  a  true  caudal  fin  never  forms.  Notochord  flexion 
does  not  take  place,  so  the  clavus  is  not  homologous  with  the 
caudal  fin.  Shortly  after  hatching,  the  huge  spines  which  char- 
acterize molid  larvae  begin  to  form.  These  reach  a  maximum 
size  at  about  the  time  the  clavus  is  formed.  As  the  massive 
spines  decrease  in  size,  small  spines  form  elsewhere,  particularly 
on  the  ventral  keel.  Also,  small  ossifications  within  the  skin 
begin  to  form,  and  these  eventually  make  up  the  carapace-like 
skin  covering.  Mo/a  and  Mastwus  pass  through  a  fairly  long 
ontogenetic  stage  between  larvae  and  juveniles  which  is  char- 
acterized by  retention  of  reduced  massive  spines,  a  deep,  com- 
pressed body  with  a  ventral  keel  and  a  shape  quite  unlike  the 
adult  (the  genus  Molacanthus  was  described  from  such  a  stage). 
Ranzania.  in  contrast,  loses  its  spines  relatively  quickly  and 
directly  assumes  the  adult  shape.  Larvae  are  heavily  pigmented 
over  the  gut  and  on  the  dorsal  surfaces. 

Department  of  Ichthyology,  The  Australian  Museum,  P.O. 
Box  A285,  Sydney,  2000,  Australia. 


Balistoidei:  Development 

A.  Aboussouan  and  J.  M.  Leis 


THE  tetraodontiform  suborder  Balistoidei  (Sclerodermi)  is  a 
small  group  of  six  families  with  about  175  recent  species 
of  great  morphological  diversity  (Tyler,  1968,  1980;  Winter- 
bottom,  1974a;Matsuura,  1979).Thesuborderisgenerallyagreed 
to  consist  of  the  six  families  (Table  115)  considered  here  (Tyler, 


1980).  However,  Winterbottom  (1974a)  has  suggested  that  the 
triacanthodids  and  triacanthids  could  be  removed  to  a  suborder 
distinct  from  all  other  tetraodontiform  fishes.  The  group  is  large- 
ly tropical  and  marine,  but  some  species  range  well  into  the 
temperate  zones,  particularly  in  Australia.  Most  species  are  bot- 


ABOUSSOUAN  AND  LEIS:  BALISTOIDEI 


451 


Table  1 15.    Range  of  Mi 

ERISTIC 

Characters 

OF  Balistoids. 

Mostly  after  Tyler, 

1980:  see  this  and  Matsuura,  1979  for  further  information. 

Tnacanthodidae 

Triacanthidae 

Balistidae 

Monacanlhidae 

Aracanidae 

Ostraciidae 

Number  of  species 

19 

7 

35 

90 

12 

15 

(after  Nelson,  1976) 

Dorsal  spines 

6 

4  or  6 

3 

1  or  2 

0 

0 

Second  dorsal  rays 

12-18 

19-26 

23-35 

22-50 

9-12 

9-11 

Anal  rays 

11-17 

13-22 

19-31 

20-62 

9-11 

8-11 

Pectoral  rays 

12-15 

12-16 

12-17 

8-16 

10-13 

9-13 

Caudal  rays 

6  +  6 

6  +  6 

6  +  6 

6  +  6 

5  +  5-6 

5  +  5 

Pelvic  spines 

1 

1 

0 

0 

0 

0 

Pelvic  rays 

0-2 

0-1 

0 

0 

0 

0 

Ventral  scales 

absent 

absent 

present 

absent  or 
present 

absent 

absent 

Vertebrae 

8 

+  12  =  20 

8+12  =  20 

7  +  10-12  = 
17-19 

6-8  +  1 1-23  = 
19-31 

9- 

-10  +  8-9  = 

18 

9  +  9-10  = 
18-19 

Caudal  fin  bones 

Epural 

1 

1 

1 

1 

0 

0 

Uroneural 

2 

1 

1  ? 

0 

0 

0 

Hypural 

3  to  5 

3 

3 

2  or  3 

2 

2 

Smallest  hypural  (5th) 

present 

present 

present 

absent  or 
present 

absent 

absent 

Parhypural 

1 

1 

1 

1 

0 

0 

Vertebrae  before  the 

8 

4-5 

5 

4-7 

5-6 

6-8 

first  second  dorsal 

pterygiophore 

Vertebrae  behind  the  last 

5-6 

7 

4-5 

4-6 

4-6 

4-6 

anal  pterygiophore 

tom-associated  in  shallow  to  moderate  depths,  but  many  tria- 
canthodids  live  in  deep  (>500  m)  water.  Most  species  have  a 
pelagic,  often  oceanic,  juvenile  stage,  and  a  few  are  pelagic 
throughout  their  lives. 

Development 

Development  of  balistoid  fishes  is  not  well  known.  Previous 
reviews  of  the  early  development  of  the  group  are  by  Breder 
and  Clark  (1947),  Tortonese  (1956).  and  Martin  and  Drewry 
(1978).  The  early  development  of  aracanids  is  entirely  unknown, 
and.  overall  information  is  available  for  only  30  species.  Often 
the  information  available  for  a  species  is  scanty.  Complete  (i.e., 
egg  to  juvenile)  information  is  available  for  only  four  or  five 
species  (Table  1 16).  This  narrow  data  base  makes  generaliza- 
tions about  development  somewhat  suspect.  However,  we  as- 
sume that  the  few  taxa  for  which  information  is  available  are 
representative. 

Few  generalizations  can  be  made  about  development  of  balis- 
toid fishes,  but  this  is  not  surprising  in  view  of  the  diversity  of 
the  adults.  A  reference  to  development  in  juveniles,  which  usu- 
ally differ  little  from  adults,  is  given  at  the  end  of  each  section. 
We  make  no  attempt  to  review  the  literature  on  juvenile  de- 
velopment. 

Eggs  (Table  //6A  — Balistoid  fishes  are  oviparous.  Pelagic  and 
demersal  eggs  are  known:  the  chorion  is  usually  smooth,  but 
may  have  limited  sculpturing;  oil  droplets  are  usually  present; 
eggs  range  in  size  from  small  (0.5  mm)  to  large  (2.0  mm)  and 
are  approximately  spherical;  incubation  times  range  from  one 
to  four  days;  development  at  hatching  vanes  widely;  the  peri- 
vitelline  space  is  narrow;  the  yolk  is  unsegmented;  and  consid- 
erable pigment  may  develop  on  the  embryo.  Parental  care  of 
eggs  ranges  from  non-existent  (pelagic  eggs)  to  considerable  (Ba- 
listidae). 


Larvae— AW  balistoid  larvae  are  pelagic.  Development  is  gen- 
erally direct  (i.e.,  no  specialized  ontogenetic  stages  between  lar- 
vae and  juveniles),  with  few  larval  specializations,  and  is  com- 
pleted at  a  small  size  (Figs.  244-25 1 ).  There  is  often  an  apparently 
unspecialized  pelagic  juvenile  stage  which  may  grow  to  a  sig- 
nificant fraction  of  the  adult  size.  Larvae  tend  to  be  deep-bodied, 
and  many  are  also  wide-bodied.  The  head  is  large  and  the  gut 
coiled  and  massive.  The  mouth  is  small.  The  head  is  usually 
rounded,  at  least  in  preflexion  larvae.  The  head  and  body  of 
young  ostraciid  larvae  are  enclosed  in  an  inflated  dermal  sac 
which  has  numerous  vesicles  (or  tubercles)  embedded  in  its 
outer  surface.  Except  in  ostraciids,  the  specialized  adult  scales 
pass  through  an  unspecialized  spinule  stage.  The  caudal  fin  is 
usually  the  last  fin  to  form.  The  reduction  in  structures,  notably 
fins,  which  characterizes  the  balistoid  fishes  is  not  a  case  of 
development  followed  by  loss— these  structures  never  develop. 
The  specialized  dentition  develops  during  the  larval  stage  di- 
rectly, without  any  intervening  generalized  or  larval  teeth.  Me- 
ristic  characters  are  summarized  in  Table  1 15  (see  Tyler.  1980 
for  further  information).  The  number  of  vertebrae  and  caudal 
fin  rays  is  low.  pelvic  fins  are  reduced  or  lacking,  anal  fins  lack 
spines,  and  dorsal  spines,  if  present,  are  few  (Table  1 1 5).  Larvae 
are  generally  moderately  to  heavily  pigmented. 

The  few  larval  morphological  specializations  which  do  occur 
are  either  developments  of  the  often  very  specialized  scales  (or 
their  precursors)  of  the  adults  or  delicate  skin  flaps,  filaments 
and  tendrils.  These  are  discussed  under  the  appropriate  family 
section.  There  are  no  specialized  ontogenetic  stages  between 
larvae  and  juveniles. 

A  shorthand  notation  will  be  used  to  designate  the  sequence 
of  fin  formation.  By  formation,  we  mean  ossification  of  the  first 
element,  with  the  exception  of  the  caudal  fin  where  completion 
of  flexion  is  meant.  However,  except  for  some  monacanthids 


452 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  1 16.     Balistoid  Taxa  for  which  Information  is  Available  on  Egg  and  Larval  Stages.  YS— yolk-sac  stage;  pre  — preflexion  stage; 
flex— flexion  stage;  post  — postflexion  stage;  U  — unstated;  dem— demersal;  pel— pelagic.  Numbers  in  parentheses  after  each  genus  refer  to  the 

number  of  species  represented.  A  blank  means  no  information  available. 


Eggs 

Lar\'ae 
developmental  stage 

Type 

Size 
(mm) 

Oil 
droplets 

YS 

Pre 

Rex 

Post 

Triacanthodidae 

Atrophacanthus  ( 1 ) 
Macrorhamphosodes  ?  ( 1 ) 

X 

X 

X 
X 

Triacanthidae 

Triacanthus  ( 1 ) 
Unidentified  (1) 

pel 

0.78 

0 

X 

X 
X 

X 

X 
X 

Balistidae 

Balisles  (2) 

dem 

U 

1 

X 

X 

X 

X 

Balistapus  ( 1 ) 
Canthidermis  (1-2?) 

dem 
dem 

0.55 
U 

0?= 

u 

X 

X 

X 

X 

Odonus  ( 1 ) 
Pseudobalisles  (3) 

dem 
dem 

U 
0.55-0.60 

u 

0?- 

Sufflamen  (3) 

dem 

0.51-0.56 

1  &  u 

X 

X 

Xanrhichlhys  ( 1 ) 
Unidentified  (>  1) 

Monacanthidae 
Alulera  (2) 
A  manses  ( 1 ) 
Anacanlhus  (1) 
Brachaluteres  ( 1 ) 
Camheriiies  ( 1 ) 
Monacanlhus  ( I ) 
Navodon  ( 1 ) 

Parika  ( 1 ) 
Pseudalutaris  ( 1 ) 
Rudanus  ( 1 ) 


Stephanolepis  (3) 

Unidentified  (>10) 

Ostraciidae 

Acanlhoslracion  ( 1 ) 

Lac  lor  ia  (2) 

Ostracion  (2) 

Tetrosomus  (1) 
Unidentified  (6) 

Rhmesoinus  (1) 


dem 


pel'     0.65-0.74'  1' 

dem  0.52  2 

dem     0.61-0.70      cluster        X 


pel         1.4-1.6  1  X 

pel         1.6-1.9         cluster        X 
pel         1.6-1.9         cluster        X 


pel 


1.4-2.0 


cluster 


X 

X         X 


X         X 

X 

X 

X         X 

X 

X         X 


X  X 

X 

X 


X 
X 
X 

X 

X 


X 
X 


X 
X 


Fraser-Brunner,  1950;  Tyler,  1968;  present  study 
Present  study 


Ohsima  and  Nakamura,  1941;  Gopinath,  1946 
Present  study 

Sanzo,  1939b';  Gamaud,  1 960;  Aboussouan,  1966; 

Lythgoe  and  Lythgoe,  1975;  Matsuura  and 

Katsuragawa.  1981 
Lobel  and  Johannes,  1980 
Nellis,  1980;  Watson  and  Walker,  pers.  comm.; 

present  study 
Fricke,  1980 
Masuda  et  al.,  1975;  Fricke,  1980;  Lobel  and 

Johannes,  1980;  Matsuura,  1982 
Ballard,  1970;  Sweatman,  pers.  comm.;  Thresher, 

pers.  comm.;  present  study 
Present  study 
Leis  and  Rennis,  1983;  present  study 

Clark,  1950;  Suzuki  et  al.,  1980;  present  study 

Present  study 

Present  study 

Leis  and  Rennis  (1983,  figure  72);  present  study 

Present  study 

Present  study 

Uchida  et  al.,  1958;  Kobayashi  and  Abe,  1962; 

Mho,  1966 
Regan,  1916;  Robertson,  1975a;  Crossland,  1981 
Lets  and  Rennis,  1983;  present  study 
Fujita,  1955;  Uchida  et  al.,  1958;  Kobayashi  and 

Abe,  1962;  Mito,  1966;  Masuda  et  al.,  1975; 

Suzuki  etal.,  1980 
Ryder,  1887;  Hildebrand  and  Cable,  1930;  Fujita, 

1955;  Uchida  et  al..  1958;  Mito,  1966;  Aboussouan, 

1966 
Leis  and  Rennis,  1983;  Watson  and  Walker,  pers. 

comm.;  present  study 

Brederand  Clark,  1947;  Palko  and  Richards,  1969; 

present  study 
Watson  and  Leis,  1974;  Moyer,  1979;  Leis  and 

Meyer,  MS;  present  study 
Watson  and  Leis,  1974;  Leis  and  Rennis,  1983;  Leis 

and  Moyer,  MS;  present  study 
Present  study 
Delsman,  1930d-';  Sanzo,  1930d';  Mito,  1962c,  1966; 

present  study 
Present  study 


Notes: 

'  Two  specimens  (L86  and  2.48  mm)  from  a  supposed  senes  of  Dactvluplerus  volnans  appear  to  be  Baiisla  capmcus.  and  one  has  the  preopercular  cluster  of  spinules.  A  4  mm  specimen 
identified  as  B  capnscus  is  also  illustrated. 

'  Lobel  and  Johannes  (1980)  descnbe  the  eggs  as  "without  visible  inclusions."  but  their  photograph  of  a  newly  hatched  B.  imdulalus  seems  to  show  an  oil  drop  in  the  yolk  sac. 

'  Eggs  identified  as  Wovodon  [sic]  convexiroslris  i= Parika  scaber)  were  described  by  Robertson  (1975a)  and  Crossland  ( 1981 ).  however  there  is  reason  to  question  this  identification-  Robertson 
(in  lill-  Nov.  1982)  notes  the  identification  and  classification  as  pelagic  of  this  egg  was  "based  on  a  small  sample  of  npe  |unfcnili/cd]  eggs  from  a  female  leatherjacket  and  a  conforming  type  in 
the  Olago  Harbour  plankton  at  that  time,"  and  that  no  eggs  were  reared.  We  feel  the  eggs  described  by  Robertson  and  Crossland  are  not  monacanthids. 

*  Misidentified  as  Tclraodon  sp. 

■  Misidentified  as  Tclraodon  honkcnn 


ABOUSSOUAN  AND  LEIS:  BALISTOIDEI 


453 


Fig.  244.  Scanning  electron  micrograph  of  the  sculptured  chorion 
of  an  unidentified  Hawaiian  ostraciid  egg.  The  micropyle  is  the  hole  in 
the  center.  The  width  of  the  field  of  bumps  is  cci.  0.5  mm. 


Fig.  245.  Triacanthid  and  ostraciid  yolk-sac  larvae  from  top  to  bot- 
tom: Triacanthus  hiaculeatus.  1.3  mm  ( 1 .4  mm  TL)  after  Ohsima  and 
Nakamura.  1941;  and  Acanthostracion  quadricornis.  2.6  mm  reared 
larva  from  Rorida.  Specimen  is  fully  enclosed  in  a  vesicular  sac  which 
is  most  inflated  over  head  and  trunk.  The  vesicles  are  omitted  in  the 
drawing.  Specimen  is  unpigmented,  but  is  probably  bleached. 


where  the  posterior  rays  of  dorsal  and  anal  fins  are  slow  to  form, 
ossification  of  all  elements  of  the  fin  could  serve  as  an  equally 
good  definition.  The  fins  will  be  indicated  by  standard  notation 
(D— dorsal,  Dsp— dorsal  spine,  etc.).  The  order  of  the  letters 
corresponds  to  the  order  of  formation.  An  equal  sign  between 
two  letters  indicates  the  fins  form  simultaneously,  a  dash  in- 
dicates the  fins  do  not  form  simultaneously. 

Triacanthodidae 

The  eggs  of  triacanthodids  are  unknown,  although  there  is  a 
dubious  report  of  pelagic  eggs  (Nikol'skii,  1961).  The  body  of 
preflcxion  and  flexion  larvae  (Fig.  247)  is  moderately  to  very 
deep,  moderately  wide  in  head  and  trunk,  and  compressed  in 
tail.  The  body  becomes  more  compressed  and  elongate  with 
growth,  but  may  remain  very  deep  until  well  after  flexion.  The 
gill  opening  is  closed  to  a  pore  in  the  smallest  available  speci- 
mens (late  preflexion).  There  is  no  dermal  sac.  The  fins  form 
D  =  A  =  P,-C-P,  =  Dsp.  The  Dsp  anlage  and  P,  buds  do  not 
form  until  after  flexion.  Although  no  early  postflexion  larvae 
are  available,  late  flexion  larvae  have  a  notochord  with  a  long 
posterior  portion  that  probably  indicates  that  the  notochord  has 
an  extended  tip  for  awhile  following  flexion.  Dermal  spinules 
first  form  in  preflexion  larvae,  and  appear  first  on  side  of  head 
(cheek,  operculum,  over  otic  vesicle)  and  laterally  on  two  small 
regions  of  the  gut  (ventral  to  P,  base  and  just  anterior  to  anus). 
The  spinules  are  unspecialized,  and  fully  cover  the  body  of 
postflexion  larvae.  The  available  larvae  of  Atrophacanthus  are 


unpigmented,  but  their  poor  condition  implies  they  could  be 
faded.  The  Macrorhamphosodcs  (?)  larva  is  moderately  and  uni- 
formly pigmented  with  small  melanophores. 

The  specimen  identified  as  Triacaruhodes  sp.  by  Weber  (1913) 
appears  to  be  a  trichiurid  (Scombroidei),  not  a  triacanthodid. 

Tyler  (1968)  describes  juvenile  development  of  several  tria- 
canthodid species. 

Triacanthidae 

Triacanthid  eggs  lack  oil  droplets  and  chorion  sculpture,  are 
pelagic,  small,  and  hatch  in  about  22  hours  (Table  1 16).  De- 
velopment at  hatching  is  not  advanced  (Fig.  245):  no  jaws  or 
pectoral  fins  are  present,  the  eye  is  unpigmented  and  much  yolk 
remains.  The  body  is  cylindrical  at  hatching  and  becomes  much 
deeper  with  growth  (Fig.  247)  and,  especially  following  flexion, 
very  compressed.  The  gill  opening  closes  to  a  pore  prior  to 
flexion.  There  is  no  dermal  sac.  The  fins  form  D  =  A  =  P|-P;  = 
Dsp-C.  The  notochord  has  an  extended  tip  following  flexion. 
The  D  and  P,  spines  become  relatively  elongate.  Dermal  spi- 
nules first  form  in  preflexion  larvae  and  appear  first  on  the  sides 
of  the  head  (cheek,  operculum,  over  otic  vesicle),  and  laterally 
on  the  posterior  portion  of  the  gut.  The  spinules  are  unspecial- 
ized  (except  for  some  terete  ones  on  the  fin  spines),  and  fully 
cover  the  body  shortly  after  flexion.  Pigment  is  heavy  on  brain 
and  gut.  and  a  single  ventral  tail  melanophore  is  present.  Fol- 
lowing yolk  exhaustion,  pigment  spreads  over  most  of  the  body 
in  a  blotchy  pattern. 


454 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  246.  Balistid  and  monacanthid  yolk-sac  larvae  from  top  to 
bottom:  Sufflainen  chrysopterus.  1.7  mm  reared  larva  from  the  Great 
Barrier  Reef  (24  hours  after  hatching);  Stephanolepis  cirrhtfer.  1.9  mm 
(2. 1  mm  TL)  after  Fujita,  1955;  and  Anacanthns  harbalus.  2.4  mm  larva 
from  a  Great  Barrier  Reef  plankton  sample  (age  unknown).  Note  anlage 
of  dorsal  fin  spine  in  occipital  region.  Mouth  is  not  fully  formed.  Frag- 
mented oil  droplets  are  present  in  the  yolk  sac.  but  are  not  illustrated. 


The  descriptions  of  lai^ae  identified  as  Triacanthus  breviros- 
tris  by  Kuthalingam  ( 1 959b)  do  not  resemble  triacanthid  larvae 
in  morphology,  sequence  of  development,  or  size.  One  can  only 
conclude  the  larvae  are  misidentified  and  the  drawings  inac- 
curate. The  eggs  identified  by  Kuthalingam  (1959b)  as  Tria- 
canthus brevirostris  are  probably  those  of  an  atheriniform  fish. 

Tyler  (1968)  describes  juvenile  development  of  several  tria- 
canthid species. 

Bahstidae 

Balistid  eggs  are  demersal,  small,  lack  chorion  ornamentation 
(but  are  adhesive),  have  a  single  oil  droplet  (Table  1 16),  and 
hatch  in  one  to  two  days.  Eggs  are  laid  in  clusters  in  shallow 
nests  on  sand  or  rubble  bottoms  and  are  guarded  by  the  adult. 
Development  of  larvae  at  hatching  is  not  advanced:  no  jaws  are 
present,  the  eye  is  unpigmented,  minimal  body  pigment  is  pres- 
ent and  much  yolk  remains  (Fig.  246).  Larvae  have  a  cylindrical, 
slightly  compressed  body  at  hatching.  The  body  quickly  be- 
comes deeper  and  then  moderately  rotund  in  the  trunk  (Figs. 
248  and  249).  The  tail  remains  compressed.  About  the  time  fins 
start  to  form,  the  larva  starts  to  become  compressed  and  this 
increases  thereafter.  In  newly  hatched  larvae,  a  slightly  inflated 
area  is  present  surrounding  the  trunk  (Fig.  246),  but  it  contains 
no  vesicles,  and  soon  disappears.  The  gill  opening  closes  to  a 
pore  just  prior  to  flexion.  The  fins  form  Dsp-D  =  A  =  P,-C. 


Fig.  247.  Late  preflexion  larvae  of  three  balistoid  families.  Small 
ticks  on  upper  and  middle  specimens  indicate  position  of  dermal  spi- 
nules.  From  top  to  bottom:  Atrophacanlhus  japonicus  (Triacanthodi- 
dae),  composite  drawing  of  three  damaged  larvae  (2.6-2.7  mm)  from  a 
Dana  Station  in  the  Philippines;  unidentified  triacanthid.  3.5  mm,  from 
the  Great  Barrier  Reef  note  small  dorsal  spine  and  pelvic  fin  bud;  and 
Acanlhostracion  QuaiJricornis  (OsUaciidae).  3.3  mm.  reared  larva  from 
Florida.  The  dermal  plates  arc  not  yet  formed,  but  ridges  on  the  body 
are  evident. 


The  first  dorsal  spine  becomes  large  and  heavily  armed  with 
barbs  before  flexion.  This  ornamentation  varies  between  species 
and  is  useful  in  identification.  The  notochord  has  an  extended 
tip  for  a  short  while  following  flexion.  Dermal  ossifications  first 


ABOUSSOUAN  AND  LEIS:  BALISTOIDEI 


455 


Table  117. 


Characters  That  Differ  Between  the  Two  Larval 
monacanthid  morphs. 


Fig.  248.  Late  to  mid  preflexion  larvae  of  two  balistoid  families. 
Small  ticks  indicate  position  of  dermal  spinules.  From  top  to  bottom: 
Canlhidenms  sufflamen  (Balistidae),  3.5  mm,  from  Puerto  Rico,  note 
small  pelvic  bud  and  prcopercularclustcr  of  spinules;  unidentified  Morph 
A  monacanthid,  3.6  mm,  from  the  Great  Bamer  Reef,  note  pigmented 
filament  at  terminus  of  pel  vie  bone  and  preopercular  cluster  of  spinules; 
Pseudatulans  nasicorms  (Morph  B  monacanthid),  4.3  mm,  from  the 
Great  Barrier  Reef,  note  pigmented  fleshy  tendrils  laterally  on  tail  and 
preopercular  cluster  of  spinules;  unidentified  Morph  C  monacanthid, 
3.0  mm  from  the  Great  Bamer  Reef  Dermal  spinules  in  this  species 
are  longer  than  in  the  other  illustrated  species.  Dorsal  spine  is  just 
beginning  to  form  (not  yet  ossified). 


Character 

Morph 

AB  (Figs.  246  and  248) 

C  (Fig.  248) 

Body  shape 

Deep  to  elongate;  be- 

Deep; becoming 

coming  angular 

deeper  with 

with  growth.  Com- 

growth, but  re- 

pressed. 

maining  rounded. 
Somewhat  rotund 
early,  becoming 
compressed. 

Cluster  of  spinules 

Small  to  large 

Absent 

on  preoperculum 

Sequence  of  fin  for- 

Dsp-D =  A-P,- 

D  =  A  =  C-P,- 

mation 

or  =  C 

Dsp 

Dorsal  fin  spine 

Early-forming,  armed 

Late-forming,  lightly 

or  unarmed,  lightly 

pigmented 

to  moderately  pig- 

mented 

Tail  pigment  in 

Present.  Ventral  or 

None 

preflexion  larvae 

dorsal  series  or 
blotches. 

Identified  taxa  in- 

Alutera. Amanses. 

Brachaluteres.  Ru- 

cluded 

Anacanthus. 

danus  (the  morph 

Cantherines.  Mon- 

C  larvae  illustrated 

acanthus,  Navo- 

by  Leis  and  Rennis 

don.  Parika,  Pseu- 

(1983,  Fig.  72)  are 

dalularis. 

Brachaluteres) 

Stephanolepis 

appear  in  the  form  of  a  small  cluster  of  relatively  long  spinules 
on  the  preoperculum  (larval  specialization).  This  cluster  appears 
within  a  few  days  of  hatching  and  persists  until  just  prior  to 
flexion.  Shortly  before  the  cluster  disappears,  dermal  spinules 
appear  in  three  areas:  laterally  on  the  cheek  ventral  to  the  cluster; 
over  the  otic  vesicle;  and  laterally  on  the  gut  from  below  the 
pectoral  base  to  near  the  anus.  These  unspecialized  spinules 
rapidly  spread  to  cover  the  body  by  mid-flexion.  They  do  not 
transform  into  the  specialized  scales  of  the  adults  until  well  into 
the  pelagic  juvenile  stage.  A  pigmented  filament  (larval  spe- 
cialization) often  develops  at  the  terminus  of  the  pelvic  bones 
(see  discussion  of  such  structures  under  Monacanthidae).  Pig- 
ment is  heavy  on  the  brain  and  gut,  and  preflexion  larvae  have 
a  series  of  melanophores  on  the  ventral  midline  of  the  tail. 
Blotches  or  bands  may  form  on  the  tail.  The  spiny  dorsal  fin  is 
heavily  pigmented  and  this  pigment  spreads  laterally  over  the 
trunk. 

Berry  and  Baldwin  (1966)  describe  juvenile  development  of 
several  balistid  species. 

Monacanthidae 

Monacanthid  eggs  are  demersal  (we  tentatively  conclude  that 
pelagic  eggs  were  wrongly  attributed  to  Parika  scaher—see  Table 
1 16),  adhesive,  small,  have  several  oil  droplets,  and  hatch  in 
about  2  days.  Eggs  are  attached  to  vegetation,  and  there  is  no 
record  of  parental  care.  Development  at  hatching  is  not  ad- 
vanced: jaws  are  absent  or  only  partially  formed,  the  eye  is 
unpigmented,  and  much  yolk  remains  (Fig.  246).  Newly-hatched 
larv  ae  are  cylindrical  and  somewhat  compressed.  Morphological 
and  developmental  diversity  among  monacanthid  larvae  is  high. 


456 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


■BSCor- 


Fig.  249.     Balistid  preflexion  larva  Xanlichthys  ringens  3.87  mm,  ASFIOl,  western  Atlantic. 


particularly  in  comparison  with  other  tetraodontiform  families. 
Leis  and  Rennis  (1983)  considered  that  three  distinct  morphs 
were  present  among  larval  monacanthids.  However,  our  studies 
of  additional  taxa  indicate  that  Morphs  A  and  B  of  Leis  and 
Rennis  (1983)  are  merely  extremes  (Fig.  248),  and  that  no  clear 
division  can  be  made  between  A  and  B.  For  example,  while  Leis 
and  Rennis  (1983)  utilized  seven  characters  to  separate  the  two 
morphs,  larvae  of  Alutera  sp.  (Fig.  250)  have  three  'A'  char- 
acters, three  'B'  characters,  and  are  intermediate  for  the  seventh. 
Morph  C  of  Leis  and  Rennis  (1983)  (Fig.  248)  is  distinct  from 
the  combined  Morph  AB  (Table  1 17).  Rapid  changes  in  body 
proportions  may  take  place  in  many  species.  Morph  AB  larvae 
are  compressed  and  become  more  so  with  growth,  while  Morph 
C  larvae  are  moderately  broad  in  gut  and  head,  but  become 
compressed  with  growth.  The  gill  opening  closes  to  a  pore  late 
in  the  preflexion  stage,  and  the  position  of  the  pore  relative  to 
the  eye  varies  with  species.  The  dorsal  spines  form  at  a  very 
early  stage  in  Morph  AB  larvae,  but  are  the  last  fin  elements  to 
form  in  Morph  C  larvae.  In  some  species,  the  first  spine  becomes 
heavily  armed  by  the  mid-preflexion  stage.  The  pelvis  may  form 
either  early  and  be  prominent  by  the  mid-preflexion  stage  or 
very  late  and  may  never  become  externally  visible,  depending 
on  species.  The  sequence  of  fin  development  is  morph-depen- 
dent  (Table  1 17).  There  is  no  dermal  sac.  The  notochord  tip  is 


long  and  persists  for  a  time  following  flexion.  If  present  (Morph 
AB),  the  small  cluster  of  spinules  on  the  preoperculum  (a  larval 
specialization)  forms  very  early  and  is  lost  before  flexion.  De- 
pending on  species,  dermal  spinules  may  first  appear  laterally 
on  the  gut  and  head,  or  on  the  forehead  and  along  the  ventral 
midline  near  the  cleithral  symphysis.  Dermal  spinules  cover  the 
body  prior  to  flexion  or  shortly  thereafter.  Pigment  is  heavy  on 
the  brain  and  over  the  gut,  but  on  the  tail,  it  varies  with  species. 

Several  species  temporarily  develop  small,  pigmented  flaps 
or  filaments  (a  larval  specialization)  on  different  portions  of  the 
body.  Alutera  sp.  (Fig.  250)  develops  an  elongate  flap  which 
originates  on  the  operculum  near  the  preopercular  spinule  clus- 
ter; Pseudaliiiaris  nasicornis  (Fig.  248)  develops  several,  elon- 
gate tendrils  laterally  on  the  tail;  and  many  species  develop  a 
filament  at  the  terminus  of  the  pelvic  bones  (Fig.  248).  The 
latter  possibly  represents  a  pelvic  fin  bud  that  atrophies. 

The  description  of  Slephanolepis  hispidus  by  Hildebrand  and 
Cable  ( 1 930)  seems  to  be  based  on  more  than  one  monacanthid 
species  (Martin  and  Drewry,  1978).  Berry  and  Vogelc  (1961) 
describe  the  juvenile  development  of  several  monacanthid 
species. 

Hildebrand  and  Cable  (1930)  state  that  the  pelvis  oi'  Sleph- 
anolepis hispidus  possibly  forms  through  coalescence  of  two 
separate  fin  buds.  In  the  material  available  to  us  (Table  1 16), 


ABOUSSOUAN  AND  LEIS:  BALISTOIDEI 


457 


^tS'i" 


/::;::. 


y.^.f 


Mor- 


Fig.  250.     Monacanthid  preflexion  larvae,  (upper)  A  manses  pulhis  3.99  mm  SL,  Y-92;  (lower)  ,-l/M;era  sp.  2.76  mm  SL,  ASF-94  western  Atlantic 
(preopercular  cluster  of  spinules  not  seen  in  these  specimens). 


the  pelvis  forms  from  a  single  unpaired  aniage  located  just  pos- 
terior to  the  cleithral  symphysis.  The  terminal  encasing  scales 
form  first  as  unspecialized  spinules,  and  at  the  same  time  a  pair 
of  pelvic  elements  begins  to  ossify.  The  pelvis  then  fuses  begin- 
ning from  its  base  (e.g.,  in  a  2.35  mm  specimen,  the  pelvis  is 
roughly  'Y'-shaped  and  fused  along  about  75%  of  its  length). 


The  "two  ventral  fins"  observed  by  Hildebrand  and  Cable  (1930) 
are  probably  dermal  flaps  similar  to  those  of  Alutera  sp. 

Development  of  the  preopercular  cluster  of  spinules,  aside 
from  Leis  and  Rennis  (1983),  has  been  described  in  published 
works  only  for  Batistes  capnscus  (Sanzo.  1939b;  Matsuura  and 
Katsuragawa,  1981)  and  Pahka  scaber  (Crossland,  1981),  al- 


Fig.  251.     Ostraciid  larvae  from  top  to  bottom:  Rhmesoinus  triqueter  2.85  mm  SL,  ASF-37,  western  Atlantic;  Laclophrys  quadncornis  2.53 
mm  SL,  and  6.0  mm  SL  ASF94,  western  Atlantic  (exhalent  gill  openings  not  shown). 


ABOUSSOUAN  AND  LEIS:  BALISTOIDEI 


459 


though  in  the  latter  it  is  illustrated  as  a  serrate  preopercular 
border.  However,  this  structure  is  present  in  all  balistids  and 
Morph  AB  monacanthids  examined  for  the  present  study,  and 
because  it  is  an  inconspicuous  structure  it  is  most  likely  that  it 
is  present  in  previously  described  taxa  and  has  been  overlooked 
(see  Fig.  250). 

Aracanidae 
Nothing  is  known  of  aracanid  eggs  or  larvae. 

Ostraciidae 

Ostraciid  eggs  are  pelagic,  large,  slightly  ovoid,  have  one  or 
more  oil  droplets  (Table  1 16)  and  hatch  in  two  to  four  days. 
There  is  some  chorion  ornamentation  surrounding  the  micro- 
pyle.  In  Indo-Pacific  species  (Ostraciinae)  this  consists  of  a  par- 
tially raised  field  of  small  bumps  surrounding  a  small  pore-like 
depression  containing  the  micropyle  (Fig.  244).  In  the  single 
Atlantic  species  examined  (Lactophrysinae),  only  the  pore-like 
depression  is  present.  Development  of  larvae  at  hatching  is  rel- 
atively advanced,  but  there  is  some  interspecific  variation  in 
how  advanced:  jaws  are  totally  unformed  to  formed  and  ap- 
parently functional,  the  eye  is  unpigmented  to  partially  pig- 
mented, dorsal  and  anal  fin  anlagen  may  be  present.  Moderate 
pigment  is  present,  much  yolk  remains,  the  gill  opening  is  re- 
stricted to  a  pore,  and  an  inflated  vesicular  dermal  sac  encloses 
head  and  trunk  (Fig.  244).  The  dermal  sac  disappears  before 
flexion.  The  larvae  are  deep-bodied  and  the  tail  is  compressed 
(Fig.  247).  Depending  on  species,  the  body  may  be  moderately 
(Rhinesomus)  to  very  wide  (Ostracion)  (Fig.  251):  the  lacto- 
phrysine  species  examined  were  more  narrow-bodied  than  the 
ostraciine  species.  Larvae  tend  to  become  wider  with  growth, 
but  never  become  as  wide  as  deep.  At  hatching,  ostraciine  larvae 


are  rotund  with  head  and  trunk  a  single  ball-like  unit,  and  they 
have  a  small  tail.  Lactophrysine  larvae  attain  this  condition 
within  a  few  days  of  hatching.  The  tail  progressively  becomes 
relatively  smaller  with  age  until  after  flexion,  and  the  ball-like 
shape  of  the  body  is  retained.  The  notochord  tip  is  small.  The 
lips  have  an  unusual  flared  structure.  The  fins  form  P.-D  =  A- 
C.  The  dermal  ossifications  do  not  pass  through  a  spinule  stage, 
but  form  directly  starting  as  thickenings  in  the  dermal  sac  which 
ossify  and  grow  out  from  their  centers.  These  eventually  coalese 
into  the  mosaic-like  armoured  carapace  characteristic  of  adults. 
The  individual  carapace  units  that  eventually  produce  spines 
and  other  ornamentation  tend  to  be  larger  and  with  more  relief 
than  other  carapace  units.  The  ossifications  become  visible  well 
before  flexion,  and  larvae  are  fully  armoured  by  the  end  of 
flexion.  Pigment  is  moderate  to  heavy  and  generally  uniform 
on  head  and  trunk,  with  the  tail  often  unpigmented. 

Le  Danois  (1961)  describes  the  juvenile  development  of  sev- 
eral ostraciid  species. 

Chorion  ornamentation  of  ostraciid  eggs  previously  has  been 
reported  only  for  Hawaiian  taxa  (Watson  and  Leis,  1974;  Leis, 
1977,  1978),  however  it  is  present  in  all  ostraciid  eggs  examined 
in  the  present  study  (Fig.  244),  albeit  reduced  to  a  pore  in  Acan- 
thostracion  quadncornis  (Table  1 1 6).  The  ornamentation  is  sub- 
tle and  confined  to  a  small  portion  of  the  chorion,  and  we  feel 
it  is  probably  present  in  all  taxa,  but  has  been  overlooked  in 
previous  descriptions. 

(A. A.)  Station  Marine  D'Endoume  et  Centre  D'Oceanog- 
RAPHiE,  Rue  Batterie  des  Lions,  13007,  Marseille, 
France;  (J.M.L.)  Department  of  Ichthyology,  The 
Australian  Museum,  P.O.  Box  A285,  Sydney,  N.S.W., 
2000,  Australia. 


Tetraodontiformes:  Relationships 
J.  M.  Leis 


IN  this  contribution  I  construct  a  phylogeny  of  tetraodontiform 
fishes  based  on  early  life  history  (ELH)  characters  and  con- 
trast this  with  phylogenies  based  on  adult  characteristics.  The 
ELH  characters  of  tetraodontiform  fishes  are  summarized  in  the 
preceding  two  contributions  (Aboussouan  and  Leis,  and  Leis, 
this  volume).  Although  in  many  cases  there  is  little  information 
available,  I  have  assumed  that  which  is  available  is  represen- 
tative, and  that  new  information  will  not  change  the  conclusions 
herein.  This  is  unlikely,  and  for  this  reason,  the  present  treat- 
ment must  be  viewed  with  caution. 

Inter-ordinal  Relationships 

The  tetraodontiform  fishes  are  usually  presumed  to  have  been 
derived  from  perciform  ancestors,  with  the  Acanthuroidei  being 
the  popular  choice  for  closest  relative  (Tyler,  1980;  Winterbot- 
tom,  1974a;  Lauder  and  Liem,  1983).  However,  D.  E.  Rosen  in 
an  unpublished  study  (pers.  comm.)  presents  evidence  sup- 
porting a  relationship  between  zeiform  and  tetraodontiform  fishes 
(see  also  Winterbottom,  1974a). 


There  is  little  in  the  early  life  history  of  tetraodontiform  fishes 
to  indicate  they  are  the  sister  group  of  the  acanthuroid  fishes. 
The  few  ELH  characters  acanthuroids  and  tetraodontiforms  share 
(small  mouth,  gas  bladder  present,  relatively  few  myomeres, 
large  head,  oviparity,  spherical  eggs  with  unsegmented  yolk)  are 
very  widespread  in  the  perciform  fishes,  and  the  larvae  are  not 
even  generally  similar  (see  Leis  and  Rennis,  1983).  Certain  char- 
acter states  (e.g.,  scale  development)  are  shared  by  the  acanthu- 
roid fishes  and  some  groups  of  tetraodontiform  fishes.  This  sit- 
uation could  be  interpreted  as  indicating  a  relationship  between 
acanthuroids  and  tetraodontiforms,  whereupon  the  character 
state  involved  would  be  viewed  as  primitive  for  the  Tetra- 
odontiformes as  a  whole.  Therefore,  the  presence  of  an  alternate 
character  state  in  some  tetraodontiform  families  would  be  viewed 
as  a  derived  condition.  This  type  of  interpretation,  while  prob- 
ably realistic,  is  avoided  here  as  it  is  fraught  with  opportunities 
for  circular  reasoning. 

Too  little  is  known  of  ELH  characters  in  zeiform  fishes  (Tighe 
and  Keene,  this  volume)  to  enable  a  proper  evaluation  of  the 


460 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Zeiformes  as  a  potential  sister  group  to  the  tetraodontiform 
fishes.  However,  at  least  7.eus  and  Capros  have  a  long  notochord 
tip;  Capros  has  early-forming  spinule-like  scales;  and  Capros 
and  Zeus  (but  not  Antigoma)  larvae  are  generally  similar  to 
some  balistoid  larvae  in  body  shape  and  pigmentation.  Thus, 
based  on  scanty  ELH  information,  there  are  some  suggestions 
of  support  for  Rosen's  proposal  of  a  zeiform-telraodontiform 
relationship. 

Present  knowledge  of  ELH  characters  does  not  help  much  in 
determining  the  inter-ordinal  relationships  of  the  tetraodonti- 
form fishes.  This  is  partially  because  there  are  no  unique,  derived 
ELH  characters  shared  by  all  tetraodontiform  fishes  (see  below). 
In  addition  the  ELH  characters  which  are  shared  between  te- 
traodontiforms  and  either  acanthuroids  or  zeiforms  (above)  are 
shared  with  other  groups  as  well,  thus  lessening  the  value  of 
these  characters  in  determining  relationships:  e.g.,  similar  body 
shape,  pigment,  reduced  number  of  vertebrae,  early-forming 
spinule-like  scales,  and  elongate  notochord  tip  are  found  in  var- 
ious combinations  in  priacanthids,  pomacanthids,  callionymids 
andlophiiform  fishes  (Leis  and  Rennis,  1983,  and  relevant  chap- 
ters in  this  volume). 

Therefore,  one  must  rely  on  ideas  of  inter-ordinal  relation- 
ships based  on  adults.  For  the  purposes  of  this  analysis,  the 
Acanthuroidei  and  the  Zeiformes  are  considered  as  alternative 
sister  groups  for  the  Tetraodontiformes. 

Order  Tetraodontiformes 

There  are  relatively  few  ELH  characters  which  apply  to  the 
Tetraodontiformes  as  a  whole,  and  fewer  still  which  could  be 
considered  derived.  The  only  characters  which  might  be  con- 
sidered derived  are  the  late  formation  of  the  caudal  fin  and  the 
various  early-forming  scale  specializations,  and  both  are  found 
in  a  few  other  percoid  and  non-percoid  groups.  The  dermal  sac 
and  some  other  derived  characters  are  probably  derived  within 
the  order  and  are  of  no  use  in  characterizing  the  order  as  a 
whole.  Other  tetraodontiform  characters  which  are  wide-spread 
among  other  fishes  are:  small  mouth,  gas  bladder  present,  rel- 
atively few  myomeres  and  fin  rays,  large  head,  no  bones  of  the 
head  with  spines,  oviparity,  basically  spherical  eggs  with  un- 
segmented  yolk,  and  transformation  to  an  unspecialized  pelagic 
juvenile  at  a  small  size.  Therefore,  I  could  find  no  uniquely 
derived  ELH  characters  shared  by  all  members  of  the  order. 

Some  features  of  the  adults  can  be  considered  paedomorphic: 
large  head,  lack  of  certain  structures  that  simply  never  form 
(Fraser-Brunner,  1950),  delayed  ossification  of  some  bones. 

Intra-ordinal  relationships 

As  noted  above,  the  Acanthuroidei  (consisting  of  the  families 
Acanthuridae,  Zanclidae  and  Siganidae)  and  the  Zeiformes  (in- 
cluding Caproidae  after  Rosen,  pers.  comm.)  will  be  considered 
as  alternative  sister  groups  to  the  Tetraodontiformes.  Therefore, 
characteristics  shared  with  the  early  life  history  stages  of  the 
Acanthuroidei,  and  particularly  the  Acanthuridae  (or  alterna- 
tively with  the  Zeiformes)  will  be  considered  primitive.  Char- 
acteristics of  acanthuroid  larvae  are  summarized  in  Leis  and 
Rennis  (1983)  and  Leis  and  Richards  (this  volume).  Character- 
istics of  zeiform  larvae  are  summarized  in  Russell  (1976)  and 
Tighe  and  Keene  (this  volume). 

Two  tetraodontiform  families  cannot  be  included  for  lack  of 
information  (Aracanidae  and  Triodonlidae)  and  these  are  not 
considered  further.  I  don't  know  how  seriously  these  omissions 
might  bias  the  results.  It  is  assumed  the  egg  characteristics  of 


the  triacanthodids  (which  are  unknown)  are  the  same  as  those 
of  the  triacanthids. 

Perhaps  surprisingly,  the  acanthuroid  and  zeiform  character 
states  differ  for  only  three  of  the  characters  used  in  the  following 
analysis.  For  these  three,  the  difference  lies  in  my  inability  to 
assign  polarity  to  the  character  if  the  zeiforms  were  chosen  as 
the  outgroup.  Thus,  it  makes  no  difference  to  the  shape  of  the 
resulting  phylogeny  (but  does  weaken  two  of  the  branch  points) 
if  the  Zeiformes  rather  than  the  Acanthuroidei  is  chosen  as 
outgroup. 

A  discussion  of  the  characters  used  follows  (Table  118):  (1) 
Egg  type— Acanthurids  (and  zeiforms)  have  pelagic  eggs,  al- 
though siganids  have  demersal  eggs.  The  demersal  eggs  of  te- 
traodontiform fishes  and  siganids  have  no  adaptations  for  being 
demersal  other  than  stickiness  or  a  mucous  mass,  and  seem 
relatively  unspecialized  for  being  demersal.  A  pelagic  egg  is 
considered  primitive.  (2)  Egg  size— Acanthuroid  eggs  are  small 
(<  1  mm),  so  eggs  larger  than  1.4  mm  are  considered  derived. 
However,  zeiform  eggs  are  medium  to  large  (0.95-2.0  mm),  so 
if  zeiform  fishes  are  accepted  as  the  outgroup,  polarity  of  this 
character  cannot  be  determined.  (3)  Oil  droplets  in  eggs— Acan- 
thuroid eggs  (and  zeiform  eggs)  have  one  or  more  oil  droplets 
in  the  yolk.  Lack  of  oil  droplets  in  eggs  is  considered  derived. 
(4)  Egg  shape— An  egg  that  is  not  spherical  is  considered  derived 
because  acanthuroid  eggs  (and  zeiform  eggs)  are  spherical.  (5) 
Chorion  sculpture  — Sculpturing  on  the  chorion  is  considered 
derived  because  acanthuroid  eggs  (and  zeiform  eggs)  are  un- 
sculptured.  (6)  Incubation  period— Acanthuroid  eggs  hatch  in 
about  two  days  or  less,  and  an  incubation  time  longer  than  this 
is  considered  derived.  Because  incubation  period  is  tempera- 
ture-dependent, it  is  possible  that  some  of  the  differences  noted 
here  are  artifacts  of  the  different  temperatures  at  which  the  eggs 
were  reared.  However,  insofar  as  it  has  been  possible  to  com- 
pare different  taxa  reared  at  similar  temperatures,  the  differences 
in  incubation  period  noted  here  seem  valid.  Incubation  times 
of  zeiform  eggs  are  poorly  known,  but  may  be  up  to  13  days  for 
Zeus.  Therefore,  if  zeiform  fishes  are  accepted  as  the  outgroup, 
polarity  of  this  character  cannot  be  determined.  (7)  Parental 
care  of  eggs— There  is  no  parental  care  of  eggs  by  fishes  with 
pelagic  eggs  including  zeiforms  and  acanthurids.  Siganids  lay 
demersal  eggs  but  no  parental  care  has  been  reported.  Therefore, 
lack  of  parental  care  is  considered  primitive.  (8)  Body  shape— 
Acanthuroid  (and  zeiform)  larvae  tend  to  be  cylindrical  to  some- 
what compressed  at  hatching  and  to  be  compressed  by  the  time 
flexion  is  complete,  although  they  may  pass  through  an  early 
preflexion  stage  which  is  more  rotund.  This  developmental  pat- 
tern is  considered  primitive.  Some  tetraodontiform  larvae  are 
extremely  rotund  throughout  development,  but  this  is  largely 
due  to  a  greatly  inflated  dermal  sac  (see  character  10).  (9)  Head 
and  gut  development  — All  balistoid  fishes  but  ostraciids  hatch 
with  a  cylindrical  to  compressed  body.  All  of  these  but  mona- 
canthid  Morph  AB  become  deeper-bodied  and  wider  in  head 
and  gut  by  the  middle  of  the  preflexion  stage  and  then  become 
compressed  by  flexion.  Morph  AB  monacanthids  never  become 
broad  in  head  and  gut.  Due  to  the  widespread  occurrence  of  the 
wide  body  development  mode  in  the  suborder,  it  is  considered 
primitive.  (10)  Vesicular  dermal  sac— Some  tetraodontiform 
larvae  have  the  head  and  trunk  enclosed  in  a  vesicular  dermal 
sac,  a  condition  not  found  in  acanthuroids  or  zeiforms  (a  very 
weakly-developed  dermal  sac  without  vesicles  similar  to  the  one 
of  yolk-sac  balistids  is  found  in  acanthurids).  This  sac  and  its 
subdermal  space  seem  to  be  the  source  of  many  of  the  dermal 


LEIS:  TETRAODONTIFORMES 


461 


Table  1 18.    Early  Life  History  Characteristics  of  the  Tetraodontiform  Fishes.  (P)  indicates  primitive,  and  (D)  derived.  (?)  indicates 

assumed,  (s)  indicates  that  character  is  secondarily  m  state  given.  (  — )  indicates  not  applicable  for  family.  See  text  for  discussion  of  characteristics. 

(*)  indicates  character  for  which  polarity  cannot  be  established  if  the  Zeiformes  is  considered  the  sister  group  of  the  Tetraodontiformes. 


Taxon 

Mona- 

Mona- 

Tnacantho- 

Tnacanthi- 

Bahsti- 

canthidae 

canthidae 

Telra- 

Diodonli- 

Characler 

didae 

dae 

dac 

A  B 

C 

Ostraciidae 

odontidae 

dae 

Molidae 

1. 

Egg  type 

P? 

P 

D 

D 

D 

P 

D 

P 

P 

*2. 

Egg  size 

P? 

P 

P 

P 

P 

D 

P 

D 

D 

3. 

Oil  droplets 

D? 

D 

P 

P 

P 

P 

P 

P 

P 

4. 

Egg  shape 

P? 

P 

P 

P 

P 

D 

P 

P 

P 

5. 

Chorion  sculpture 

P? 

P 

P 

P 

P? 

D 

P 

P 

P 

*6. 

Incubation  period 

P? 

P 

P 

P 

P? 

D 

D 

D 

D 

7. 

Parental  care  of  eggs 

P? 

P 

D 

P 

P? 

P 

P-D 

P 

P 

8. 

Body  shape 

P 

P 

P 

P 

P 

D 

D 

D 

D 

9. 

Head  and  gut  development 

P 

P 

P 

D 

P 

— 

— 

— 

— 

10a. 

Vesicular  dermal  sac 

P 

P 

P 

P 

P 

D 

D 

D 

D 

10b. 

Dermal  sac  inflation 

— 

— 

— 

— 

— 

D 

P 

D 

D 

11a. 

Opercular  pore  A 

P? 

P 

P 

P 

P 

D 

D 

D 

D 

lib. 

Opercular  pore  B 

— 

— 

— 

— 

— 

D 

P 

D 

D 

12. 

Scale  development 

P 

P 

P 

P 

P 

D 

D 

D 

D 

13. 

Very  large  spines 

P 

P 

P 

P 

P 

P 

P 

P 

D 

14. 

Preopercular  cluster 

P 

P 

D 

D 

Ps 

P 

P 

P 

P 

15. 

Long  notochord  tip 

P 

P 

P 

P 

P 

P 

D 

D 

D 

16. 

Dorsal  spine  development 

D 

D 

P 

P 

Ds 

— 

— 

— 

— 

17a. 

Dorsal  spines  A 

P 

P 

D 

D 

D 

D 

D 

D 

D 

17b. 

Dorsal  spines  B 

P 

P 

P 

D 

D 

— 

— 

— 

— 

17c. 

Dorsal  spines  C 

P 

P 

P 

P 

P 

D 

D 

D 

D 

18. 

Dsp,  P,sp  formation 

D 

P 

P 

P 

P 

— 

— 

— 

— 

19. 

Pelvic  fin 

P 

P 

D 

D 

D 

D 

D 

D 

D 

20. 

Pelvis 

P 

P 

P 

P 

P 

D 

D 

D 

D 

*21a. 

Caudal  fin  rays  (£  1  1) 

P 

P 

P 

P 

P 

D 

D 

D 

D 

21b. 

Caudal  fin  rays  (s  10) 

P 

P 

P 

P 

P 

D 

P 

D 

D 

21c. 

Caudal  fin  rays  (none) 

P 

P 

P 

P 

P 

P 

P 

P 

D 

22. 

Pectoral  development 

P 

P 

P 

P 

P 

D 

D 

D 

D 

23. 

Body  width 

P 

P 

P 

P 

P 

P 

P 

D 

P 

specializations  of  the  Tetraodontiformes.  A  dermal  sac  is  con- 
sidered derived  (10a).  A  strongly  inflated  dermal  sac  (with  a 
large  subdermal  space)  linking  head  and  trunk  in  a  ball-like  unit 
is  considered  a  further  derivation  from  the  presence  of  a  sac 
(lOb).  (II)  Restriction  of  gill  opening  to  a  pore  — In  perciforms 
with  restricted  gill  openings  (no  zeiform  fishes  have  restricted 
gill  openings),  the  opercles  are  fully  open  in  early  larvae  and 
gradually  close  to  a  pore.  The  assumed  primitive  condition  in 
Tetraodontiformes  is  for  closure  to  a  pore  to  occur  after  some 
rays  of  the  median  fins  have  ossified  (11a)  because  this  is  closest 
to  the  perciform  condition.  Having  the  opening  closed  to  a  pore 
at  hatching  is  considered  derived  from  closure  at  the  end  of  the 
yolk-sac  stage  (lib).  (12)  Scale  development— The  specialized 
scales  of  adult  tetraodontiform  fishes  form  in  two  ways:  directly 
or  by  first  passing  through  a  relatively  unspecialized  spinule 
stage.  The  intermediate  spinule  stage  is  considered  primitive 
because  it  is  present  in  acanthurids  and  at  least  Capros  in  the 
zeiforms.  In  acanthurids  small  spine-like  scales  change  into  tri- 
angular scales  arrayed  in  vertical  rows,  and  these  spine-like 
scales  form  first  on  the  lower  gut  and  laterally  on  the  head,  in 
the  same  place  they  first  form  in  most  tetraodontiform  fishes 
which  have  them.  In  Capros  small  spinules  form  prior  to  flexion 
and  eventually  cover  the  whole  body  before  differentiating  into 
scales.  (13)  Very  large  dermal  spines  (larval  specialization)— 
The  ver>'  large,  fixed  spines  with  conical  or  pyramidal  bases  of 
molids  are  unique  and  are  considered  derived.  ( 1 4)  Preopercular 
spinule  cluster  (larval  specialization)— This  cluster  of  spinules 


is  unique  to  balistids  and  most  monacanthids  and  is  considered 
a  derived  character.  (15)  Long  notochord  tip— The  notochord 
may  extend  well  past  the  caudal  fin  aniage,  and  if  so,  following 
flexion  it  will  protrude  dorsal  and  parallel  to  the  caudal  rays  (to 
about  '/:  their  length)  for  a  time.  This  condition  was  initially 
considered  derived  because  it  is  absent  in  acanthuroid  fishes. 
However,  in  the  Tetraodontiformes  the  long  notochord  tip  is 
absent  only  in  taxa  in  which  the  tail  becomes  greatly  reduced 
(i.e.,  Diodontidae,  Ostraciidae,  Molidae).  Therefore,  it  seems 
better  to  regard  the  long  notochord  tip  as  a  primitive  characler 
within  the  order,  but  a  character  derived  after  the  supposed  split 
from  the  acanthuroid  fishes.  The  absence  of  this  structure  within 
the  order  is  thus  derived.  Zeiform  larvae  (Zeus.  Capros)  have 
an  elongate  notochord  tip  very  similar  to  that  of  balistoid  fishes, 
so  this  is  considered  the  primitive  condition.  (16)  Dorsal  fin 
spine  development  sequence— Dorsal  fin  spine  development 
prior  to  dorsal  fin  soft  ray  development  is  considered  primitive 
because  this  is  the  condition  in  acanthuroids  and  zeiforms.  ( 1 7) 
Dorsal  fin  spines— Acanthuroid  fishes  have  4  to  14  dorsal  fin 
spines,  and  zeiform  fishes  5  to  10.  Therefore,  in  the  tetraodon- 
tiform fishes,  the  most  primitive  character  state  is  the  greatest 
number  of  spines  (i.e.,  4-6  of  triacanthodids  and  triacanthids). 
The  intermediate  derived  condition  is  a  reduction  in  this  num- 
ber to  three  spines  ( 1 7a).  From  the  intermediate  condition  are 
derived  one  or  two  spines  ( I  7b)  and  no  spines  ( I  7c).  ( 1 8)  Initial 
formation  of  fin  spines— The  presence  of  dorsal  fin  spine  aniage 
and  pelvic  fin  buds  prior  to  flexion  in  fishes  that  have  late- 


462 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


< 

Q 

UJ 

< 

Q 

J- 

U 

o 

2 

< 

< 

1- 

O 
Q 
O 

Q 

_1 
O 

o 

Q 

Z 

13.21c 


2.IOb,llb.l5.2lb 


TETRAODONTOIDEA 
6. 8.10a. I  la. 12. 17c, 20. 21a.  22 


TRIACANTHOIDEI 


TETRAODONTOIDEI 


TETRAODONTIFORMES 


Fig.  252.  Phylogeny  of  tetraodontiform  fishes  based  on  early  life  history  characters  (excluding  Aracanidae  and  Triodontidae  for  which  no 
information  is  available).  Numbers  refer  to  characters  (see  text)  and  are  located  on  the  branch  which  possesses  the  derived  state.  Characters  2,  6 
and  21a  would  be  deleted  if  the  Zeiformes  is  accepted  as  the  sister  group  of  the  Tetraodontiformes.  Note  that  character  1  occurs  m  two  places 
indicating  conflict  with  the  accepted  classification. 


forming  dorsal  spines  (character  16)  is  considered  primitive 
because  it  is  closer  to  the  presumed  ancestral  (i.e.,  acanthuroid 
and  zeiform)  condition  of  early-forming  dorsal  and  pelvic  spines 
than  is  late  formation  oftheanlage.  (19)  Pelvic  fin  — Acanthuroid 
(and  zeiform)  fishes  have  a  pelvic  fin  formula  of  at  least  I,  3. 
Presence  of  a  pelvic  fin  is  primitive,  and  its  absence  is  derived. 
(20)  Pelvis— The  pelvis  is  present  in  acanthuroid  (and  zeiform) 
fishes,  and  its  absence  is  considered  derived.  (21)  Caudal  fin 
rays- Acanthuroids  have  16-17  principal  caudal  fin  rays.  The 
maximum  number  (i.e.,  1 2)  in  tetraodontiform  fishes  is  consid- 
ered primitive.  The  intermediate  derived  condition  is  <  1 1  rays 
(2 1  a).  The  next  most  advanced  condition  is  <  1 0  rays  (21b),  and 
the  most  advanced  condition  is  the  complete  absence  of  the 
caudal  fin  (21c).  Zeiform  fishes  have  1 1-15  principal  caudal  fin 
rays,  so  if  zeiforms  are  accepted  as  the  outgroup,  polarity  of  2  la 
cannot  be  determined,  while  21b  and  21c  would  not  change. 
(22)  Pectoral  fin  development— The  pectoral  fin  in  acanthuroid 
(and  probably  zeiform)  fishes  develops  after  or  simultaneously 
with  the  dorsal  and  anal  fin  soft  rays,  and  this  is  considered 
primitive.  (23)  Body  width  — The  condition  of  body  width  > 
body  depth  found  in  the  Diodontidae  is  unique  in  the  Tetra- 
odontiformes and  is  considered  derived. 

Phylogenetic  Analysis 

Relationships  within  the  Tetraodontiformes  based  on  ELH 
characters  are  presented  in  Fig.  252.  In  the  following  section,  I 


will  contrast  the  present  phylogeny  with  three  phylogenies  based 
on  adult  characters  (Fig.  253):  myology  (Winterbottom,  1 974a), 
external  and  internal  characters  (Tyler,  1980)  and  osteology 
(Rosen,  pers.  comm.).  Lauder  and  Liem's  (1983)  review  of  in- 
terrelationships of  tetraodontiform  fishes  depends  heavily  on 
Winterbottom's  (1974a)  work  and,  for  my  purposes  here,  is 
identical  to  his  phylogeny.  Therefore,  Lauder  and  Liem's  (1983) 
phylogeny  will  not  be  considered  separately.  The  ELH-based 
phylogeny  exactly  matches  none  of  the  three  adult-based  schemes, 
but  is  closest  to  Rosen's  (pers.  comm.),  differing  only  in  place- 
ment of  the  Tetraodontidae.  Two  cautions  should  be  kept  in 
mind:  1)  Rosen's  (pers.  comm.)  study  is  primarily  concerned 
with  inter-ordinal  relationships,  and  the  portion  dealing  with 
intra-ordinal  relationships  of  the  Tetraodontiformes  is  based  on 
relatively  few  characters;  and  2)  the  present  phylogeny  has  lim- 
itations flowing  from  exclusion  of  two  families  and  many 
subfamilial  taxa  due  to  lack  of  information. 

There  is  most  agreement  between  the  four  phylogenies  in  the 
question  of  the  relationship  of  the  triacanthodids  (Figs.  252, 
253).  The  present  phylogeny  and  those  of  Winterbottom  (1974a) 
and  Rosen  (pers.  comm.)  agree  in  the  erection  of  the  suborder 
Triacanthoidei  as  the  sister  group  to  all  other  tetraodontiform 
fishes.  Tyler  (1980)  includes  the  triacanthodids  in  the  balistoid 
line,  but  this  is  a  result  of  philosophy  of  classification  more  than 
anything  else  (Tyler,  pers.  comm.). 

The  four  phylogenies  are  evenly  divided  on  the  question  of 


LEIS:  TETRAODONTIFORMES 


463 


TYLER 


WINTERBOTTOM 


I   TRIACANTHODIDAE 

2  TRIACANTHIDAE 

3  BALISTIDAE 

4  MONACANTHIDAE 


5  OSTRACIIDAE 

6  TETRAODONTIDAE 

7  DIODONTIDAE 

8  MOLIDAE 


Fig.  253.  The  two  published  adult-based  phylogenies  of  tetraodon- 
tiform  fishes  which  were  tested  by  the  ELH-based  phylogeny.  These 
phylogenies  were  modified  by  omitting  the  two  families  which  could 
not  be  included  in  the  ELH-based  phylogeny.  After  Tyler  (1980)  and 
Winterbottom  (1974a).  Numbers  refer  to  the  families  listed  at  bottom. 
Rosen's  unpublished  phylogeny  is  not  shown. 


ostraciid  relationships  (Figs.  252,  253),  indicating  that  further 
study  is  required.  Winterbottom  ( 1 974a)  and  Tyler  ( 1 980)  place 
the  ostraciids  with  the  balistoids,  a  view  reinforced  by  a  recent 
reassessment  of  their  data  (Winterbottom  and  Tyler,  1983).  The 
present  phylogeny  and  Rosen's  (pers.  comm.),  however,  place 
the  ostraciids  with  the  tetraodontoids.  A  relationship  between 
ostraciids  and  tetraodontoids  was  suggested  by  Sakamoto  and 
Suzuki  (1978)  based  on  general  similarity  of  larvae. 

The  three  adult-based  phylogenies  regard  the  tetraodontids 
as  the  sister  group  of  the  diodontids  (Fig.  253).  This  differs 
significantly  from  the  ELH-based  phylogeny  (Fig.  252)  which 
regards  the  tetraodontids  as  the  sister  group  of  all  other  tetra- 
odontoids (including  ostraciids).  The  trichotomy  between  these 
"other  tetraodontoids"  in  Fig.  252  cannot  be  resolved  at  present. 
Further  study  is  indicated. 

The  balistoids  (Monacanthidae-Balistidae)  branch  off  in  a 
convincing  manner,  but  not  without  problems.  The  phylogeny 
as  depicted  in  Fig.  252  requires  that  demersal  eggs  (1)  be  in- 
dependently derived  in  balistoids  and  tetraodontids.  Although 
this  is  quite  possible,  it  brings  into  question  the  validity  of  using 
demersal  eggs  as  a  derived  character  to  define  the  Balistoidea. 
Morph  C  monacanthids  lack  the  preopercular  spine  cluster  (14) 


which  characterises  all  other  balistoids.  I  conclude  that  this  is 
a  secondary  loss  and  that  the  delayed  development  of  the  dorsal 
fin  spine  in  Morph  C  is  independently  derived  (thus  not  indi- 
cating a  relationship  with  triacanthoids). 

All  phylogenies  agree  on  the  close  relationship  of  monacan- 
thids and  balistids.  Indeed,  in  the  present  study  (Fig.  252),  they 
were  separated  by  only  two  ELH  characters,  (17b)  loss  of  a  fin 
spine,  and  (7)  parental  care  of  eggs,  about  which  there  is  little 
information  and  which  is  variable  in  tetraodontids.  Although 
the  present  phylogeny  is  nominally  consistent  with  Matsuura's 
(1979)  phylogeny,  Winterbottom  (1974a)  considered  monacan- 
thids and  balistids  to  be  subfamilies,  and  the  ELH-based  phy- 
logeny presented  here  has  done  little  to  clarify  this  conflict. 

There  is  some  indication  from  ELH  characters  of  divergences 
within  families,  but  the  amount  that  can  be  said  is  severely 
limited  by  the  small  number  of  taxa  for  which  ELH  characters 
are  known.  The  diodontids  seem  very  conservative  but  some 
species  of  Chilomycterus  have  a  specialized  ontogenetic  stage 
between  larvae  and  juveniles  (" Lyosphaera"):  this  supports  re- 
moval of  these  species  to  a  separate  genus  (study  in  progress). 
Within  the  ostraciids,  the  two  subfamilies  are  separated  by  de- 
gree of  chorion  ornamentation,  and  to  a  lesser  degree  by  de- 
velopment at  hatching.  The  specialized  "Molacanthus"  stage 
separates  Afola  and  Masturus  from  Ranzania  in  the  Molidae. 
Balistids  seem  very  conservative  in  development.  Tetraodontids 
vary  greatly  in  development  at  hatching,  parental  care  of  eggs, 
and  perhaps  in  a  number  of  other  characters.  Too  few  taxa  are 
known  within  the  triacanthodids  and  triacanthids  for  any  state- 
ments to  be  made  here.  Monacanthids  have  the  most  variation 
in  ELH  characters  within  the  order,  some  of  which  has  already 
been  referred  to  (Aboussouan  and  Leis,  this  volume).  There 
seems  to  be  a  great  deal  of  potential  in  the  use  of  ELH  characters 
for  phylogenetic  studies  in  the  Monacanthidae,  but  first,  devel- 
opmental series  for  more  species  and  genera  must  be  established. 
I  have  attempted  to  use  ELH  characters  independently  as  a 
test  of  phylogenies  based  on  adult  characters.  Where  the  two 
types  of  phylogenies  support  each  other,  confidence  in  the  phy- 
logeny is  increased.  Where  differences  appear,  further  study,  or 
re-interpretation  of  existing  data  is  called  for  to  resolve  the 
differences. 

In  conclusion,  the  present  classification  should  be  viewed  with 
caution  because  there  are  relatively  few  taxa  for  which  eady  life 
history  information  is  available.  Monophyly  of  the  tetraodon- 
tiform  fishes  could  not  be  established  using  ELH  characters. 
The  present  ELH-based  phylogeny  and  those  of  Winterbottom 
(1974)  and  Rosen  (pers.  comm.)  agree  in  the  creation  of  a  sep- 
arate suborder  for  triacanthoid  fishes;  Tyler  (1980)  disagrees 
with  this  placement.  Tyler  (1980)  and  Winterbottom  (1974a) 
agree  in  placing  the  Ostraciidae  in  the  Balistoidea,  in  contrast 
to  inclusion  of  the  Ostraciidae  within  the  Tetraodontoidea  as 
proposed  here  and  by  Rosen  (pers.  comm.).  My  placement  of 
the  Tetraodontidae  is  in  conflict  with  previous  phylogenies  based 
on  adult  characters.  In  other  areas,  the  ELH-based  phylogeny 
is  in  agreement  with  the  three  adult-based  phylogenies.  The 
different  placements  of  the  Tetraodontidae  and  in  particular  the 
Ostraciidae  in  the  present  classification  warrant  further  inves- 
tigation of  tetraodontiform  interrelationships. 

Department  of  Ichthyology,  The  Australian  Museum,  P.O. 
Box  A285,  Sydney,  N.S.W.,  2000,  Australia. 


Percoidei:  Development  and  Relationships 
G.  D.  Johnson 


AS  the  largest  and  most  diverse  of  the  perciform  suborders, 
the  Percoidei  exemplifies  the  inadequacies  that  charac- 
terize perciform  classification.  Regan  (1913b)  defined  the  Per- 
coidei "by  the  absence  of  the  special  peculiarities  which  char- 
acterize the  other  suborders  of  the  Percomorphi  [=Perciformes]," 
and  seventy  years  of  research  in  systematic  ichthyology  have 
failed  to  produce  a  more  meaningful  definition.  In  the  absence 
of  even  a  single  shared  specialization  uniting  the  percoids,  the 
monophyly  of  this  great  assemblage  of  fishes  is  doubtful.  In  spite 
of  our  inability  to  adequately  define  the  Percoidei,  or  because 
of  it,  half  of  the  approximately  145  families  of  perciform  fishes 
are  usually  referred  to  this  suborder.  Greenwood  et  al.  (1966) 
listed  71  percoid  families  in  their  "highly  tentative"  familial 
classification  of  the  Perciformes,  and  Nelson  ( 1976)  stated  that 
the  Percoidei  contains  72  families,  595  genera  and  about  3,935 
species. 

Percoids  are  best  represented  in  the  nearshore  marine  envi- 
ronment and  form  a  significant  component  of  the  reef  associated 
fish  fauna  of  tropical  and  subtropical  seas.  A  few  groups  are 
primarily  epipelagic  or  mesopelagic.  Association  with  brackish 
water  occurs  in  many  nearshore  marine  families,  some  of  which 
have  one  or  more  exclusively  freshwater  members,  but  only 
four  families  are  primarily  restricted  to  freshwaters,  the  north 
temperate  Percidae  and  Centrarchidae,  the  south  temperate  Per- 
cichthyidae  (with  one  brackish  water  species)  and  the  tropical 
Nandidae. 

In  a  practical  sense,  the  suborder  Percoidei  serves  the  Per- 
ciformes in  much  the  same  capacity  as  the  Serranidae  once 
served  the  Percoidei  itself  as  a  convenient  repository  for  those 
"generalized"  perciform  families  that  cannot  obviously  be  placed 
elsewhere.  I  have  treated  the  percoids  in  a  similar  sense  here, 
one  of  practicality  and  convenience.  1  do  not  intend  to  imply 
or  formulate  hypotheses  about  the  monophyly  of  the  Percoidei 
or  to  consider  their  intrarelationships  as  a  whole.  My  major 
objectives  are  to  provide  some  preliminary  documentation  of 
the  variability  of  a  number  of  character  complexes  among  adults 
and  larvae  of  those  fishes  we  now  call  percoids,  to  suggest  what 
1  believe  to  be  promising  avenues  of  future  research  and  to  offer 
some  specific  examples  illustrating  the  utility  of  larval  mor- 
phology in  elucidating  percoid  phylogeny. 

Classification 

As  here  defined  (Table  1 1 9)  the  Percoidei  includes  80  families 
and  1 2  incertae  sedis  genera,  making  it  by  far  the  largest  and 
most  diverse  suborder  of  teleostean  fishes.  The  overall  limits  of 
the  suborder  are  only  slightly  modified  from  Greenwood  et  al. 
(1966).  The  Pomacentridae,  Embiotocidae  and  Cichlidae  are 
excluded  because  of  their  recent  placement  in  the  Labroidei  by 
Kaufman  and  Liem  (1982).  The  suborder  Acanthuroidea  is 
treated  separately  in  this  volume,  but  a  recent  hypothesis  (Mok 
and  Shen,  1983),  with  which  1  concur,  based  on  additional  evi- 
dence, suggests  a  close  relationship  between  acanthuroids  and 
the  Scatophagidae.  The  affinities  of  the  questionably  monophy- 
letic  Nandidae  remain  unresolved  (Lauder  and  Liem,  1983), 
and  although  the  nandids  are  provisionally  included  in  my  list 


of  percoid  families,  they  were  not  considered  in  the  larval  and 
adult  tables.  The  genus  Elassoma.  formerly  a  member  of  the 
family  Centrarchidae,  is  excluded  from  the  Percoidei,  for  rea- 
sons discussed  below.  The  monophyly  of  the  suborder  Trachi- 
noidei,  as  defined  by  Greenwood  et  al.  (1966)  is  suspect,  and 
the  affinities  of  families  such  as  the  Mugiloididae,  Percophidae, 
Chiasmodontidae  and  others  may  lie  with  the  percoids.  How- 
ever, these  families  are  treated  elsewhere  in  this  volume,  and 
of  the  "trachinoids,"  only  the  Opistognathidae  are  here  included 
as  percoids. 

Although  the  overall  limits  of  the  Percoidei  are  similarly  per- 
ceived in  my  classification  and  that  of  Greenwood  et  al.  (1966), 
substantive  discrepancies  result  from  differences  in  concepts  of 
family  limits.  For  example,  my  Serranidae  (Johnson,  1983)  in- 
cludes the  Pseudogrammidae  and  Grammistidae  of  Greenwood 
et  al.  (1966).  Leptohrama  is  treated  as  a  monotypic  family  sep- 
arate from  the  Pempherididae  (Tominaga,  1965),  epigonids  are 
treated  as  a  separate  family,  etc.  The  high  percentage  of  mono- 
typic families  that  has  historically  characterized  percoid  clas- 
sification is  a  disturbing  but  unavoidable  problem  that  can  only 
be  remedied  with  a  better  understanding  of  percoid  intrarela- 
tionships. In  my  classification  (Table  1 19),  26  of  the  80  families 
are  monotypic  and  12  genera,  which  lack  family  names,  are 
retained  incertae  sedis.  Families  and  incertae  sedis  genera  are 
arranged  alphabetically  for  easy  reference  and  to  avoid  any  in- 
ference of  affinity  based  on  sequence.  The  classification  of 
Springer  (1982)  was  followed  for  most  families  treated  by  him 
and  otherwise  that  of  Nelson  ( 1 976).  Below,  1  discuss  differences 
between  my  classification  and  that  of  Spnnger  ( 1 982)  or  that  of 
Nelson  (1976),  and  present  some  new  information  about  fa- 
milial relationships.  Early  life  history  information  contributed 
substantially  to  some  of  these  modifications. 

Acropomatidae  and  Symphysanodon— The  "oceanic  per- 
cichthyids"  of  Gosline  (1966)  do  not  share  the  defining  char- 
acteristics of  the  Percichthyidae  (see  below),  and  are  treated  here 
as  a  separate  family,  including  the  following  genera— v^cropowa, 
Apogonops.  Doederleinia  (=Rhomhoscrranus),  Malakichthys, 
Neoscombrops.  Synagrops  and  V'erilus.  I  know  of  no  synapo- 
morphy  that  unites  the  acropomatids,  and  further  work  will  be 
necessary  to  test  their  monophyly.  Larvae  of  four  genera  are 
known.  Those  of  Acropoma  (Fig.  254C),  Doederleinia  (Fig.  254D) 
and  Malakichthys  are  quite  similar,  but  those  of  Synagrops  {Fig. 
254B)  differ  in  pigmentation,  body  form,  and  the  presence  of 
more  extensive  head  spination.  Although  the  larvae  of  Sym- 
physanodon (Fig.  254A)  are  unique  in  their  possession  of  horn- 
like frontal  spines,  they  are  otherwise  remarkably  similar  to 
those  of  Synagrops  (Fig.  2548),  suggesting  that  these  two  genera 
may  be  closely  related. 

Callanthiidae  and  Grammatidae.— Springer  (1982)  noted  that 
"there  is  little  evidence  to  unite"  the  five  genera  he  included  in 
the  family  Grammatidae.  I  concur  with  this  and  treat  two  of 
these  genera,  Callanthias  and  Grammatonotus  as  a  distinct  fam- 
ily, the  Callanthiidae  (currently  under  revision  in  collaboration 


464 


JOHNSON:  PERCOIDEI 


465 


with  W.  D.  Anderson).  Callanthiids  share  a  flat  nasal  organ 
without  laminae,  a  lateral  line  that  runs  along  the  base  of  the 
dorsal  fin,  ending  near  its  terminus  or  continuing  along  the 
dorsolateral  margin  of  the  caudal  peduncle,  and  a  midlaleral 
row  of  modified  scales  that  bear  a  series  of  pits  and/or  grooves. 
The  larvae  of  these  two  genera  appear  dissimilar  (Fig.  255E,  F), 
but  specimens  of  Grammatonotus  smaller  than  1 3  mm  are  un- 
known. Stiginatonolus  (based  on  a  small,  now  lost  specimen) 
was  reported  to  have  three  opercular  spines,  and  probably  rep- 
resents a  larval  or  juvenile  anthiine  serranid.  The  family  Gram- 
matidae,  as  considered  here,  contains  only  Gramma  and  Li- 
pogramma. 

Carangidae,  Coryphaenidae,  Echeneididae,  Rachycentridae  and 
Nematistiidae.  — See  discussion  on  utility  of  larval  morphology. 

Coracinidae,  Drepanidae  and  Ephippididae.  — The  family 
Ephippididae,  as  defined  here,  contains  the  following  genera: 
Chaetodipterus.  Ephippus.  Parapsetttis.  Platax.  Proteracanthus. 
Rhinoprenes  and  Tripterodon.  Ephippidids  exhibit  considerable 
diversity  in  several  features  that  are  more  commonly  conser- 
vative among  percoids,  such  as  scale  morphology  and  the  struc- 
ture and  arrangement  of  median  fin  supports  and  predorsal 
bones.  Nonetheless,  monophyly  of  the  family  is  supported  by 
shared  specializations  of  the  gill  arches  that  include  reduction 
or  absence  of  the  basihyal,  absence  of  the  interarcual  cartilage, 
a  relatively  large  first  pharynogobranchial  and,  most  notably,  a 
peculiar  comblike  series  of  large  blunt  rakers  loosely  associated 
with  the  anterior  margin  of  the  broadened  first  epibranchial. 
Springer  (1982;  pers.  comm.),  following  some  previous  authors 
(Jordan,  1923;  Golvan,  1965)  included  Parapsettus  in  the  Scor- 
pididae.  Rhinoprenes  was  previously  treated  as  a  monotypic 
family,  possibly  related  to  the  Scatophagidae  (Munro,  1967), 
and  Proteracanthus  as  a  girellid  (Norman,  1966).  Although 
Drepane  may  be  related  to  the  ephippidids,  it  does  not  share 
the  branchial  specializations  described  above,  and  lacking  fur- 
ther evidence  of  a  direct  relationship,  I  treat  it  separately.  Based 
on  other  features  of  the  gill  arches  a  close  relationship  between 
Drepane  and  Coraciniis  seems  likely.  In  both  genera  the  basihyal 
is  embedded  in  thick  connective  tissue  and  is  tightly  bound 
along  the  anteroventrally  sloping  median  junction  of  the  hy- 
pohyals.  In  addition,  an  unusual  moveable  articulation  between 
the  hypohyals  and  the  anterior  ceratohyal  allows  for  dorsoven- 
tral  rotation  of  the  ceratohyal.  Pending  further  investigation  of 
the  possible  affinities  of  these  two  genera,  I  retain  them  as  mono- 
typic families.  Larval  morphology  could  provide  important  in- 
formation in  resolving  the  relationships  among  the  five  ephip- 
pidid  genera,  Drepane  and  Coraciniis,  but  to  date,  only  the 
larvae  of  Chaetodipterus  have  been  described  (Fig.  256G). 

Elassoma— In  an  extensive  comparison  of  the  acoustico-lat- 
eralis  system  of  the  Centrarchidae,  Branson  and  Moore  (1962) 
placed  the  pygmy  sunfishes,  genus  Elassoma.  in  a  separate  fam- 
ily, based  on  over  20  "major  characteristics."  These  include 
numerous  reductions  in  the  laterosensor>'  system  (e.g.,  absence 
of  a  lateral-line  canal  on  the  body,  absence  of  all  infraorbitals 
except  the  lacrimal,  absence  of  the  mandibular  and  angular 
lateralis  canals,  etc.),  presence  of  numerous  free  neuromasts  of 
a  distinctive  form,  rudimentary  olfactory  organ,  gill  membranes 
broadly  united  across  the  isthmus,  rounded  caudal  fin,  and  cy- 
cloid scales.  To  these,  I  add  the  following  reductive  features  of 
Elassoma,  not  shared  by  the  Centrarchidae:  basisphenoid  ab- 
sent; endopterygoid  absent;  ectopterygoid  absent  or  fused  to 


Table  1 1 9.    List  of  the  Families  and  incertae  sedis  Genera  of  the 
Suborder  Percoidei.  *  Families  with  a  single  genus. 


Acanthoclinidae 

Acropomatidae 

Ambassidae 

Aplodactylidae 

Apogonidae 

Arripididae* 

Banjosidae* 

Balhyclupeidae* 

Bramidae 

Caesionidae 

Caesioscorpis 

Callanlhiidae 

Carangidae 

Caristiidae* 

Cenlracanthidae 

Centrarchidae 

Cenlrogenysidae* 

Centropomidae 

Cepolidae 

Chaetodontidae 

Cheilodactylidae 

Chironemidae 

Cirrhitidae 

Congrogadidae 

Coracinidae* 

Coryphaenidae* 

Datnioides 

Dinolestidae* 

Dmoperca 

Drepanidae* 

Echeneididae 


Emmelichthyidae 

Enoplosidae* 

Ephippididae 

Epigonidae 

Gerreidae 

Giganthiidae* 

Girellidae 

Glaucosomatidae* 

Grammatidae 

Haemulidae 

Hapalogenys 

Hemiliiljanus 

Howella 

Inermiidae 

Kuhliidae* 

Kyphosidae 

Lactariidae* 

Lateolahrax 

Latrididae 

Leiognathidae 

Leplobramidae* 

Lethrinidae 

Lobotidae* 

Lutjanidae 

Malacanthidae 

Menidae* 

Microcanthidae 

Monodactylidae* 

Moronidae 

Mullidae 

Nandidae 


Nematistiidae* 

Nemipteridae 

Neoscorpis 

Opistognathidae 

Oplegnathidae* 

Ostracoberycidae* 

Parascorpididae* 

Pempherididae 

Pentacerotidae 

Percichthyidae 

Percidae 

Plesiopidae 

Pomacanthidae 

Pomatomidae* 

Polypnon 

Priacanthidae 

Pseudochromidae 

Rachycentridae* 

Scatophagidae 

Sciaenidae 

Scombropidae* 

Scorpididae 

Serranidae 

Sillaginidae 

Simperca 

Sparidae 

Stereolepis 

Symphysanodon 

Terapondiae 

Toxotidae* 


palatine;  palatine  with  a  single  notch-like  articulation  with  eth- 
moid cartilage;  predorsals  usually  absent,  a  single  bone  present 
in  some  (vs.  3-7  in  centrarchids);  branchiostegals  5  (vs.  6-7); 
principal  caudal  rays  6-7  -I-  7-8  (vs.  9  +  8);  hypurals  1-2  and 
3-4-5  fused. 

Branson  and  Moore  (1962)  concluded  that  "either  the  elas- 
somids  diverged  from  the  centrarchid  stock  early  in  the  history 
of  the  group  or  they  have  entirely  different  affinities."  Subse- 
quent classifications  (Greenwood  et  al.,  1966;  Nelson,  1976) 
have  continued  to  treat  Elassoma  as  a  subfamily  of  the  Cen- 
trarchidae, presumably  accepting  the  conclusion  of  Eaton  (1953, 
1956)  that  Elassoma  is  a  neotenous  centrarchid,  with  most  of 
its  distinctive  features  having  arisen  through  paedomorphosis. 
Weitzman  and  Fink  (1983)  attributed  similar  reductions  in  the 
laterosensory  system  of  small  characids  to  paedomorphosis  and 
suggested  that  these  characters  may  be  quite  labile.  These  and 
other  osteological  reductions  similar  to  those  of  Elassoma  are 
found  in  other  small  fishes  such  as  gobioids  (Springer,  1983) 
and  cyprinodontoids  (Parenti,  1981),  but  I  know  of  no  such 
extreme  examples  among  small  percoids. 

That  the  reductive  specializations  of  Elassoma  actually  rep- 
resent character  states  of  earlier  developmental  stages  of  cen- 
trarchids has  never  been  clearly  demonstrated  or  even  ade- 
quately investigated,  and  comparative  studies  of  the  osteological 
development  of  these  fishes  would  be  necessary  to  answer  this 
question.  However,  a  crucial  point,  that  seems  to  have  been 
overlooked,  is  the  absence  of  any  other  evidence  suggesting  a 
close  relationship  between  Elassoma  and  the  Centrarchidae. 
Although  I  know  of  no  morphological  specialization  that  defines 
the  family,  all  centrarchids  exhibit  a  similar  mode  of  nest-build- 


466 


ONTOGENY  AND  SYSTEM ATICS  OF  FISHES -AHLSTROM  SYMPOSIUM 


Fig.  254.  (A)  Symphysanodon  sp.,  5.1  mm  SL;  (B)  Acropomatidae— Sv/!agropi  sp.,  8.5  mm  SL;  (C)  \cTopomn\iidie—Acropomajaponicuin, 
6.0  mm  SL,  from  Y.  Konishi  (unpubl.);  (D)  \cropomai\Aae  —  Doederleinia  herycoides.  8.0  mm  SL,  from  Okiyama  (1982b);  (E)  Polyprion  ameri- 
canus.  12.2  mm  TL,  from  Sparta  (1939a);  (F)  Slereolepis  doederleim.  1.1  mm  SL,  from  Okiyama  (1982b);  (G)  X^ohoxKAat  —  Lobotessunnamensis. 
6.0  mm  TL,  from  Uchida  et  al.  (1958);  (H)  Hapahgenys  sp.,  7.3  mm  SL,  from  Okiyama  (1982b). 


ing  and  parental-care  behavior,  and  this  behavioral  "synapo-  the  search  for  its  origins  to  the  Centrarchidae.  Quite  the  con- 

morphy"  is  not  shared  by  Elassoma  (Breder  and  Rosen,  1966;  trary,  I  believe  the  affinities  of  Elassoma  will  be  shown  to  lie 

M.  F.  Mettee,  pers.  comm.).  Consequently,  though  Elassoma  outside  the  Percoidei  and,  perhaps,  outside  the  Perciformes. 
may  be  a  product  of  paedomorphosis,  I  see  no  reason  to  limit  My  preliminary  findings  indicate  that  Elassoma  possesses  a 


i 


JOHNSON:  PERCOIDEI 


467 


Fig.  255.  (A)  Ambassidae—  I'elamhassis  jacksonensis.  5.5  mm  SL;  (B)  Opislognathidae—Opislognarhus  sp..  6  mm  SL;  (C)  Pseudochromidae, 
8.1  mm  SL.  from  Leis  and  Renins  (1983);  (D)  Acanlhochnidne—Acanlhoclirwslrilmealus.  10.0  mm,  from  Crossland  (1982);  (E)  Callanthiidae  — 
Grammatonotus  laysanus.  13.7  mm  SL,  from  Leis  and  Rennis  (1983);  (F)  CaWanXhwdae—Callanlhiaspelontanus.  8  mm  TL,  from  Fage  (1918); 
(G)  Con%xogzA\(i?Le—Congrogadus  suhducens.  1 1.8  mm  SL;  (H)  Monodactylidae— A/ono(/acO'/i«  sebae.  5.2  mm  SL,  from  Akatsu  et  al.  (1977); 
(I)  Pempherididae  — Pfm/jAmi  sp.,  5.5  mm  SL,  from  Leis  and  Rennis  (1983);  (J)  Op\e%mi\\\\dae-Oplcf>nathus  fasciatus.  7.5  mm  SL. 


number  of  salient  features  (not  mentioned  above)  that  cast  doubt 
on  its  affinities  with  the  Percoidei.  The  second  preurai  centrum 
bears  a  full  neural  spine,  and  there  are  no  autogenous  haemal 
spines.  Strong  parapophyses  begin  on  the  first  centrum,  and 
pleural  ribs  may  begin  on  the  first,  second  or  third  vertebra. 
The  first  neural  arch  is  fused  to  its  respective  centrum.  The 
pelvic  fin  is  inserted  well  behind  the  pectoral  fin  base  and  the 
pelvic  girdle  docs  not  contact  the  cleithra.  The  first  pharyngo- 
branchial  and  interarcual  cartilage  are  absent  and  what  is  ap- 
parently the  uncinate  process  of  the  first  epibranchial  articulates 


directly  with  the  second  pharyngobranchial.  The  fourth  phar- 
yngobranchial,  usually  cartilaginous  in  percoids,  is  absent.  The 
proximal  base  of  the  medial  half  of  the  uppermost  pectoral  ray 
does  not  extend  laterally  to  form  a  process  for  articulation  with 
the  scapular  condyle  (also  true  of  at  least  some  cyprinodontoids 
and  gobioids).  Finally,  the  ossified  portion  of  the  ethmoid  con- 
sists of  two,  closely  applied,  disc-like  bones,  a  condition  listed 
as  one  of  the  defining  characteristics  of  the  Atherinomorpha  by 
Rosen  (1964)  and  Rosen  and  Parenti  (1981).  (They  did  not 
discuss  the  distribution  of  this  character  among  other  groups. 


468 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  256.  (A)  Luljanidae  —  Luljanus  campechanus.  7.3  mm  SL,  from  Collins  et  al.  (1980);  (B)  Caesionidae— Carao  sp.  or  Gyinnocaesio  sp., 
7.8  mm  SL,  from  Leis  and  Rennis  (1983);  (C)  Leiognathidae— unidentified,  4.8  mm  SL;  (D)  Menidae— A/cne  maculala.  4.6  mm  SL;  (E)  Mala- 
canlh'idat—Caulolalilus  princeps.  6.0  mm  SL,  from  Moser  (1981);  (F)  Ma\acanlh\dae—Hoplolatilusfronlicinclus  (head  only),  15  mm  SL,  from 
Dooley  (1978);  (G)  Ephippididae—Chaclodiplerusfaber,  9  mm,  from  Hildebrand  and  Cable  (1938);  (H)  Pomacanthidae—Centropyge  sp.,  4.4 
mm  SL,  from  Leis  and  Rennis  (1983). 


JOHNSON:  PERCOIDEI 


469 


but  I  have  observed  a  similar  condition  in  the  gobiid  Dormi- 
tator.) 

Elassoma  seems  to  exiiibit  a  confusing  mosaic  of  character 
states  variously  shared  with  atherinomorphs,  cyprinodontoids, 
percopsiforms,  perciforms,  and  gobioids.  Resolution  of  the  evo- 
lutionary affinities  of  this  genus  could  be  important  to  our  un- 
derstanding of  acanthomorph  interrelationships,  and  I  intend 
to  examine  this  problem  more  fully. 

Epigonidae.  —  Eraser  ( 1 972a)  treated  Epigonns.  Florencwlla  and 
Rosenhlattia  as  a  subfamily  (Epigoninae)  of  the  Apogonidae, 
but  I  find  no  evidence  to  suggest  that  these  genera  are  closely 
related  to  other  apogonids.  They  are  primitive  with  respect  to 
apogonids  in  possessing  two  pairs  of  uroneurals  and  a  procurrent 
spur  (Johnson,  1975),  but  specialized  in  several  features  listed 
below.  Moreover,  the  two  anal  spines  of  epigonines  and  apo- 
gonids, usually  cited  as  evidence  of  their  close  relationship,  are 
not  homologous  (see  discussion  on  median  fins).  The  Epigonidae 
are  here  recognized  as  a  distinct  family,  including  Brinkmanella. 
Sphyraenops  and  Eraser's  epigonines.  These  five  genera  share 
the  following  specializations:  rostral  cartilage  greatly  enlarged, 
ascending  processes  of  premaxillaries  reduced  or  absent;  pre- 
maxillary  articular  cartilages  enlarged;  endopterygoids  large, 
metapterygoids  notably  reduced;  infraorbitals  more  than  six. 
The  larvae  of  Sphyraenops  (Pig.  257A)  resemble  those  of  Epigo- 
nus  (Fig.  257B)  but  differ  in  possessing  well-developed  head 
spination. 

Girellidae,  Kyphosidae,  Microcanthidae,  Neoscorpis.  Parascor- 
pididae  and  Scorpididae.  — Springer  (1982;  pers.  comm.),  fol- 
lowing Jordan  (1923)  and  Golvan  (1965),  included  microcan- 
thids,  Neoscorpis.  Parascorpis  and  scorpidids  in  the  family 
Scorpididae,  but  no  convincing  evidence  for  uniting  them  has 
been  presented,  and  they  are  treated  separately  here.  The  Scor- 
pididae is  here  restricted  to  Scorpis.  Medialuna.  Lahracoglossa 
and  Bathystethus.  The  latter  two  genera  were  treated  as  a  sep- 
arate family,  Labracoglossidae,  by  Springer.  Scorpidids  share 
similar  meristic  and  osteological  features  (not  derived)  and  com- 
parable scale  morphology.  An  unusual  small  slip  of  muscle  ex- 
tends from  the  basioccipital  to  the  first  vertebra  in  Scorpis  and 
Lahracoglossa.  but  its  presence  has  not  been  confirmed  in  the 
other  two  genera.  The  larvae  of  Scorpis  and  Bathystethus  are 
undescribed  hut  those  of  Lahracoglossa  (Fig.  258A)  and  Me- 
dialuna (Fig.  258B)  share  a  similar  body  form,  generalized  head 
spination,  late  fin  development  and  pigment  pattern  with  larvae 
of  the  Girellidae  (Fig.  258C).  Girellids  are  specialized  in  several 
osteological  features  with  respect  to  the  Scorpididae  (see  Table 
1 20)  and  have  a  unique  adductor  mandibulae  in  which  A,  inserts 
on  the  lateral  surface  of  the  dentary  (Johnson  and  Fritzsche, 
in  prep.).  The  distinctive  larval  form  shared  by  scorpidids  and 
girellids  suggests  that  they  may  be  sister  groups.  Convincing 
evidence  supporting  a  close  relationship  between  the  Scorpi- 
didae and  the  Microcanthidae  (Microcanthus.  Atypichthys  and 
Neatypus)  or  the  Kyphosidae  (Kyphosus,  Seclator  and  Her- 
nwsilla)  is  lacking.  Furthermore,  the  larvae  of  the  latter  two 
families  (Figs.  259G,  J)  do  not  possess  the  salient  features  of 
scorpidid  or  girellid  larvae,  but  more  closely  resemble  those  of 
the  Teraponidae  (Fig.  259H).  The  larvae  of  Neoscorpis  and  Par- 
ascorpis are  unknown,  and  available  anatomical  information  is 
insufficient  to  clarify  the  systematic  position  of  these  two  genera. 

Malacanthidae. — See  discussion  on  utility  of  larval  morphology. 


Moronidae  {Morone  and  Dicentrarchus),  Lateolahrax  and  5/>j- 
/perca.— Gosline  (1966)  included  the  Moronidae  (using  the  name 
Roccus).  Lateolahrax  and  Siniperca  (=Coreoperca)  in  his  "es- 
tuarine  and  freshwater  percichthyids."  I  treat  these  separately, 
because  I  lack  evidence  of  their  affinities  with  the  Fercichthyi- 
dae,  with  one  another,  or  with  any  other  percoid  group.  It  is 
interesting  to  note  that  the  Moronidae  share  one  of  the  two 
synapomophies  of  the  Centropomidae  described  by  Greenwood 
(1976)— the  lateral  line  extends  almost  to  the  posterior  margin 
of  the  caudal  fin.  In  addition,  moronids  have  an  auxilliary  row 
of  lateral  line  scales  on  the  caudal  fin  above  and  below  the  main 
row,  as  does  the  centropomid  Lates.  Although  both  of  these 
conditions  occur  elsewhere  in  generalized  percoids  (e.g.,  Neo- 
scorpis. some  species  of  Lutjanus.  and  the  percid  subfamily 
Luciopercinae)  and  may  actually  be  primitive  for  the  Percoidei 
(Springer,  1983),  the  possibility  of  a  moronid-centropomid 
relationship  seems  plausible  and  should  probably  be  investi- 
gated further.  Unfortunately,  as  is  typical  of  most  fresh  or  brack- 
ish water  spawners,  the  larvae  of  these  groups  (Fig.  260)  exhibit 
relatively  direct  development  and  consequently  offer  little  phy- 
logenetic  information. 

Percichthyidae.  — The  Percichthyidae  of  Gosline  (1966)  repre- 
sents a  polyphyletic  assemblage  defined  on  the  basis  of  shared 
primitive  features.  I  am  unable  to  find  synapomorphies  that 
support  recognition  of  the  assemblage  as  a  monophyletic  group. 
I  restrict  the  Percichthyidae  to  the  following  genera,  which  occur 
only  in  freshwaters  of  Australia  and  South  America:  Percolates 
(brackish  water),  Plectroplites.  Macquaria.  Maccullochella.  Per- 
cichthys,  Percilia.  Bostockia.  Gadopsis.  Nannoperca.  Edelia.  and 
Nannatherina.  The  monophyly  of  the  family  is  supported  by  a 
series  of  nested  synapomorphies,  only  a  few  of  which  are  men- 
tioned here.  The  scales  of  most  of  these  genera  are  similar  and 
unlike  those  of  the  excluded  genera  in  having  the  posterior  field 
filled  with  simple,  only  slightly  amputated  (see  McCully,  1970), 
needle-like  ctenii  (those  of  Bostockia.  Gadopsis  and  Nannath- 
erina are  secondarily  cycloid).  The  three  most  generalized  gen- 
era. Percolates,  Plectroplites.  and  Macquaria  are  very  similar 
biochemically  [MacDonald  (1978)  synonymized  them  on  this 
basis],  and  the  latter  two  share  two  morphological  specializa- 
tions with  Macidlochella.  Percichthys.  Percilia.  Bostockia  and 
Gadopsis:  enlarged  sensory  pores  on  the  dentary  and  a  separate 
inner  division  of  adductor  mandibulae  section  A,.  The  three 
most  derived  genera,  Nannoperca.  Edelia  and  Nannatherina 
(heretofore  treated  as  kuhliids)  share  with  Bostockia  a  similar 
vertebral  number  (29-33),  a  distinctive  asymmetrical  nasal  ro- 
sette, and  a  number  of  reductive  specializations  (absences  of  the 
subocular  shelf,  procurrent  spur,  and  supracleithral  sensory  ca- 
nal, reduced  numbers  of  procurrent  caudal  rays,  dorsal  spines, 
branchiostegals  and  trisegmental  pterygiophores,  and  an  inter- 
rupted or  absent  lateral  line).  Systematic  placement  of  the  enig- 
matic Gadopsis  has  proved  problematic,  even  in  recent  years. 
It  has  generally  been  treated  as  a  monotypic  family  and  variously 
assigned  to  the  Percoidei  (Greenwood  et  al.,  1 966),  Ophidioidei 
(Gosline,  1 968),  Perciformes  with  proposed  affinities  to  the  Tra- 
chinoidei  and  Blennioidei  (Rosen  and  Patterson,  1969)  or  a 
separate  order  Gadopsiformes  (Scott,  1962).  The  percoid  affin- 
ities of  Gadopsis  are  manifest  in  the  anatomy  of  the  dorsal  gill 
arches,  caudal  skeleton  and  median  fin  supports.  Its  affinities 
with  the  Percichthyidae  are  indicated  by  a  number  of  features 
shared  with  some  percichthyid  genera,  including  the  configu- 
ration of  the  adductor  mandibulae  noted  above.  Gadopsis  shares 


470 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  257.  (A)  Epigonidae—Sphyraenops  bairdianus.  6.8  mm  SL;  (B)  EpxgomAae—Epigonus  sp.,  14.0  mm  SL;  (C)  Howella  sp.,  6.0  mm  SL; 
(D)  \po%.on\dne  —  Pseudamia  sp.  or  Pseudamiops  sp.,  8.7  mm  SL,  from  Leis  and  Rennis  (1983);  (E)  Apogonidae  — foa  brachvgrainma.  4.2  mm 
SL,  from  Miller  et  al.  (1979);  (F)  Apogonidae  — unidentified,  4.2  mm  SL,  from  Leis  and  Rennis  (1983);  (G)  Apogonidae— unidentified,  5.0  mm 
SL,  from  Leis  and  Rennis  (1983);  (H)  Sciaenidae— 5rW///er  tanceotalus.  6.2  mm  SL,  from  Powles  (1980). 


the  asymmetrical  nasal  ro^eXXe  oi Bostockia.  Nannoperca.  Edelia 
and  Nannalherina  and  all  reductive  specializations  of  those  gen- 
era noted  above,  except  the  reduced  lateral  line  and  branchio- 
stegal  number.  Specializations  shared  with  Bostockia  alone  in- 
clude a  tubular  anterior  nostril  placed  near  the  margin  of  the 
lip  and  absences  of  the  basisphenoid,  medial  tabular,  and  third 
epural.  Based  on  this  evidence,  Gadopsis  appears  to  be  most 
closely  related  to  Bostockia,  however  it  bears  a  strong  superficial 
resemblance  to  Macullochella  and  shares  the  premaxillary  fre- 
num  of  that  genus. 

Adult  Morphology 

The  scope  of  morphological  diversity  exhibited  within  the 
Percoidei  surpasses  that  of  all  other  perciform  suborders.  Al- 
though many  percoids  have  a  generalized  bass-like  or  perch- 
like physiognomy,  extremes  of  adult  body  form  range  from  deep 


bodied,  compressed  or  "slabsided"  fishes,  such  as  the  ephip- 
pidids,  chaetodontids  and  menids  to  extremely  elongate  forms 
like  the  cepolids  and  the  eel-like  congrogadids.  Add  to  this  the 
exceptional  variability  in  fin  conformation,  ornamentation  of 
head  bones,  squamation.  jaw  configuration,  and  internal  osteo- 
logical  features,  and  the  suborder  Percoidei  presents  an  im- 
pressive heterogeneous  array  of  forms.  Lack  of  progress  in  elu- 
cidating percoid  phylogeny  is  largely  attributable  to  this 
somewhat  overwhelming  diversity  and  the  ostensible  conver- 
gence (particularly  in  reductive  traits)  that  seems  to  have  char- 
acterized percoid  evolution.  To  date,  no  familial  phylogeny, 
cladistic  or  otherwise,  has  been  proposed  for  the  suborder.  The 
limits  and  monophyly  of  many  of  the  component  families  are 
not  clearly  defined  and  the  affinities  of  numerous  genera  remain 
unresolved.  Superficial  knowledge  of  basic  percoid  anatomy  and 
an  inadequate  understanding  of  character  distribution  and  vari- 


JOHNSON:  PERCOIDEI 


471 


Fig.  258.  (A)  Scorpididae— Labracoglossa  argenliventris,  9.9  ininSL;{S)ScoTpididat—Medialunacaliforniensis,  10.1  mm  SL;(C)Girellidae— 
Girella  nigricans.  10.9  mm  SL;  (D)  Leptobramidae  — Z,fp/o/)rama  mulleri.  7.2  mm  SL;  (E)  CheWodacXyWdm—Palunolepishrachydactylus.  8.3  mm 
SL;  (F)  C\Tr\\\\\dae—Amhlycirrhituspinos.  13.2  mm  SL;  (G)  PoTm\om\dae— Pomatomus  satlalrix.  7.3  mm  TL.  from  Pearson  (1941);  (H)  Nem- 
ipteridae— unidentified,  5.1  mm  SL,  from  Leis  and  Rennis  (1983);  (I)  Spandae—Acanthopagrus  cuvieri.  8  mm  SL,  from  Hussain  et  al.  (1981);  (J) 
Cenlracanlhidae— Plerosmaris  axillaris,  7.7  mm  SL,  from  Brownell  (1979). 


ability,  basic  to  cladistic  outgroup  comparison,  have  seemingly 
inhibited,  or  at  least  hindered,  meaningful  comparative  studies 
within  the  Percoidei. 

Because  the  group  is  so  large,  these  problems  will  necessarily 
continue  to  plague  studies  of  percoid  relationships.  Outgroup 
comparisons  based  on  a  single  family  are  speculative  without 
evidence  for  a  sister  group  relationship,  and  broader  surveys  of 
each  character  are  frequently  impractical  if  not  impossible.  One 


approach  that  can  gradually  alleviate  this  problem  is  the  cu- 
mulative tabulation  of  characters  and  character  states.  Com- 
parative tables  document  the  distribution  of  morphological  fea- 
tures throughout  the  suborder  and  the  variability  of  these  features 
within  families,  and  they  accordingly  offer  the  most  complete 
foundation  for  outgroup  comparison.  Furthermore,  they  pro- 
vide information  about  the  plasticity  of  various  complexes,  al- 
low identification  of  characters  most  frequently  subject  to  con- 


472 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


JOHNSON:  PERCOIDEI 


473 


.% 


^-'^V,-:-. 


Fig.  260.  (A)  Centrarchidae— .•)«iWop//fM  rupestris.  10.5  mm  TL,  from  Fish  (1932);  (B)  Simperca  (=Coreoperca)  kawamebari.  9.0  mm  TL, 
from  Imai  and  Nakahara  (1957);  (C)  Pemdae  —  Perca  flavescens.  14.2  mm  TL,  from  Mansueti  (1964);  (D)  Percichthyidae— A/aa-w//oc/!e//a 
macquanensis,  size  unknown,  from  Dakin  and  Kesteven  (1938);  (E)  Lateolabrax  japomcus.  13.7  mm  TL,  from  Mito  (1957b);  (F)  Moronidae  — 
Morone  amencana.  13.2  mm  TL,  from  Mansueti  (1964);  (G)  Centropomidae— Ce«rrapo«ii«  undecimatis.  6.3  mm  SL,  from  Lau  and  Shafland 
(1982). 


vergence  and  convincingly  document  the  uniqueness  of  derived 
features.  With  this  in  mind,  I  have  compiled  information  about 
selected  morphological  features  of  adults  (Table  1 20)  and  larvae 
(Table  121)  for  each  percoid  family  or  inceriae  sedis  genus.  This 
information  was  compiled  from  the  literature  (particularly  the 
meristic  data)  and  from  my  own  examination  of  cleared  and 
stained  specimens  and  radiographs.  Data  for  a  few  groups  were 
compiled  by  experts  working  on  those  groups.  For  many  fam- 
ilies. I  examined  at  least  one  representative  of  each  genus,  but 
obviously  this  was  not  always  possible  and  only  in  a  few  of  the 
smaller  families  were  all  species  examined.  As  a  consequence, 
this  data  will  not  reflect  the  full  range  of  variability  for  every 
family  but  should  represent  a  reasonably  close  approximation. 
Most  features  considered  in  Table  1 20  are  discussed  below. 

Fins— The  primitive  perciform  complement  of  one  spine  and 
five  rays  (I,  5)  in  the  pelvic  fin  is  the  most  consistent  feature  of 


percoid  fins.  A  single  spine  is  always  present  and  fewer  than 
five  soft  rays  are  found  only  in  the  Acanthoclinidae  (I.  2),  Con- 
grogadidae  (I,  2-4  or  absent),  Plesiopidae  (I,  4),  Pseudochro- 
midae  (I,  3-5)  and  the  percichthyid  Gadopsis  (I,  1). 

The  primitive  and  most  common  number  of  principal  caudal 
fin  rays  (branched  rays  +  2)  is  9  +  8.  Where  reductions  occur 
(in  18  families)  they  usually  involve  one  fewer  principal  ray 
dorsally  and/or  ventrally  and  are  frequently  consistent  within 
families,  e.g..  8  +  7  in  Cheilodactylidae.  Chironemidae.  Cir- 
rhitidae.  Latrididae  and  Mullidae.  and  8  +  8  in  Acanthoclini- 
dae, Priacanlhidae,  and  Scatophagidae.  The  most  extreme  re- 
duction (4-6  branched  +  4-8  branched)  is  seen  in  the 
Congrogadidae.  The  only  apparent  increases,  10  +  9  found  in 
some  grammatids  and  plesiopids,  do  not  result  from  an  in- 
creased number  of  rays  articulating  with  the  hypurals,  but  from 
branching  of  the  outermost  hypural-associated  rays.  Numbers 
of  procurrent  or  secondary  caudal  rays  dorsally  and  ventrally 


Fig.  259.  (A)  Ge:m\(ia.e— Eucinostomus  sp.,  8.7  mm  SL;  (B)  HaemuWAae  —  Xenistius  californiensis.  6.5  mm  SL;  (C)  HaemuXiAae — Pseudo- 
prislipoma  nigra.  5.8  mm  SL,  from  Leis  and  Rennis  (1983);  (D)  HaemuVxdae-Conodon  nobdis.  9.8  mm  SL;  (E)  Mullidae.  8.2  mm  SL.  from 
Miller  et  al.  (1979);  (F)  Sillaginidae— .S'///tig<)  .«/;ama.  9.0  mm  TL,  from  llchida  et  al.  (1958);  (G)  M\!:TocaM\\'\dae  —  Microcanlhussthgatus.  7.1 
mm  TL.  from  Uchida  et  al.  (1958);  (H)  Tcraponidae— F/icrapo/i  iheraps.  9.5  mm.  from  Zvjagina  (1965b);  (I)  Emmelichthyidae— £o'''"'octe 
schlegeh.  6.9  mm  TL.  from  Nakahara  (1962);  (J)  Kyp\\o%\dae- Kyphosus  anerascens.  9.8  mm  TL.  from  Uchida  et  al.  (1958). 


474 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  261.     {A)Coryphaemd!ie—Coryphaenahippurus.  8.5  mm  SL;  (B)  Rachycenlridae—Rachycenlroncanadum.  9.0  mm  SL;  (C)Echeneididae- 
Echeneis  sp.,  8.8  mm  SL;  (D)  Caristiidae— Canif/i«  sp.,  10.1  mm  SL;  (E)  Bramidae— firaAjja  dussuimeri,  6.5  mm  SL,  from  Mead  (1972). 


range  from  0  in  the  Congrogadidae  to  19  in  the  Sillaginidae,  the 
most  common  numbers  being  8-14. 

One  of  the  most  variable  aspects  of  percoid  physiognomy  is 
the  form  and  composition  of  the  dorsal  fin.  Even  the  most 
consistent  feature,  the  presence  of  spines,  does  not  characterize 
all  percoids.  Absence  of  dorsal  spines  in  six  percoid  families 
appears  to  have  originated  by  at  least  two  different  mechanisms. 
In  Bathyclupea,  it  is  obvious  that  the  spines  have  been  lost 
because  the  spinous  pterygiophores  are  still  present  and  the  soft 
rays  occupy  a  position  posterior  to  them.  In  Coryphaena,  how- 


ever, Potthoff  (1980)  showed  that  although  the  anteriormost  3- 
4  pterygiophores  bear  soft  rays,  they  are  of  the  type  that  normally 
support  spines.  This  suggests  that  the  absence  of  spines  in  Cory- 
phaena is  the  result  of  transformation,  rather  than  loss,  of  pre- 
existing elements.  Absence  of  spines  in  the  Bramidae,  Caristi- 
idae,  some  cepolids  and  some  congrogadids  is  also  probably  the 
result  of  transformation. 

Spines  are  present  anteriorly  in  the  dorsal  fin  of  all  other 
percoids,  ranging  from  1  in  some  malacanthids  and  pseudo- 
chromids  to  XXI  in  some  acanthoclinids.  Dorsal  soft  rays  range 


Fig.  262.  (A)  Chaetodontidae— unidentified,  10  mm,  from  Burgess  (1978);  (B)  C\ae\oAon\\Aae—Forcipigerlongiroslris.  17  mm  TL,  from 
Kendall  and  Goldsborough  (191 1);  (C)  Chaetodontidae— C/jWotom  sp.  or  Coradion  sp.,  6.5  mm  SL,  from  Leis  and  Rennis  (1983);  (D)  Scato- 
phagidae— 5ca/o/7AagMi  argiis,  10  mm  SL  from  Weber  and  de  Beaufort  (1936);  (E)  Scombropidae  — 5fom/)TOpi  hoops.  6.2  mm  SL,  from  Uchida 
et  al.  (1958);  (F)  Lethrinidae— Z.e?/!n>ii/i  nematacanlhus,  6.1  mm  SL,  from  K.  Mori  (unpubl.);  (G)  CepoMiae—Acanthocepola  sp.,  9.7  mm  SL, 
from  Okiyama  (1982b);  (H)  Priacanlhidae  — unidentified.  4.6  mm  SL.  from  Leis  and  Rennis  (1983);  (I)  Priacanthidae— /'nacanr/iMisp.,  10.9  mm 
SL,  from  Leis  and  Rennis  (1983);  (J)  PenXace^toUd&c  —  Pseudopentaceros  richardsoni.  15  mm  SL. 


JOHNSON:  PERCOIDEI 


475 


476 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  120.  Selected  Morphological  Features  of  Adult  Percoidel  Abbreviations  and  definitions:  SS— supernumerary  (non-serial)  spines 
(or  soft  rays)  on  first  anal  pterygiophore  (see  Johnson,  1980);  D— dorsal  fin;  A— anal  fin;  Tnseg.  pteryg.  —  pterygiophores  with  proximal,  medial 
and  distal  radials  separate;  Stay  — separate  bony  element  posterior  to  ultimate  pterygiophore  in  D  and  A;  Predorsal  formulae  — based  on  Ahlstrom 
et  al,  (1976);  P— pterygiophore  with  no  supernumerary  spines  or  soft  rays;  H  — hypurals;  E— epurals;  U  — uroneurals;  Ah— autogenous  haemal 
spines;  pH  — parhypural;  UR  — urostyle;  Proc  spur— procurrent  spur  (see  Johnson,  1975);  PU3  cart- radial  cartilage  anterior  to  neural  and  haemal 
spines  of  third  preural  centrum;  BR  — branchiostegals;  lAC— interarcual  cartilage;  Cy— cycloid;  Ct— ctenoid,  ctenii  free  from  posterior  margin; 
Ct— ctenoid,  ctenii  continuous  with  posterior  margin;  and  br— branched  caudal  fin  rays.  With  the  exception  of  (SS),  parentheses  enclose  features 

known  to  characterize  only  some  members  of  a  group. 


Vertebrae 

Dorsal  fin 

Tbiseg. 

PTERYG- 

D 

—  Slav 
A 

Pelvic 
fin 

Predorsal  formulae 

CAUDAL   RN 

Pnncipal 

Anal  fin  (SSl 

Procurrent 

Acanthoclinidae 

10  + 

11  + 

18 

17 

XVIII-XXI,  3-4 
VIII-XI,4(l-2) 

h 

1,2 

0/0/1/1  +  1/ 

8  +  8 

3-4  +  3-4 

11  + 

18 

12  + 

18 

13  + 

16 

Acropomatidae 

10  + 

15 

VII-X-I,  8-10 
or 

'■a* 

1,5 

/0+0/0  +  2/1  +  I/ 
0/0/0  +  2/1/ 

9  +  8 

9-13  +  9-13 

IX-X,  10 

0/0/0  +  2/1  +  1/ 

II-III,  7-9  (2) 

Ambassidae 

10  + 

14 

Vll-I,  8-11 
111,7-10(2) 

0 
0 

1,5 

0/0/0+1/1/ 
0/0/1  +  1/1/ 

9  +  8 

7-11  +  7-10 

Aplodactylidae 

15  + 

16  + 

20 
19 

XIV-XXI,  16-21 
III,  6-9  (2) 

I* 

1,5 

0/0  +  0/2+1/1/ 
0/0  +  0/2/1  +  1/ 

8-9  +  7 

12-14+12-13 

Apogonidae 

10  + 
10  + 

14 
15 

VI-VIII-I,  8-14 
11,8-18(1) 

m* 

1,5 

various: 

0/0/0  +  2  or  1/ 

to 
///2  or  1/ 

9  +  8 

6-10  +  5-10 

Arripididae 

10  + 

15 

IX,  13-19 

\* 

I,  5 

0/0/0  +  2/1  +  1/ 

9  +  8 

111,9-10(2) 

7  +  6 

Banjosidae 

11  + 

14 

X,  12 

9 

1,5 

0/0  +  2/1/1/ 

9  +  8 

III,  7 

6  +  5 

Bathyclupeidae 

9  +  22 
10  +  21 

9 

i- 

1,5 

0/0/0//P+ P/P/P+ P/P/P/P/ 1  / 

9  +  8 

I,  26-27  (0) 

9  +  8 

Bramidae 

Braminae 

T;36- 

-47 

30-38 

I* 

1,5 

0/0/0/P/P/l  +  l/ 
0/0/0/P/l/ 

9  +  8 

21-30(2-3) 

7-8  +  7-8 

Pteraclinae 

T;45- 

-54 

46-57 
39-50  (?) 

'-a- 

1,5 

l  +  l  +  l  +  l,etc./l/l/ 

9  +  8 

5  +  5 

Caesionidae 

10  + 

14 

IX-XV,  9-21 
111,9-13(2) 

w 

1.5 

0/0/0  +  2/1  +  1/ 
/0  +  0/2/I  +  1/ 

9  +  8 

7-10  +  5-10 

Caesioscorpis 

10  + 

15 

XI,  20-21 
III,  18-20(2) 

6-8 
6-8 

1,5 

0/0  +  0/2/1 

9  +  8 

9-11+8-10 

Callanthiidae 

10  + 

14 

X-XI,  8-12 

'v 

1,5 

0/0/2/1  +  1/ 

8-9  +  7-8 

111,9-12(2) 

5-9  +  5-9 

Carangidae 

10  + 
10  + 

14 
15 

IV-VIII-I,  17-44 
I-II-I,  15-39(1-2) 

0-3 
0-2 

1,5 

see  section  on 
Carangidae 

9  +  8 

8-14  +  8-12 

10  + 

16 

10  + 

17 

11  + 

13 

11  + 

14 

Caristiidae 

T:  35 

-40 

32-40 
18-21  (1) 

O9 
0  ■ 

1,5 

l  +  1  +  I  +  l  +  l  +  I/l/I/ 
l  +  I  +  I  +  I/I/I/ 

9  +  8 

7-8  +  7-8 

Centracanthidae 

10  + 

14 

XI-XIII,  9-11 
111,9(2) 

'-'*  + 
1-4 

1,5 

0/0  +  0/2+1/1/ 
0/0/0  +  2/1/ 

9  +  8 

9-10  +  8-10 

Centrarchidae 

11- 

+ 

4 

V-XIII,  9-16 

3-6 
3-6^ 

1,5 

3-7  predorsals 
1-3  sup.  spines 

9  +  8 

II-VIII,  8-19  (2-3) 

5-10+5-9 

15- 

9 

T:28 

-32 

Centrogenysidae 

11  + 

14 

XIIl-XIV,  9-11 
III,  5  (2) 

0 
0  + 

I,  5 

0/0  +  2/1/1/ 

7  +  7 
5  +  5 

JOHNSON:  PERCOIDEI 


477 


Table  120.    Extended. 


SKELETON 

H/E/U/Ah 
H  Fusions 

Proc 
spur 
"PU3 

can. 

BR 

lAC 

Scales 

3/3/0/0 

+ 

6 

+ 

Ct 

pH-l-2;3-4-UR 

or 

Cy 

5/3/2/2 

+   + 

+ 

7 

+ 

Ct 
or 
Cy 

3/2/1/1 
1-2:  3-4 

+ 

6 

+ 

Cy 

5/3/1/2 

? 

6 

7 

Cy 

2-5/2-3/0-1/0-2 
various 

+ 

7 

+ 

Ct 
or 
Cy 

5/3/2/2 

+  - 
+ 

7 

+ 

Ct 

5/3/2/2 

+   + 

7 

+ 

Ct 

5/3/2/2 

+  + 

+ 

7 

+ 

Cy 

5/3/2/2 

+  - 
+ 

7 

+ 

Ci' 

5/3/2/2 

+  - 
+ 

7-8 

+ 

Cf 

3  or  5/3/2/2 
(1-2;  3-4) 

7 

7 

+ 

Ct 

5/3/2/2 

+  + 
? 

,      7 

+ 

Ct 

3/2-3/1/1 
1-2;  3-4 

+ 

6 

+ 

Ct 

2-3/2-3/1-2/2 
1-2;  3-4-(5) 



7-9 

+ 

Cy 

5/3/?/2 

3/3/2/2 

1-2;  3-4 

5/3/1-2/2 


7 

+ 

Cy 

6 

+ 

Ct 

6-7 

+ 

Ct 

from  as  few  as  3  in  some  acanthoclinids  to  89  in  some  cepolids. 
Within  families,  the  range  of  dorsal  fin  ray  counts  may  be  rel- 
atively restncted  as  in  the  Lutjanidae  (X-XII,  10-17)  or  quite 
broad,  as  in  the  Sciaenidae  (VII-XV  — 1,  17-46). 

In  most  percoids  the  spinous  and  soft  portions  of  the  dorsal 
fin  are  continuous,  but  gradual  shortening  of  the  posteriormost 
spines  results  in  a  variously  developed  cleft  or  apparent  sepa- 
ration. Where  this  cleft  is  present,  the  ultimate  spine  is  notably 
longer  than  the  penultimate  and  is  considered  to  form  the  first 
element  of  the  soft  dorsal  portion  of  the  fin.  Some  groups  (e.g., 
Pseudochromidae,  Grammatidae,  Plesiopidae,  Priacanthidae, 
Cepolidae)  do  not  develop  this  cleft.  Others,  such  as  the  Ac- 
ropomatidae,  Ambassidae,  Apogonidae,  Emmelichthyidae,  En- 
oplosidae  and  Epigonidae  have  such  a  well-developed  cleft  that 
the  spinous  and  soft  portions  of  the  fin  appear  completely  sep- 
arate. Pterygiophores  usually  continue  beneath  the  resultant  gap 
and  may  even  bear  minute  spines.  The  Mullidae  and  Echenei- 
didae  are  exceptional  in  having  no  pterygiophores  below  this 
gap.  Extreme  separation  of  the  spinous  and  soft  dorsal  fins  oc- 
curs only  in  the  Echeneididae,  where  the  spinous  dorsal  fin  has 
been  modified  as  an  attachment  disc  and  has  moved  far  forward 
to  cover  the  flattened  cranium. 

The  anal  fin  of  percoids  is  less  variable  in  form  and  com- 
position than  the  dorsal  fin.  The  most  common,  and  apparently 
primitive  condition  is  three  anal  spines.  The  first  anal  pteryg- 
iophore  is  larger  than  the  succeeding  pterygiophore  and  bears 
the  first  two  spines  in  supernumerary  (non-serial)  association 
and  the  third  in  serial  association  (see  Johnson,  1980).  Scato- 
phagids  and  some  chaetodontids  and  pomacanthids  have  four 
spines,  the  first  two  supernumerary.  Centrarchids  have  from 
two  to  eight  spines,  pentacerotids  from  two  to  six  and  gerreids 
from  three  to  five,  all  with  the  first  two  supernumerary.  The 
only  other  percoids  characterized  by  more  than  three  anal  spines 
(eight  to  eleven)  are  the  Acanthoclinidae,  where  one  or  two  may 
be  supernumerary.  Several  groups  have  fewer  than  three  anal 
spines,  and,  as  in  the  dorsal  fin,  it  is  important  to  understand 
the  nature  of  this  reduction.  Apogonids,  for  example,  have  only 
two  spines  and  only  one  of  these  is  supernumerary,  suggesting 
that  the  anteriormost  spine  was  lost.  The  mesopelagic  epigon- 
ines  (sensii  Eraser,  1 972a)  have  usually  been  treated  as  a  subfam- 
ily of  the  Apogonidae,  for  they  also  have  only  two  anal  spines. 
The  epigonine  anal  spines,  however,  are  both  supernumerary 
(as  are  those  of  the  Sciaenidae),  suggesting  that  the  usually  spi- 
nous third  (serial)  element  has  not  transformed  into  a  spine. 
Hence,  the  two  anal  spine  conditions  of  epigonines  and  apo- 
gonids are  not  homologous.  In  bathyclupeids,  the  single  anal 
spine  is  serially  associated  with  the  first  pterygiophore,  sug- 
gesting that  the  first  two  spines  have  been  lost.  Only  a  few  groups, 
Bramidae.  Caristiidae,  Congrogadidae,  Coryphaenidae  and  some 
cepolids  and  grammistine  serranids,  lack  anal  spines.  The  pres- 
ence of  1-3  supernumerary  elements  on  the  first  pterygiophore 
in  all  these  groups  indicates  transformation  rather  than  loss  of 
the  pre-existing  spines.  Anal  soft  rays  range  in  number  from  4 
in  the  Acanthoclinidae  to  101  in  the  Cepolidae  and,  with  some 
exception,  the  range  of  variability  within  families  is  comparable 
(frequently  within  two  or  three  rays)  to  thai  of  the  dorsal  soft 
rays. 


4/3/1/2 
1-2 


Ct 


Predorsal bones.  —  In  most  percoids,  one  to  three  strut-like  bones 
precede  the  anteriormost  pterygiophores  of  the  dorsal  fin.  It  has 
been  proposed  (Smith  and  Bailey,  1961),  but  never  conclusively 
demonstrated,  that  these  predorsal  bones  were  derived  from 
true  pterygiophores  that  once  bore  spines  or  rays,  but  Eraser 


478 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 

Table  120.    Continued. 


Dorsal  fin 


Anal  fin  (SS) 


Triseg. 

PTERYG. 

D 

-Slay 
A 


Predorsal  formulae 

CAUDAL  RN 

Pnncipal 

fin 

Procurrent 

1,5 

0/0/0  +  2/1  +  1/ 
0/0/0+1/1  +  1/ 

9  +  8 

8-12  +  7-11 

Cenlropomidae 

10  +  14 

11  +  14 

Cepolidae 

Cepolinae 

1 2  +  44-66 

14  +  55 

16  +  53 

Owstoniinae 

11  +  17 

13  +  16 

14  +  16 

Chaetodontidae 

10  +  14 

11  +  13 

Cheilodactylidae 

14  +  21 

Chironemidae 

13  +  20 

14  +  19 

Cirrhitidae 

10  +  16 

Congrogadidae 

12-19 

+ 

34-64 

Coracinidae 
Coryphaenidae 

Datnioides 

Dinolestidae 

Dmoperca 

Drepanidae 

Echeneididae 

Emmelichthyidae 

Enoplosidae 
Ephippididae 

Epigonidae 
Gerreidae 


10  +  15 

13-15 
+ 

17-19 
T:  30-34 
10  +  14 


10  +  15 

11  +  14 

10  +  14 


VII-VIII.  8-13 
III,  6-9  (2) 


O-lII,  65-89 
0-1,62-101 


III-IV,  21-27 
I-II,  14-19 

Vl-XVl.  15-30 

III-V.  14-23(2) 

XIV-XXII.  15-39 

111,6-19(2) 

XlV-XVl,  15-21 

III,  6-8  (2) 

X, 11-17 


III,  5-7(2) 


O-II.  33-76 
28-63 


X.  18-23 

III,  13-14(2) 

52-66 

23-31  (2) 


XII,  15 


III,  15-28(2) 


VII-VlII-1.  7-10 
I-III,  7-9(1-2) 


IX-X.  9-17 
III-V,  13-17(2) 


0-5 
0-4 


111.9(2) 

1 

10  +  17 

Vliii-I,  18-19 
1,26-27(1) 

0-1 
0-1 

10  +  16 

XI,  17-19 
III,  11-13(2) 

4 
3 

10  +  14 

XIII-IX,  19-22 
111,  17-19(2) 

0 
0 

12-18 

+ 

14-22 

T:26-40 

IX-XXVIII- 17-42 
II,  14-36(1) 

0 
0 

10  +  14 

XI-XIV,  9-12 
III,  9-1 1  (2) 

6-8 
6-8 

10  +  16 

VIII-I,  14-15 
III,  14-15(2) 

1-7 
4-8 

10  +  14 

V-IX,  18-40 

0(11*) 

0 

0^ 
0-(  +  ) 
0 

0  + 
0 

0  + 
0 
0  + 

0  _ 
0 


1* 

0  _ 
0 


5 

1 

J^ 
1 

4 
3 

0(  +  ) 


+ 


+ 


0(11*) 
*Ephippus 


0-1 


(  +  ) 


0-1 


0-2 
0-2 


I,  5 
1,5 

I,  5 

I,  5 
I,  5 
I,  5 


1,5 


mm 


HUM 


0/0  +  2/1/ 

0/2/1/1/ 

0  +  0/2+1  +  1/1/ 

0  +  0/2+1/1  +  1/ 

0/0/2+1/1/ 

0/0+0/2/1  +  1/ 
0/0/0  +  2/1  +  1/ 


I,  2-4       0/0/0/P+  1 

or  0/0//P+1 

absent       0/0//P  +  P 
///P+P/ 
1,5         0/0/0  +  2/1  +  1/ 


l-3+l  +  l  +  l,etc./l/I/ 


5 

0/0/0  +  2/1/ 

5 

0/0/0+1/1  +  1/ 
0/0/0/1  +  1/ 

5 

0/0/0  +  2+1/1/ 

5 

0/0  +  0/2/1  +  1/ 

5 

absent;  D, 
on  head 

5 

0/0/0  +  2/1/ 

5 

0/0/0  +  2/1  +  1/ 

5 

0/0  +  0/2/1  +  1/ 
0+0  +  0//2/1  +  1/ 
0  +  0/2/1  +  1/ 
0/0/0+1/1/ 
0  +  0/0/P/l 

5 

0/0/0  +  2/1  +  1/ 
/0+0/0  +  2/1  +  1/ 
/O  +  0/O+l/l/ 

5 

0/0/0  +  2/1  +  1/ 

6-7  +  6 
1-2+1-2 

8  +  7 
3-4  +  3-4 

9  +  8 
2-4  +  2-3 

8  +  7 
9-1 1 +8-10 

8  +  7 
13-16+10-12 

8  +  7 
9-14+10-13 

4-6  br  +  4-8  br 
0-4+0-3 


9  +  8 
9  +  8 
9  +  8 


10- 

-14+10- 

9  +  8 

6  +  7 

9  +  8 
11  +  11 

9  +  8 
13+12 

9  +  8 

5+4 

9  +  8 

14 

8-13  +  7-13 


9  +  8 
7-8  +  7-8 


9  +  8 

5-6  +  5-6 

9  +  8 

3-7  +  3-6 


9  +  8 
9-10  +  7-10 

9  +  8 
9-11+9-10 


JOHNSON:  PERCOIDEI 


479 


Table  120.    Continued.    Extended. 


(  AUnAl 
SMIEION 

H  E  11  Ah 
H  Fusions 

Proc 
spur 

Wi 

cart. 

5/2-3/1-2/2 

+   + 

— 

+ 

2-4/1-2/1/2? 

__ 

(1-2;  3-4;  5  absent) 

7 

3-4/3/1/2 

_  _ 

(1-2;  3-4) 

- 

5/3/2/2 

5/2-3/1/2 



5/3/1/2 

_  _ 

— 

? 

5/3/2/2 



— 

+ 

2/0-2/0/0-1 



pH-1-2;  3-4-UR 


5/3/2/2 


3/1/1/2 
1-2;  3-4 


+  + 


Ct 


6 

+ 

Cy 

6 

+ 

Cy 

6 

+ 

Ct 

3-6 

- 

Cy 

6 

- 

Cy 

6 

r 
or 

Cy 

Cy 

Ct 

Cy 


5/3/2/2 

+  + 

+ 

6 

+ 

Ct 

- 

5/3/2/2 

+  + 
? 

7 

+ 

Cy 

5/3/2/2 

+  + 

7 

+ 

Ct 

5/2-3/2/2 

+  + 

+ 

6 

+ 

Cy 

5/2/1/1-2/ 

-    +(-) 

^   8-11 

- 

Cy 

5/3/2/2 

+  + 

7 

r 

Ct 

— 

? 

or 

5/3/2/2 

+  + 
? 

7 

+ 

Cy 

4-5/3/2/2 

+  +(-  -)* 

6 

— 

Ct 

(2-3) 

(  +  ) 
*Rhinoprenes 

or 
Cy 

5/3/2/2 

+  + 

7 

+ 

Ct 

+ 

or 

3  or  5/3/2/2 
(1-2;  3-4) 

+  + 
+ 

6 

+ 

Ct 

( 1 972a)  argued  that  the  first  three  predorsal  elements  of  percoids 
may  represent  supraneurals.  Ahlstrom  et  al.  (1976)  recognized 
the  importance  and  utility  of  considering  patterns  of  predorsal 
bones  in  early  life  history  studies,  and  further  developmental 
studies  could  resolve  the  origin  of  these  elements. 

The  most  common  and  presumably  primitive  number  of  pre- 
dorsal bones  in  percoids  is  three;  Table  120  shows  that  over 
half  of  91  percoid  groups  (families  and  incertae  sedis  genera) 
have  three  predorsal  bones  exclusively,  with  three  predorsals 
occurring  in  at  least  some  members  of  66  groups.  The  first  dorsal 
pterygiophore  inserts  in  the  third  intemeural  space  in  at  least 
some  members  of  69  groups,  bears  two  supernumerary  spines 
in  some  members  of  69  groups  and  exhibits  both  conditions  in 
57  groups.  Therefore,  the  most  common  and  ostensibly  prim- 
itive predorsal  formulae  (using  that  defined  by  Ahlstrom  et  al., 
1976)  for  the  Percoidei  are  0/0/0  +  2/  and  0/0  +  0/2/.  The  0/0/1 
pattern,  considered  by  Smith  and  Bailey  (1961)  to  be  primitive 
for  percoids  occurs  in  only  six  families,  frequently  in  the  more 
derived  members.  Furthermore,  Fraser  (1972a)  noted  that  der- 
ivation of  the  0/0/0  +  2/  or  0/0  +  0/2/  patterns  from  the  0/0/1 
pattern  by  backward  shift  of  the  first  dorsal  spine,  hypothesized 
by  Smith  and  Bailey  (1961),  is  untenable  and  inconsistent  with 
pterygiophore  interdigitation.  On  the  other  hand,  the  0/0/0  +  2/ 
pattern  could  be  easily  derived  by  a  posterior  shift  of  the  first 
dorsal  spine  in  the  0/0/1  +  1/  pattern  that  characterizes  many 
beryciforms,  including  holocentrids  and  diretmids.  This  latter 
pattern  is  found  among  percoids  only  in  some  ambassids. 

Departures  from  the  primitive  predorsal  pattern  have  appar- 
ently arisen  independently  in  many  families.  In  anterior  shifts 
of  the  dorsal  fin  origin  a  compound  first  pterygiophore  with  two 
supernumerary  spines  is  frequently  retained,  but  it  is  invariably 
absent  in  posterior  shifts.  A  possible  conclusion  is  that  forward 
shifts  result  from  anterior  displacement  of  the  pterygiophores, 
whereas  posterior  shifts  result  only  from  loss  of  spines.  Reduc- 
tions in  numbers  of  predorsal  bones  to  fewer  than  three  are 
almost  certainly  the  result  of  simple  losses  as  opposed  to  trans- 
formations, even  when  these  reductions  are  accompanied,  as 
they  are  occasionally  (e.g.,  Chaetodontidae,  Scatophagidae,  Pen- 
tacerotidae,  Priacanthidae),  by  a  forward  shift  of  the  dorsal  fin 
origin.  Transformations  of  pre-existing  predorsal  bones  to  spi- 
nous pterygiophores  would  require  the  unlikely  addition  of  de 
novo  spines  and  distal  radials,  and  modification  of  musculature. 
More  than  three  "predorsal  bones"  are  found  in  at  least  some 
members  of  13  percoid  groups,  with  a  corresponding  posterior 
shift  of  the  dorsal  fin  origin.  The  additional  elements  are  usually 
distinguishable  from  the  anterior  three  ("true")  predorsals.  In 
Bathyclupea,  Braminae,  some  carangids,  Congrogadidae,  Mene, 
Neoscorpis.  Platax.  some  pempheridids  and  Toxotidae,  these 
additional  spineless  elements  (designated  P  in  Table  120)  re- 
semble pterygiophores,  may  have  separate  distal  elements,  and 
often  articulate  with  succeeding  similar  elements  or  with  the 
anterior-most  spine-bearing  pterygiophore.  In  Bathyclupea  and 
Toxotidae.  they  are  also  separated  from  the  true  predorsals  by 
one  intemeural  space.  In  the  remaining  groups  with  more  than 
three  "predorsals"  (some  percichthyids  and  centrarchids,  Brink- 
manella  and  Leptohrama).  the  additional  elements  are  not  mor- 
phologically distinguishable  from  the  anterior  three,  but,  as  in 
the  other  groups,  the  dorsal  fin  originates  posterior  to  the  third 
intemeural  space  (except  in  Brinkinanella),  and  it  seems  likely 
that  these  elements  were  also  derived  from  pre-existing  pteryg- 
iophores. Studies  of  the  sequence  of  development  of  predorsal 


480 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  120.    Continued. 


Anal  lit!  (SS) 


Triseg. 

PTERYG, 

D 

-Slav 
A 


Pelvic 
fin 


Predorsal  formulae 


CAL'DAL  RN 

Pnncipal 

Procurrenl 

9  +  8 

10  +  9 

9  +  8 

9-14  +  8-12 

9  +  8 

7-8  +  7-8 

8-10  +  7-9 

2-8  +  2-7 

9  +  8 

9-14  +  8-13 

9  +  8 

6  +  5-6 

9  +  8 

11  +  10 

9  +  8 

9-10  +  9-10 

9  +  8 

10-12+10-12 

9  +  8 

11-13+10-12 

9  +  8 

Giganthiidae 

10  +  15 

Girellidae 

11  +  16 

11  +  17 

14  +  20 

Glaucosomatidae 

10  +  15 

Grammatidae 

10  +  15 

10  +  17 

10  +  18 

Haemulidae 

10+  16 

11  +  16 

Hapalogenys 

10  +  14 

Hemilutjanus 

10  +  15 

Howella 

10  +  16 

Inermiidae 

12  +  14 

13  +  13 

Kuhliidae 

10  +  15 

Kyphosidae 

10  +  15 

10  +  16 

Lactariidae 

10  +  14 

Laleolabrax 

17  +  18 

Latrididae 

14  +  21 

Leiognathidae 

10  +  14 

Leptobramidae 

10  +  14 

Lethrinidae 

10  +  14 

Lobotidae 

11  +  13 

12  +  12 

Lutjanidae 

10  +  14 

Malacanthidae 

10  +  14 

11  +  14 

11  +  16 

Menidae 

10  +  14 

Microcanlhidae 

10  +  15 

Monodactvlidae 

10  +  14 

10  +  15 

Moronidae 

11  +  14 

12  +  13 

IX.  13 

111,8(2) 

XII-XVl.  11-17 

III,  10-13(2) 


VIII,  1  1 


111,9(2) 

XII-XIV.  8-10 

III,  7-11  (2) 

IX-XIV,  1 1-26 
111,6-13(2) 

XI-XIV,  14-19 
111,9-13(2) 
X, 10-11 
III,  9  (2) 
VIII-I,  9 
111.7(2) 
X-IIorXVlI,  9- 


10 


111,8-10(2) 

X.  9-12 

111,9-13(2) 

X-XII.  11-15 

III,  10-16(2) 

VlI-VIII-1.  20-22 

III,  25-28  (2) 

XII-XIV,  12-16 

111,7-10(2) 

XIV-XXIII,  23-40 

III,  18-35(2) 

VIII,  15-16 

III,  14(2) 

IV,  16-18 


111.26-30(2) 

X.  9-1 1 
111,8-10(2) 
XII,  15-16 

III.  11  (2) 
X-XII,  10-17 

111,7-11  (2) 

I-X,  14-60 

I-II,  11-55(1-2) 

IV,  38-42 
III,  28-32  (2) 
X-XI.  16-22 
111,  13-19(2) 

VII-IX.  26-36 

III,  27-37  (2) 

VIll-X-1,  10-13 

111,9-12(2) 


6 
0-1 
0-1 

7 
0 


0 

0 

0 

6 

6 

0 

0 

0 

0 

4-7 

5-8 

1-7 

3-5 

3-5 

2-3 

6 

3 

0 

0 

15-16 

13 

0 

0  " 
2-3 
2-3  ' 
0 

0  " 
1-7 
1-7  " 
?-l7 
?-14  " 

0 

O" 
10-16 
10-15  ' 
0-2 
0-1  " 
2-4 
2-4  " 


1.5 

0/0/2/1  +  1/ 

1.  5 

0/0  +  0/2/1  +  1/ 

0/0/2/1  +  1/ 

I.  5 

0/0/0  +  2/1  +  1/ 

1,5 

0/0/0  +  2/1  +  1/ 

0/0/0+1/1  +  1/ 

I,  5 

0/0/0  +  2/1/ 

0/0  +  0/2+1/1/ 

0/0  +  0/2/1/ 

1,5 

0/0  +  0/2/1  +  1/ 

1,5 

0/0/0  +  2/1  +  1/ 

1,5 

0/0/0  +  2/1  +  1/ 

1,5 

/0/0  +  2/1/ 

/0  +  0/0  +  2/1/ 

1,5 

0/0/0  +  2/1  +  1/ 

0/0/0  +  2/1/ 

1,5 

0/0/0  +  2/1/ 

1,5 

0/0/0  +  2/1  +  1/ 

1,5 

0/0/0+1/1  +  1/ 

1,5 

0/0  +  2/1  +  1/1/ 

0  +  0/2/1  +  1/1/ 

1,5 

0/1/1/1/ 

I,  5 

0/0/0/0//P/P/P/P/P/ 1  / 1  +  1  + 1  / 

I,  5 

0/0  +  0/2+1/1/ 

1,5 

0/0/0  +  2/1  +  1/ 

I,  5 

0/0/0  +  2/1  +  1/ 

0/0  +  0/2/1  +  1/ 

1,5 

0/0/2/1  +  1/ 

/0+0/2/1  +  1/ 

//2+1  +  1  +  1  +  1/1  +  1  +  1/ 

1.5 

0/0  +  0/P/l/ 

1,5 

0/0  +  0/2/1  +  1/ 

1,5 

0/0/0+1/1  +  1/ 

1.5 

0/0/0  +  2/1  +  1/ 

0/0/0  +  2/1/ 

0/0/0/2+1/ 

8-10  +  8-10 
9  +  8 
9  +  8 
9  +  8 


13+12-13 

8  +  7 

14+12 

9  +  8 

9-10  +  7-9 

9  +  8 

6-8  +  7 

9  +  8 

7-9  +  7-9 

9  +  8 

3-5  +  3-5 

9  +  8 

8-13  +  8-13 

9  +  8 

10-13  +  9-13 

9  +  8 

4  +  3-4 

9  +  8 

7-10  +  7-10 

9  +  8 

6  +  5-6 

9  +  8 

10-13  +  9-13 


JOHNSON:  PERCOIDEI 


481 


Table  120.    Continued.    E,\tended. 


1  \i  r^Ai 

SKFt  I^TON 

H/E  U/Ah 
H  Fusions 

Proc 
spur 

Pili 

can. 

BR 

lAC 

Scales 

5/3/1/2 

1 

7 

+ 

Ct 

5/3/1/2 

-(r)- 

6 

+ 

Ct 

— 

+ 

5/3/2/2 

+  + 

7 

+ 

Cl 

2-3/3/0/2 

+ 

6 

+ 

Ct 

1-2;  3-4-UR 

5  absent 

5/3/2/2 

+  + 
? 

7 

+ 

Ct 

5/3/2/2 

? 

7 

+ 

Ct 

5/3/2/2 

+  + 

7 

7 

+ 

Ct 

5/3/2/2 

+ 

7 

+ 

Ct' 

3/3/2/2 
1-2;  3-4 

+  + 
? 

7 

+ 

Ct 

5/3/2/2 

+  + 

+ 

6 

+ 

Ct 

5/3/2/2 

+  + 

+ 

7 

+ 

Ct 

3/3/2/2 
1-2;  3-4 

+  + 
+ 

7 

+ 

Cy 

5/3/2/2 

+  + 

+ 

7 

+ 

Ct 

4/3/1/2 
3-4 

? 

6 

7 

Cy 

3/3/1/2 
1-2;  3-4 



5 

- 

Cy 

4/3/2/2 
3-1 

+  + 
? 

6 

+ 

Ct 

5/3/2/2 

? 

6 

+ 

Ct 

5/3/2/2 

+  - 

6 

+ 

Ct 

3  or  5/3/2/2 
(1-2;  3-4) 

7 

+ 

Ct 

5/3/2/1-2 

+ 

6 

+ 

Ct 

2/3/0/0/ 
1-2-3-4-UR 

7 

+ 

Cy 

5/3/2/2 

+  + 
(+) 

7 

+ 

Ct 

2-5/3/2/2 
(1-2;  2-3;  3-4) 

+  - 
+ 

7 

+ 

Ct 

5/3/2/2 

+  + 

+ 

7 

+ 

Ct 

bones  in  relation  to  the  development  of  the  dorsal  fin  may  prove 
useful  in  determining  the  homologies  of  these  additional  ele- 
ments as  well  as  the  first  three  predorsals. 

Caudal  skeleton.— The  primitive  percoid  caudal  skeleton  con- 
sists of  one  parhypural  with  a  well-developed  hypurapophysis, 
five  hypurals,  two  pairs  of  uroneurals.  three  epurals,  one  ural 
centrum,  a  low  neural  crest  on  PU2  and  autogenous  haemal 
spines  on  PU2  and  PU3.  This  configuration  is  found  in  at  least 
some  members  of  54  percoid  groups. 

The  most  common  reductions  involve  fusion  of  hypurals  one 
and  two  and  hypurals  three  and  four  and  loss  of  the  posterior 
uroneural  pair.  Loss  of  one  epural  occurs  in  only  14  groups,  and 
epurals  are  completely  lacking  only  in  some  congrogadids.  More 
extreme  reductions,  including  various  combinations  of  fusions 
of  the  hypurals  with  the  parhypural  and/or  urostyle,  loss  or 
fusion  of  the  anterior  uroneural  pair  and  fusion  of  the  autoge- 
nous haemal  spines,  occur  in  only  a  few  groups:  Acanthoclini- 
dae,  some  apogonids,  Congrogadidae,  Grammatidae,  Menidae, 
Mullidae,  Opstognathidae,  Plesiopidae,  and  Pseudochromidae. 

The  second  preural  centrum  bears  a  full  neural  spine  in  only 
two  groups,  Echneididae  and  Nandidae,  except  for  occasional 
anomalous  specimens.  This  full  neural  spine  must  be  second- 
arily derived  in  the  echeneidids  because  these  fishes  are  un- 
questionably closely  related  to  other  percoids  that  bear  the  usual 
reduced  neural  crest  on  PU2  (see  discussion  on  utility  of  larval 
morphology).  Unfortunately,  evidence  for  the  origin  of  this  de 
novo  spine  in  echeneidids  is  lacking.  Although  it  may  represent 
a  captured  first  epural  (there  are  only  two  in  echeneidids),  it  is 
attached  and  of  full  length  at  its  earliest  appearance  in  ontogeny. 
Another  possibility  is  that  the  second  preural  centrum  of  other 
percoids  has  been  lost  in  echeneidids,  so  that  the  last  centrum 
bearing  a  full  neural  spine  actually  corresponds  to  PU3.  How- 
ever, presence  of  the  usual  autogenous  haemal  spines  on  both 
PU2  and  PU3  in  echeneidids  refutes  this  hypothesis.  The  sig- 
nificance of  a  full  neural  spine  on  PU2  in  the  Nandidae  is  un- 
clear, since  the  affinities  of  this  family  with  the  Percoidei  remain 
problematic. 

The  presence  of  a  procurrent  spur  and  of  radial  cartilages 
anterior  to  the  neural  and  haemal  spines  of  PU3  are  probably 
primitive  features  (Johnson,  1975,  1983).  The  procurrent  spur 
is  developed  to  some  extent  in  50  percoid  groups,  all  but  ten  of 
which  have  a  primitive  caudal  complex.  Reductions  among  these 
ten  groups  usually  involve  only  simple  hypural  fusion.  The  pro- 
current  spur  is  never  present  in  groups  with  fewer  than  9  -F  8 
principal  rays.  Third  preural  radial  cartilages  are  found  in  45 
of  66  groups  examined  for  them,  about  half  of  which  have 
primitive  caudal  complexes. 

I'crtchral  number.  —  \ c'rXf:bra.\  number  ranges  from  23  to  about 
78  in  percoids.  Gosline  (1968,  1971)  noted  that  the  "basal 
number"  of  vertebrae  in  percoids  is  24-25  (10  -I-  14-15),  and 
this  number  characterizes  45  of  the  91  groups  treated  in  Table 
1 20;  63  groups  have  24-27  vertebrae.  Twenty-two  groups  have 
vetebral  counts  greater  than  thirty,  but  only  five  have  more  than 
40  vertebrae.  Only  priacanthids  and  scalophagids  have  fewer 
than  24  (10  +  13). 

G///a/-c/!e5.  — Primitively,  percoid  gill  arches  contain  the  follow- 
ing elements:  one  basihyal,  four  basibranchials  (the  fourth  car- 
tilaginous), three  pairs  of  hypobranchials,  five  pairs  of  cerato- 
branchials,    four    pairs    of  epibranchials,    four    pairs    of 


482 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  120.    Continued. 


Anal  fin 

(SS) 

VII-VIII 

-1,8-9 

I-II.  5- 

8(0-1) 

IX,  2 

6-29 

III,  16- 

X,  9 

III,  7- 

-17(2) 

-10 

-8(2) 

VI-VIII 

,  20-22 

III.  23- 
X-XII, 

-26  (2) 
11-22 

Triseg, 

PTERYG, 
D 

—  Sla\ 
A 


Pelvic 
fin 

Predorsal  formulae 

1,5 

0/0/0-1-2/1  +  1/ 

0/0/0-1-1/14-1/ 

0/0/2/ l-t-l/ 

1,5 

0-I-0-I-0//2/1/ 

1,5 

0/0/2-1-1/1/ 

1,5 

0/0/0 -i-P/P/P-l-1/ 

1,5 

/OH-O/l/l-l-l/ 

/O/l/H-l/ 

//l/l-l-l/ 

1,5 

0/0-1-0/2/1-1-1/ 

0/0/2/1  +  1/ 

1,5 

0/0/0  +  2/1  +  1/ 

1,5 

0/0  +  0/2+1/1/ 

1,5 

0/0/0+1/1  +  1/ 

0/0/0+1/1/ 

0/0/0+P/l  +  l 

1,5 

0/0  +  2/1/ 

0/0/2/1  +  1/ 

CA1.1DAL  FIN 

Pnncipal 


Mullidae 

10  +  14 

Nematistiidae 

10  +  14 

Nemipteridae 

10  +  14 

Neoscorpis 

10  +  15 

Opistognathidae 

10  +  15-21 

11  +  16-23 

12  +  18-19 

13  +  18-20 

Oplegnathidae 

10  +  15 

Ostracoberycidae 

10  +  15 

10  +  16 

Parascorpididae 

12  +  15 

Pempherididae 

10  +  15 

Pentacerotidae 

12  +  12 

12+  13 

13+11 

13+  12 

13  +  13 

13  +  14 

Percichthyidae 

10-15 

+ 

15-23 

T;25-36 

Gadopsis 

21  +  26 

Percidae 


Plesiopidae 


T:31-50 


10  +  15 


II-III,  10-20(1-2) 


XI- 

XII, 

11-22 

111 

11- 

18(2) 

IX,  8 

-10 

III,  7- 

8(2) 

XI- 

XII, 

14-17 

111 

13- 

15(2) 

IV 

-VII 

7-12 

III 

17- 

45(2) 

IV- 

-XV 

,8-29 

II-VI,  6-17(?-2) 


VII-XI.  8-18 
111,7-13(2) 


X-XII,  25-28 

III,  17-19(2) 

V-IXX-0-lII,  7-24 

1-11,4-15(1-2) 


IX-XV,  7-21 


+ 

16-25 

T:26-35 


I-III,  13-30 


i- 


^* 


II 


4-7 


13 

—  + 
11 


0-16 
0-15 


^  +  (-) 


6-16 


10  +  16 

III,  8-23  (2) 

6-19 

12  +  25 

Pomacanthidae 

10  +  14 

IX-XV,  15-33 

0 

III-IV.  14-25(2) 

0 

Pomatomidae 

11  +  15 

VIl-IX,  23-28 

2-3 

III,  22-28  (2) 

3 

Polyprion 

13  +  14 

XI-XII,  11-13 

8-10 

111,8-10(2) 

5-6 

Priacanthidae 

10+  13 

X,  11-15 

0-1 

111,9-16(2) 

0-1 

Pseudochromidae 

10-13 

1-111,21-37 

0-many 

0-many 


I,  5 

variable: 

0-5  predorsals 

0-2  sup.  spines 

I,  1 

//O/O/l/l/ 

1,5 

0/1/1/1/ 

/l/l/l/ 

////I/l/ 

/////l/l/ 

1,4 

0/0/0  +  2/1  +  1/ 

0/0/2/1  +  1/ 

0/0/0+1/1  +  1/ 

0/0/1/1  +  1/ 

1.5 

0/0/2/1  +  1/ 

0//2/1  +  1/ 

1,5 

0/0/0+1/1  +  1/ 

I,  5 

0/0/0+2/1  +  1/ 

1,5 

0  +  2/1/1/1/ 

2/1/1/1/ 

.3-5 

0/0/0  +  2/1  +  1/ 

0/0/0+1/1  +  1/ 

0/0/2/1  +  1/ 

0/0/1/1  +  1/ 

8  +  7 


8-10+8-10 


9  +  8 


9-10  +  8-9 
9  +  8 


8-11+8-11 

9  +  8 
9  +  8 

6-7  +  6-7 

3-8  +  3-7 

9  +  8 

9  +  8 

9  +  8 

10-12+10-11 

9+8 
11+9 

9  +  8 

3-7  +  3-7 


9  +  8 

3-7  +  3-6 


9  +  8 


5- 

16  +  5- 

9  +  8 
5  +  5 

-14 

8 

-9  +  7- 

-8 

10- 

15  +  8- 

-13 

9- 

10  +  8- 

-9 

3-10  +  3-9 


9  +  8 


4  +  3-4 

9  +  8 

9-10  +  8-9 

9  +  8 

8-9  +  7-8 

8  +  8 

4-6  +  4-6 

7-9  +  7-8 

5-7  +  5-7 


JOHNSON:  PERCOIDEI 


483 


Table  120.    Continued.    Extended. 


(  M   DM 
SKI  1  t  ION 
H    h    I  1    Ah 

H  Fusions 

Proc 

cart 

BR 

lAC 

Scales 

3/2/1-2/2 
1-2;  3-4-UR 

? 

4 

- 

Ct 

3/3/2/2 
1-2;  3-4 



7 

+ 

Cy 

5/3/2/2 

+ 

6 

r 
or 

Ct 

5/3/2/2 

+  + 

6 

+ 

Ct 

2-3/3/0/1 

+ 

6 

+ 

Cy 

pH-1-2;  3-4-UR; 

(5  absent) 

5/3/2/2 

+  + 
? 

7 

+ 

Ct 

5/3/2/2 

+  + 

+ 

7 

+ 

Ct' 

5/3/2/2 

+  + 

7 

+ 

Ct 

3  or  5/3/1-2/0-2 

-(  +  )    + 

7 

+ 

Ct 

(1-2;  3-4) 

+ 

or 
Cy 

5/3/1/2 

+    -(  +  ) 

7 

+ 

Ct 

_ 

+ 

pharyngobranchials,  and  an  interarcual  cartilage  between  the 
uncinate  process  of  epibranchial  1  and  pharyngobranchial  2. 
The  first  pharyngobranchial  is  rod-like  and  serves  to  suspend 
the  dorsal  gill  arches  from  the  neurocranium.  The  fourth  phar- 
yngobranchial is  reduced  and  cartilaginous,  but  consistently  bears 
a  well-developed  dermal  tooth  plate,  as  do  the  second  and  third 
pharyngobranchials  and  the  fifth  ceratobranchials.  Small  tooth- 
plates  on  the  second  and  third  epibranchials  are  variously  pres- 
ent or  absent. 

Reductive  departures  from  the  primitive  branchial  complex 
are  few  and  involve  only  the  basihyal,  first  pharyngobranchial 
or  interarcual  cartilage.  The  basihyal  is  reduced  or  absent  in 
ephippidids.  Pseudochromids  lack  a  first  pharyngobranchial 
(Springer  et  al.,  1977).  Of  88  percoid  groups  examined  for  it, 
only  1 3  lack  a  well-developed  interarcual  cartilage  and  at  least 
three  of  these  (Cirrhitidae,  Emmelichthyidaeand  Nemipteridae) 
may  have  a  vestigial  element.  The  remaining  eleven  groups 
completely  lack  the  interarcual  cartilage,  but  most  have  an  un- 
cinate process  with  the  cartilaginous  tip  separated  by  a  decided 
gap  from  the  second  pharyngobranchial  and  frequently  pointing 
away  from  it.  This  condition  differs  from  the  primitive  state  (as 
represented  in  most  beryciforms)  wherein  the  uncinate  process 
of  the  first  epibranchial  directly  contacts  that  of  the  second 
pharyngobranchial,  and  suggests  that  these  percoids  have  sec- 
ondarily lost  the  interarcual  cartilage.  A  condition  resembling 
that  of  the  beryciforms  was  observed  among  percoids  only  in 
some  anthiin  serranids,  where  it  must  be  secondary.  In  eche- 
neidids  the  uncinate  process  of  the  first  epibranchial  also  artic- 
ulates directly  with  that  of  the  second,  but  there  is  a  concom- 
itant extreme  reduction  of  the  main  arm  of  the  first  epibranchial 
not  seen  in  beryciforms.  Again  this  condition  must  be  derived 
if  the  relationships  of  the  echeneidids  are  as  postulated  here  (see 
discussion  on  utility  of  larval  morphology). 


7/2-3/1-2/2 

+ 

+  (-  -) 

5-7 

+ 

Ct 

■> 

+  (-) 

or 
Cy 

5/2/1/2 



7 

+ 

Cy 

5/3/1/2 



5-8 

+ 

Ct 

3/3/0-2/1 
l-2-(pH);  3-4-UR 


5/3/2/2 


5/3/2/2 


5/3/2/2 


5/3/2/2 


3/2-3/0/0-1/ 
(pH)- 1-2;  3-4-UR 


+  (-) 


— 

6 

+ 

Ct 

+ 

or 



6 

+ 

Ct' 

-1- 

+  + 

7 

+ 

Cy 

-1-  + 

7 

+ 

Ct 

7 



6 

- 

Ct' 

_  _ 

6 

_ 

Ct 

Scales.— The  unpublished  work  of  McCully  (1961)  on  compar- 
ative anatomy  of  serranid  scales  provides  an  excellent  illustra- 
tion of  the  wealth  of  information  available  in  the  scales  of  per- 
coid fishes  that  has  largely  been  ignored  in  systematic  studies. 
More  recent  work  on  ctenoid  scales  of  other  groups  (DeLamater 
and  Courtenay,  1973a,  b,  1974;  Hughes,  1981)  using  scanning 
electron  microscopy  also  demonstrates  the  systematic  value  of 
ctenoid  scales.  Details  of  the  scale  morphology  of  most  percoids 
are  unknown.  On  a  gross  level,  three  basic  scale  types  (Ct,  Ct' 
and  Cy  in  Table  120)  are  found  among  percoids.  Although  be- 
ryciforms and  some  myctophids  are  said  to  have  ctenoid  scales, 
these  scales  (Ct')  differ  from  the  type  possessed  by  most  percoids 
and  other  perciforms  (Ct).  In  beryciforms  and  myctophids  the 
"ctenii"  are  continuous  spinous  projections  from  the  lateral  sur- 
face and  posterior  margin  of  the  scale.  A  few  percoids  (Bramidae, 
Epigonidae,  Howella.  Pomacanthidae,  Priacanthidae,  Ostraco- 
berycidae  and  Scatophagidae)  possess  similar  scales  that  may 
represent  retention  of  the  plesiomorphic  beryciform  condition, 
or  may  have  been  secondarily  derived.  In  the  "true"  ctenoid 
scale  that  characterizes  most  percoids  (59  groups),  the  ctenii  are 
separate  bony  plates,  or  scalelets  (McCully,  1961,  1970),  that 
are  continually  added  in  the  posterior  field  as  the  scale  grows. 
In  most  groups  the  posterior  field  becomes  filled  with  remnants 
of  old  ctenii,  the  tips  of  which  are  amputated  (or,  more  likely, 
resorbed),  as  each  new  row  of  ctenii  is  added.  In  a  few  groups, 
however  (e.g.,  anthiine  serranids  and  callanthiids),  only  a  pri- 
mary and  secondary  row  of  marginal  ctenii  are  evident.  This 
second  variation  of  "true"  ctenoid  scale  also  characterizes 


484 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  120.    Continued. 


Triseg. 

rALIDAL  I-TN 

Vertebrae 

Dorsal  tin 

PTERYG. 
D 

—  Slav 

A 

Pelvic 
fin 

Predorsal  formulae 

Pnncipal 

Anal  fin  (SS) 

Procurrent 

Rachycentridae 

11   + 

14 

VIl-IX,  26-34 
1-11,22-28(1) 

0 
0 

I,  5 

/l  +  l/l/l/ 

9  +  8 

15-16+12-14 

Scatophagidae 

10  + 

13 

Xl-Xll,  16-18 
IV,  14-16(2) 

0 
0  + 

I,  5 

0/0  +  2/1/1/ 
0  +  0/2/1/1/ 

8  +  8 

4-6  +  4-5 

Sciaenidae 

10- 

+ 

5 

VII-XV-1,  17-46 
1-11,5-23(1-2) 

l^ 

1,5 

0/0/0  +  2/1  +  1/ 

//2+l  +  l  +  l,etc./ 

9  +  8 

7-10  +  6-9 

12- 

S 

T:24- 

-29 

Scombropidae 

10  + 

16 

VIlI-lX-1,  12-13 

h 

1,5 

0/0/0  +  2/1  +  1/ 

9  +  8 

II,  11-12(1) 

11  +  10 

Scorpididae 

10  + 
10  + 

15 
16 

IX-X,  22-28 
III.  17-28(2) 

m^ 

1,  5 

0/0/0  +  2/1  +  1/ 

9  +  8 

11-13+10-12 

Serranidae 

10  + 

11  + 

14 
13 

VI-XIl,  9-24 
or 

0-24 
0-19  "^ 

I,  5 

0/0/0  +  2/1  +  1/ 
0/0  +  0/2/1  +  1/ 

8-9  +  7-8 

3-12  +  3-10 

10  + 

15 

II-IV,  20-29 

0/0/2/1  +  1/ 

10  + 

16 

II-III,  6-22  (2) 

0  +  0/2/1/1/ 

11  + 

15 

or 

0/0/1/1  +  1/ 

10  + 

18 

13-17(1) 

0/0/P/l  +  l/ 
0//P/1  +  1/ 

//l/l  +  l/ 

Sillaginidae 

14-20 

+ 

X-XIII-I,  16-27 

11,  14-26 

^ 

I,  5 

0/0/0/1  +  1/ 
0/0/0/1/ 

9  +  8 

17-19+14-19 

19-27 

0//0/1/1  +  1/ 

T:33- 

-44 

Siniperca 

12  + 

13  + 

16 
15 

XI-XV,  10-17 
111,7-13(2) 

7-10 
4-6 

1,5 

0/0/0  +  2/1  +  1/ 
0/0/0/2/1  +  1/ 

9  +  8 

6-12  +  6-12 

13  + 

18 

Sparidae 

10  + 

14 

X-XIII,  10-15 
111,7-14(2) 

1-4 

1,5 

0/0+0/2+1/1/ 

9  +  8 

7-11+7-11 

Stereolepis 

12  + 

14 

XI-XII,9-10 
III,  7-9  (2) 

W 

1.5 

0/0+0/2/1  +  1/ 
0/0/0  +  2/1  +  1/ 

9  +  8 

10-11+8-9 

Symphysanodon 

10  + 

15 

IX,  10 

w 

1,5 

0/0/0  +  2+1/1/ 

9  +  8 

III,  7-8  (2) 

12-14+12-14 

Teraponidae 

10  + 
10  + 

15 
16 

XI-XIV,  8-14 

111,7-12(2) 

0 
0  + 

1,5 

0/0+0/2/1  +  1/ 
0/0+0/1/1  +  1/ 

9  +  8 

9-10  +  5-8 

10  + 

17 

0+0/0  +  2/1/1/ 

11  + 

14 

11  + 

16 

Toxotidae 

10  + 

14 

IV-Vl,  11-14 
III,  15-18(2) 

0(+) 
0  - 

I,  5 

0/0/0//P/P/1/ 
0/0/0//P/1/ 

9  +  8 

4-5  +  4-5 

0//0//P/P/1/ 

Aphredoderus.  gobies  and  some  flatfishes,  and  the  mechanism 
of  growth  of  the  posterior  field  is  not  understood.  As  shown  by 
the  authors  mentioned  above,  there  is  extensive  diversity  in 
configuration  and  processes  of  formation  of  marginal  and  sub- 
marginal  ctenii,  and  this  diversity  undoubtedly  holds  useful 
phylogenetic  information. 

The  third  major  scale  type  found  among  percoids  is  the  cy- 
cloid scale  (Cy  in  Table  1 20),  characteristic  of  most  groups  below 
the  Percomorpha.  Although  the  cycloid  scales  of  some  percoids 
may  represent  a  plesiomorphic  state,  they  are  clearly  secondary 
in  a  number  of  families  where  they  occur  only  in  some  members 
(acanthoclinids,  acropomatids,  apogomds,  ephippidids,  pem- 
pheridids,  percichthyids,  sciaenids  and  serranids).  Cycloid  scales 
also  characterize  all  members  of  two  groups  of  percoid  families. 


each  of  which  probably  comprises  a  monophyletic  lineage.  The 
cirrhitoid  fishes  (Aplodactylidae,  Cheilodactylidae,  Chironem- 
idae,  Cirrhilidae,  and  Latrididae)  have  large  cycloid  scales  of 
similar  morphology,  and  the  carangoid  fishes  (Carangidae,  Cor- 
yphaenidae,  Nematistiidae,  Rachycentridae  and  Echeneididae) 
have  very  small  adherent  cycloid  scales.  Cycloid  scales  char- 
acterize five  other  families  of  moderate  size,  Ambassidae,  Ce- 
polidae.  Congrogadidae,  Leiognathidae  and  Opistognathidae. 
Otherwise,  cycloid  scales  are  restricted  to  a  few  monotypic  fam- 
ilies and  tncertae  sedis  genera  (Bathyclupeidae,  Caristiidae,  Di- 
nolestidae,  Drepamdae,  Enoplosidae,  Lactanidae,  Menidae,  Po- 
matomidae,  Scombropidae  and  Siniperca).  The  widespread 
occurrence  of  true  ctenoid  scales  in  the  Percoidei,  including  most 
of  the  less  specialized  forms,  and  the  distribution  of  cycloid 


JOHNSON:  PERCOIDEI 


485 


Table  120.    Continued.    Extended. 


t  AliDAl 

Proc 

SKELETt)N 

H/E/U/Ah 
H  Fusions 

spur 

cart. 

BR 

lAC 

Scales 

5/3/2/2 

h 

7 

+ 

Cy 

5/3/1/2 

6 

- 

CV 

5/3/2/2 

+    +(--) 

7 

+ 

Ct 

— 

? 

or 

Cy 

5/3/2/2 

+   + 

+ 

7 

+ 

Cy 

3  or  5/3/2/2 
(1-2:3-4) 

+   + 
(+) 

7 

+ 

Ct 

3  or  5/3/1/2 
(1-2;  3-4) 

- 

6-7 

+ 
(-) 

Ct 
or 
Cy 

5/2-3/1-2/2 

+  + 

+ 

6 

+ 

Ct 

4-5/3/1/2 

r(-)   + 

7 

+ 

cv 

(3-4) 

- 

3  or  5/3/2/2 
(1-2:  3-4) 

+ 

6 

+ 

Ct 

5/3/2/2 

+  + 

7 

+ 

Ct 

5/3/2/2 

-  + 
+ 

7 

+ 

Ct 

5/3/2/2 

+  + 

+ 

6 

+ 

Ct 

5/3/0-1/2 


+  - 


Ct 


fully  coalesced  by  hatching.  Most  members  of  the  three  primary 
freshwater  families.  Centrarchidae,  Percichthyidae  and  Percidae 
have  demersal  eggs  as  do  some  members  of  the  Ambassidae 
and  Teraponidae,  however  only  six  families  of  exclusively  ma- 
rine percoids  are  known  to  possess  non-buoyant  eggs.  The  Acan- 
thoclinidae.  Congrogadidae,  Plesiopidae  and  Pseudochromidae 
have  specialized  demersal  eggs  with  adhesive  threads  that  bind 
them  together  in  attached,  sheet-like  (Plesiopidae)  or  free,  spher- 
ical (Acanthoclinidae  and  Pseudochromidae)  masses  that  are 
guarded  by  the  male.  These  eggs  also  have  numerous  small  oil 
globules  that  gradually  coalesce  with  a  single,  much  larger  glob- 
ule. The  possibility  that  these  four  families  are  closely  related 
has  remained  unresolved  (Bohlke,  1960a;  Springer  et  al.,  1977), 
but  the  similar  egg  morphology  and  parental  care  shared  by 
them  may  represent  synapomorphies  not  heretofore  considered. 
The  other  two  marine  families  with  adhesive  demersal  eggs, 
Apogonidae  and  Opistognathidae  are  oral  brooders,  and  oral 
brooding  has  also  been  reported  for  the  plesiopid  Assessor  (Allen 
and  Kuiter,  1976). 

Larvae 

Diversity  of  general  body  form  and  morphological  special- 
ization among  the  larvae  of  percoid  fishes  is  extensive,  and.  as 
with  the  adults,  no  single  feature  shared  by  larval  percoids  char- 
acterizes the  suborder.  Representative  postflexion  larvae  of  62 
percoid  groups  are  illustrated  in  Figs.  254-262.  Larval  serranids 
and  carangids  were  excluded  from  these  figures  because  they  are 
illustrated  elsewhere  in  this  volume.  I  was  unable  to  obtain 
specimens  or  illustrations  of  larvae  of  the  remaining  30  groups 
and  most  are  probably  unknown,  or  at  least  undescribed.  Of 
these,  19  are  monotypic. 

Larval  body  form  ranges  from  elongate  to  deep-bodied,  by 
the  criteria  of  Leis  and  Rennis  (1983),  and  frequently,  but  not 
always,  reflects  adult  body  form.  Thus,  some  of  the  most  deep- 
bodied  percoid  larvae  are  found  among  the  Chaetodontidae, 
Pomacanthidae,  Menidae,  Bramidae,  and  Caristiidae,  whereas 
the  elongate  Congrogadidae  and  Cepolidae  have  elongate  larvae. 
On  the  other  hand,  the  moderately  elongate  larvae  of  groups 
like  the  Girellidae  or  the  Cirrhitidae  are  not  particularly  reflec- 
tive of  the  adult  body  form,  nor  are  the  deeper-bodied  larvae 
of  the  Emmelichthyidae. 

In  Table  121,  selected  aspects  of  known  larvae  of  percoid 
families  and  mcertae  sedis  genera  are  given.  This  table  should 
prove  a  useful  guide  to  identification  of  postflexion  larval  per- 
coids at  the  family  level,  particularly  when  used  in  conjunction 
with  the  meristic  data  in  Table  1 20  and  the  illustrations  in  Figs. 
254-262.  Features  included  in  Table  121  are  discussed  below. 


scales  just  described,  suggests  that  cycloid  scales  in  most  per- 
coids have  been  secondanly  acquired. 

Development 

Eggs 

Most  percoids  have  buoyant,  spherical  eggs  about  1mm  in 
diameter,  with  a  single  oil  globule.  The  total  size  range  is  about 
.5  to  4.6  mm,  but  eggs  larger  than  2  mm  are  found  only  in  a 
few  freshwater-associated  groups,  Centrarchidae,  Moronidae. 
Percichthyidae,  Percidae,  Siniperca  and  Teraponidae,  and  in 
the  marine  Echeneididae  (Table  121).  Multiple  oil  globules  oc- 
cur in  some  centrarchids,  percichthyids  and  sciaenids,  and  in 
Hapalogenys.  moronids  and  Polyprion,  but  they  are  generally 


Fin  development.  —  Formation  of  median  fin  rays  occurs  at  very 
small  sizes  in  most  percoids.  Flexion  may  begin  as  early  as  2.5- 
3  mm  and  is  complete  in  most  groups  by  4-5  mm,  at  which 
time  the  full  complement  of  principal  caudal  rays  is  present. 
Dorsal  and  anal  fin  rays  begin  to  form  during  or  shortly  after 
flexion  and  are  usually  complete,  including  spinous  rays,  by  5- 
8  mm.  Size  at  flexion  and  completion  of  full  median  fin  ray 
complements  is  relatively  consistent  within  families,  the  range 
usually  not  varying  more  than  2  mm.  Groups  characterized  by 
notably  later  flexion  (6-18  mm)  include  the  Caristiidae,  Cen- 
tracanthidae,  Centrarchidae,  Cheilodactylidae,  Girellidae,  La- 
tcolabrax,  Moronc.  Percichthyidae,  Percidae,  Polyprion.  Scor- 
pididae,  Sillaginidae,  and  Siniperca.  These  groups  also  exhibit 
somewhat  delayed  dorsal  and  anal  fin  ray  completion  (7-18 
mm).  Among  marine  percoids,  the  most  extreme  delay  in  com- 


486 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  121.  Selected  Early  Life  History  Features  of  Percoidel  Parentheses  enclose  features  known  to  characterize  only  some  members  of 
a  group.  Head  spination  abbreviations— Supraoccipital:  SI  —small  peak-like  crest;  S2  — SI  with  serrations;  S3  — large  vaulted  spine-like  crest  with 
serrations;  S4  — low  serrated  median  ridge;  S5— entire  surface  rugose.  Frontal:  Fl  —entire  surface  rugose;  F2— one  or  more  parallel  or  converging 
serrated  ridges;  F3  — serrated  supraorbital  ridges;  F4  — single  spine  on  supraorbital  ridge;  F5  — large  posteriorly  projecting  serrated  spine.  Preopercle: 
PI— posterior  margin  with  moderate  to  large  simple  spines;  P2  — PI  plus  lateral  ridge  with  one  or  more  small  simple  spines;  P3  — P2  with  spine 
at  angle  notably  elongate;  P4  — P3  with  marginal  spines  serrate;  P5  — posterior  margin  and  sometimes  lateral  ridge  with  very  small  spines  or 
serrations.  Other  bones  with  simple  spines,  serrations  or  serrated  ridges:  Op— opercle;  Sb  — subopercle;  lo— interopercle;  Ta— tabular;  Pt  — 
posttemporal;  Scl  — supracleithrum;  CI— cleithrum;  La  — lacrimal;  Co— circumorbitals;  Na— nasal;  Mx  — maxillary  shaft;  D— dentary;  Br— bran- 
chiostegals;  Pe  — pterotic;  Pa  — parietal;  Sp— sphenotic.  Sequence  of  completion  of  fin  rays:  A.  D,-A-D|-P,-P|;  B.  D,-P,-D,-A-P,;  C.  P,-P,-D,-A- 
D,;  D.  P|-D,-A-D|-P,;  E.  A-D,-P,-P|-Di;  F.  P,-D,-A-D|-P|.  Egg  type:  P— pelagic,  buoyant;  D— demersal;  A  — adhesive;  M  — egg  mass;  O— oral 

brooder. 


Size  (mir 

) 

Sequence 

Text 

D  &  A  rays 

First 

of  fin 

Head 

Other 

Taxon 

figures 

Egg  type 

Egg 

Halch 

Flex 

complete 

scales 

completion              spination 

specializations 

Acanthoclinidae 

255D 

D,  A,  M 

-1.4 

-4.7 

5-6 

7 

7 

A 

P5 

None 

Acropomatidae 

254A-D 

P 

? 

9 

-4 

-5 

12-15 

A 

(SI),(S4),(F2), 
F3,  (P4),  (P5). 
Op,  Sb,  lo,  Pt 
Scl,  (Pel), 
(Co),  (D),  (Pe) 

(D  and  P, 
spines  ser- 
rate) 

Ambassidae 

255A 

(D,  A)(P) 

.7-.  8 

1.8 

-3.5 

5.5-6 

9-10 

A 

P5 

None 

Apogonidae 

257E>-G 

D,  A,  M,  O 

<1 

2.5-3 

3-4 

4-6 

12  or  > 

A(B) 

(S1),(S5),(F1), 
(P2),  (P3), 
(P5),  (Op), 
(Sb),  (Pt) 
(.  .  .?) 

(Fl),  PI,Op, 

(Elongate  D 
spines  and 
P,  rays) 

Bramidae 

261E 

P 

9 

-3 

4-7 

6-10 

7-10 

C(D) 

Spinous  scales 

Sb,  lo 

(large  P,  and 

Callanthiidae 

255E-F 

P 

? 

7 

5 

7 

7-14 

A 

P2,  Op,  Sb,  lo, 
Pt 

(S1),(F3),(F4), 

None 

Carangidae 



P 

.7-1.3 

1-3.5 

-3-5 

-6-10 

-7-14 

A(D) 

(Elongate  D 

P3,  (P4),  Pt, 

spines  and 

Scl,  (Pe) 

P2  rays) 

Caristiidae 

261D 

P 

1.1-1.3 

2.3-2.9 

-7 

-8 

7 

A 

P5 

None 

Centracanthidae 

258J 

P 

1.1-1.3 

2.3-2.9 

6-7 

8-9 

7 

A 

P5 

None 

Centropomidae 

260G 

P 

.7 

1.4-1.5 

3.6-3.8 

-7 

-14 

A 

P5 

None 

Centrarchidae 

260A 

D,  A 

.8-2.8 

2.2-5.5 

6-9 

-7-13 

-14-18 

A 

None 

None 

Cepolidae 

262G 

P 

.7 

<3 

8-9 

7-9 

7 

A 

S3,  F1,F3,  P4, 
Scl,  D 

None 

Chaetodontidae 

262A-C 

P 

.7-.9 

1.5-2.0 

4-5 

5-8 

7-11 

A(B) 

All  exposed 
head  bones 
thick  and  ru- 
gose. 

Pt  and  Scl  ex- 
panded poste- 
riorly. 

P  expanded  to 
cover  cheek 
and  with 
broad  flat 
spine  poste- 
riorly. 

(P,  spine  long 
and  serrate) 

(Ant.  D  spines 
long  and  ru- 
gose) 

Cheilodactylidae 

258E 

P 

.9-1 

2.9-3.3 

7-8 

10-12 

-10 

A 

None 

Postlarvae 
deep,  com- 
pressed, sil- 
very to  70- 
90  mm 

Cirrhitidae 

258F 

P 

? 

7 

-4 

-8 

10  or  < 

A 

P5 

Chin  barbel 

Congrogadidac 

255G 

D,  A,  M 

7 

7 

7 

7 

7 

7 

None 

None 

Coryphaenidae 

26IA 

P 

-1.6 

-4 

6.5-7.5 

D  1 3-24 
A    8-11 

-25-30 

E 

F4,  P2,  Pt 

Minute  epithe- 
lial "prick- 
les" by 
-6  mm; 
"swollen" 
pterotics 

Echeneididae 

261C 

P 

1.4-2.6 

4.7-7.5 

5-9 

D  12-30 
A    6-12 

-  1 5-30 

E 

None 

Large  hook- 
like teeth  on 
dentary 

Emmelichthyidae 

2591 

P 

? 

7 

? 

7 

7 

A 

Pl,Op,  lo,  Pt, 

Scl 

None 

JOHNSON:  PERCOIDEI 

Table  121.    Continued. 


487 


.Size  (mm) 

Spfiiipnr*' 

Text 

D  &  A  rays 

First 

of  fin 

Head 

Other 

Taxon 

figures 

Egg  lype 

Egg 

Hatch 

Flex 

complete 

scales 

completion              spination 

specializations 

Ephippididae 

Chaetodipterus 

256G 

P 

~l 

-2.5 

-4 

-5 

-8-9 

A 

S1,F3.  P2.  Op, 
lo,  Ta,  Pt 

Spinous  scales 
to  - 1 5  mm 

Epigonidae 

Epigonus 

257B 

P 

? 

7 

7 

7 

7 

7 

None 

None 

Sphyraenops 

257A 

P 

7 

7 

7 

7 

-12 

7 

S1,S5,  F1,F3, 

None 

P3,  Op,  Pt,  Pe 

Gerreidae 

259A 

P 

.6-.  7  5 

-1.4 

3.5-4.4 

-6 

>15 

A 

P5,  (Scl) 

None 

Girellidae 

258C 

P 

~1 

-2.3 

-6 

11-13 

-15-16 

A 

P5,  Scl 

None 

Haemulidae 

259B-D 

P 

.8-1.0 

1.7-2.8 

3.9-5.4 

6-8 
(earlier 

in 
P.  nigra) 

-13 
(much 
earlier  m 
Conodon) 

A 

(F3),  (P1),(P5), 
(Op),  Sb,  lo, 
(Pt),  Scl,  Pe 
(also  F2,  Pel, 
La,  Co,  Na,  D 
in  Conodon) 

None  (spinous 
scales  in 
Conodon) 

Hapalogenys 

254H 

P 

1.2 

-3 

-4.5 

-5-6 

>10 

F 

S3.  S5,FI,F3, 
P2,  Op,  Sb, 
lo,  Pt,  Scl,  La 
Pe 

P,  preco- 
cious, large 

Howella 

257C 

P 

? 

7 

-3.5 

-4.5 

7 

A 

P5,  Op,  lo,  Pt, 

CI 
P5.  Op,  Sb,  lo. 

None 

Kyphosidae 

259J 

P 

~1 

2.4-2.9 

3.8-5.5 

6-7 

-7 

A 

Spinous  scales 

Scl,  Pel 

Lateolabrax 

260E 

P 

1.3-1.4 

4.4-4.6 

-9 

-15 

>I5 

A 

P5 

None 

Leiognathidae 

256C 

P 

.6 

1.4 

-4 

-5 

7 

7 

S3,  F3,  P4,  Pt. 

Scl 

Ant.  D  spines 
serrate 

Leptobramidae 

258D 

7 

? 

7 

9 

>8 

>9 

7 

None 

None 

Lethrinidae 

262F 

P 

.7-.8 

1.3-1.7 

4.4-5.2 

5.5-7.0 

8 

A 

S3,  F3,  P4,  Op, 
Sb,  lo,  Ta,  Pt, 
Scl.  Pel,  U, 
Co,  Mx,  D, 
Pe 

Spmous  scales 

Lobotidae 

254G 

P 

~1 

7 

<6 

<6 

-8 

7 

S3,  S5,  FI,F4, 
P2,  Op,  Sb, 
lo,  Pt,  Scl 

P,  large  (pre- 
cocious?) 

Lutjanidae 

256A-B 

P 

~.5-.8 

1.7-2.2  4.2-5.3 

5-6 

-12-14 

B 

F3,  (P2).  (P3), 

(2nd  dorsal 

(including 

(P4),  Op,  lo. 

spine  and  P, 

Caesionidae) 

Pt,  Scl.  Pel 

spine  and 
soft  rays 
elongate) 
(anterior  D, 
A, and  P, 
spines  ser- 
rate) 

Malacanthidae 

256E-F 

P 

1.2 

2.2-2.6 

4-6 

5-8 

3-4 

A 

F2,  F3,  P4,  Op, 
Sb,  lo,  Ta,  Pt, 
Scl,  La,  Na, 
D,  Pe,  Pa 

Spinous  scales 
to  -30  mm 
or  >;  fused 
nasals 

Mcnidae 

256D 

7 

? 

7 

<4.5 

<4.5 

7 

9 

None 

None 

Microcanthidae 

259G 

P 

7 

7 

-4 

5-6 

-15 

A 

P2,  Op,  Sb.  lo, 
Pt,  Scl 

None 

Monodactylidae 

255H 

P 

.6-.7 

1.8 

3.5-4.0 

5-6 

9-10 

F 

F3,  P2.  Op,  lo, 
Pt,  Scl 

P,  large,  pre- 
cocious 

Moronidae 

260F 

(P)  (D,  A) 

.7-4.6 

1.7-3.7 

7-9 

10-13 

16-25 

A 

P5 

None 

Mullidae 

259E 

P 

.6-.9 

1.6-3.4 

3.5-4.5 

-7 

-12-13 

A 

None 

Silvery,  pelag- 
ic postlarvae 
to  -40- 
60  mm 

Ncmipteridae 

258H 

P 

.7-.8 

1.5-1.6 

-4 

6-8 

-11 

A 

None 

None 

Opistognathidae 

255B 

D,  A,  M.  O 

7 

9 

-5.5 

-7 

9 

A 

P5,  lo 

None 

Oplegnathidae 

255J 

P 

.9 

2.3 

-5 

-7 

-12 

A 

P5,  Op,  lo,  Scl 

None 

Pemphendidae 

2551 

P 

7 

7 

3.6-4.3 

-6 

7 

F 

P5,  lo,  Scl 

P,  precocious 

Pentacerotidae 

262J 

P 

? 

7 

7 

7 

-12 

7 

S3,  S5,  FI,F3, 
F4,  P4,  Op, 
Pt,  CI,  La,  Pe 

P,  spines  ser- 
rate; spinous 
scales 

Percichthyidae 

260D 

(P)(D,  A) 

1.2-4.2 

3.1-9.0 

7-9 

9-13 

10-20 

D 

None 

None 

488 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  121.    Continued. 


Size  (mm) 


Text 
hgures 


Egg  type 


Egg 


D  &  A  rays 
Hatch  Flex  complete 


First 
stales 


Sequence 

of  fin 
completion 


Head 
spinalion 


Other 

specializations 


Percidae 
Plesiopidae 

Polypnon 

260C 
254E 

(P)(D,  A)(M) 
D,  A,  M,  (O) 
P 

.7-2.8 
-.9  X  .6 
1.6 

Pomacanthidae 

256H 

P 

.7-.9 

Pomatomidae 
Priacanthidae 

258G 
262H-I 

P 
P 

.8-1.2 

9 

Pseudochromidae     255C 
Rachycentridae         26  IB 


Scatophagidae 


262D 


Sciaenidae 

257H 

P 

Scombropidae 

262E 

P 

Scorpididae 

258A-B 

P 

Serranidae 

Serraninae 

- 

P 

Anthiinae 



P 

Epinephelinae 
Epinephelini 


Grammistini 


4.7-8.7      7-15  9-18 
2.8-2.9        ?  ? 

3.7       -7  -9 

1.5-1.8  3.4-4.3  4-5 


2-2.5      5-6  ~7 

?  4-5  ~7 


.7-1.0 


~4 


-4-5 


1.4-2.4  -4-5 


<2.9     3.3-4.6       -6 


13-24 


2.5-2.8 


-12 
-6 


D.  A,  M 

? 

3-4 

4.4-5.8       -8 

12-13 

P 

1.2-1.4 

9 

-7         D  16-18 
A    9-10 

30-35 

-4 


7-1.3 

1.5-2.5  3.0-4.6 

-5-9 

14-20 

? 

?             ? 

>6 

9 

? 

?          -6 

9-10 

9-11 

8-1.0 

-2.2    4.3-5 

~6 

-11-12 

6-.  8 

1.2-1.4  3.5-5 

4.6-5.5 

~6->l 

>15 


A(D) 

(P5) 

None 

9 

9 

9 

A 

S2,  F3,  P2,  Op, 
Sb,  lo,  Pt,  Scl 

None 

A 

F3,  P2,  Sb,  lo. 

Spinous  scales 

Ta.  Pt,  Scl, 

to  17- 

La,  Co,  Na,  D 

19  mm 

A 

P5 

None 

A 

S3,  F2,  F3,  P4, 

D„A,  P, 

Op,  Sb,  lo. 

spines  and 

Ta,  Pt,  Scl, 

soft  rays  ser- 

La, Co,  Na, 

rate;  spinous 

D,  Br 

scales  to 
—  20  mm 

A 

P5 

None 

E 

F4.  P2,  Pt 

Minute  epithe- 
lial pnckles 
by  ~6  mm 
"swollen" 
ptenotics 

D 

Most  exposed 
head  bones 
thick  and  ru- 
gose; P  and 
Pt  expanded 
posteriorly;  Pt 
with  posterior 
spatulate 
"spine";  Pe 
swollen  and 
with  separate 
rugose 
"shield" 

Spinous  scales 

A(B) 

(S4),  (F3),  (P2), 
P5,  Pt,  Scl,  lo 

None 

A 

S2,  F3,  P4,  Op, 
Sb,  lo.  Pt 

None 

A 

P2,  Op,  Sb,  lo, 
Scl 

None 

A(B) 

P2,  Op,  Sb.  lo, 
Pt,  Scl 

None 

B(A) 

(SI),(S2),(F1), 

(D,.  A  and  P, 

(F2),  (F3), 

spines  ser- 

(F4). (P3), 

rate)  (ant  D 

(P4),  Op, 

spines  and 

Sb,  lo,  (Ta). 

P.,  rays  elon- 

Pt, Scl,  (La). 

gate) 

(Co),  (D), 

(Pe),  (Pa) 

B 

(Fl),  F3,  P4, 

D,,  A  and  P^ 

Op,  Sb,  lo,  Pt, 

spines  ser- 

Scl 

rate;  second 
D|  spine 
and  P, 
spines  elon- 
gate 

P„D„D„ 

P2,  Op,  Sb, 

Ant  D  spines 

A,P2 

lo 

flexible, 
elongate, 
pigmented; 
P,  large, 
precocious 

JOHNSON:  PERCOIDEI 


489 


Table  121.    Continued. 


Text 
5gures 

Egg  type 

Size  (mm 

Head 
1              spmation 

Taxon 

Egg 

Hatch 

Flex 

D  &  A  rays 
complete 

First 
scales 

of  fin 
completio 

Other 
specializations 

Liopropomini 

P 

~1 

7 

9 

~6 

7 

D„D„A. 
P„P, 

(F4).  P2,  Op, 
do) 

Ant  D  spines 
flexible, 
elongate,  or- 
namented 

Sillaginidae 

259F 

P 

.6-.7 

1.3 

-6 

-9 

9 

A 

P5,  Pt 

None 

Siniperca 

260B 

D,  A 

-2 

~5 

-10 

-11 

9 

A 

P5 

None 

Sparidae 

2581 

P 

.8-1.2 

2.0-2.7 

4-7 

6-11 

8-20 

A 

P5.  lo,  Pt.  Scl 
(also  S2,  F3, 

None  (spinous 
scales  in  Pa- 

Stereolepis 

254F 

P 

? 

7 

7 

<7 

>10 

7 

P2,  in  Pagrus 
F3.  P2,  Sb.  lo. 
Pt.  Scl,  Pe 

grus) 
None 

Symphysanodon 

254A 

P 

? 

7 

3.5-4.0 

-4.5 

-13-14 

A 

F2,  F5,  P4,  Pt. 
Scl,  Ta,  La, 
Co,  D,  Pe 

None 

Teraponidae 

259H 

(P)(D) 

.7-2.8 

1,7-3.7 

-4-8 

-7-11 

14-18 

A 

P5,  Op,  Sb,  lo, 
Scl,  CI,  Pel 

None 

pletion  of  dorsal  fin  rays  (12-30  mm)  occurs  in  the  elongate 
larvae  of  Coryphaena.  Rachycentron  and  the  Echeneididae. 

The  most  commonly  observed  sequence  of  fin  completion 
(pattern  A  in  Table  1 2 1 )  is  that  described  for  Moronehy  Fritzsche 
and  Johnson  (1980)  and  for  Anisolremus  by  PotthofTet  al.  ( 1 984). 
Soft  rays  of  the  dorsal  and  anal  fins  begin  to  form  during  or  just 
prior  to  flexion.  Fin  rays  appear  first  near  the  future  middle  of 
these  fins  and  are  added  in  an  anterior  and  posterior  direction. 
Full  complements  of  dorsal  and  anal  soft  rays  are  usually  achieved 
at  about  the  same  time  as  the  full  principal  caudal  fin  ray  com- 
plement. The  spinous  dorsal  fin  is  completed  next  (usually  from 
posterior  lo  anterior)  followed  by  the  pelvic  and  pectoral  fins. 

Precocious  development  of  the  anterior  portion  of  the  spinous 
dorsal  and  the  pelvic  fins,  pattern  B,  is  usually  associated  with 
ornamentation  and/or  elongation  of  the  spines.  It  characterizes 
all  larvae  of  lutjanids  and  epinepheline  serranids,  and  a  few 
apogonids,  chaetodontids  and  sciaenids.  In  liopropomine  ser- 
ranids, the  anterior  portion  of  the  spinous  dorsal  is  precocious, 
but  the  pelvic  fins  develop  last.  Precocious  development  of  pec- 
toral and  pelvic  fins,  pattern  C,  is  unique  to  some  members  of 
the  Bramidae.  Pattern  D,  precocious  pectorals  only,  is  found  in 
scatophagids,  some  bramids,  and  interestingly,  is  also  shared 
by  the  freshwater  Percichthyidae  and  some  Percidae.  The  pec- 
toral fin  and  anterior  portion  of  the  spinous  dorsal  are  precocious 
in  the  serranid  tribe  Grammistini.  In  pattern  E,  the  full  anal  fin 
ray  complement  tends  to  be  complete  prior  to  that  of  the  dorsal, 
and  the  spinous  dorsal  is  the  last  fin  to  be  completed.  This 
pattern  is  unique  to  the  echeneoid  fishes  (Coryphaenidae,  Ra- 
chycentridae  and  Echeneididae).  Pattern  F,  in  which  only  the 
pelvics  are  precocious,  is  found  in  Hapalogenys.  Monodactyl- 
idae  and  Pempherididae. 

Scales.  — MoiX  percoids  begin  to  develop  scales  well  after  com- 
pletion of  fins  near  the  end  of  the  larval  period,  frequently  after 
settling.  In  several  families  (e.g.,  Chaetodontidae,  Cheilodac- 
tylidae,  Cirrhitidae,  and  Scorpididae)  unspecialized  scales  first 
appear  at  or  slightly  before  completion  of  the  median  fins  and 
are  thus  present  during  the  late  larval  stages.  Larvae  of  a  few 
groups  are  characterized  by  early  development  of  specialized 


spinous  scales  that  eventually  transform  into  the  typical  adult 
ctenoid  scale.  In  the  ephippidid  Chaetodiplerus.  the  haemulid 
Conodon.  malacanthids,  pomacanthids  and  scatophagids  these 
consist  of  small,  roughly  circular,  non-imbricate  bony  plates 
from  the  center  of  which  one  to  several  spines  project  outward 
at  right  angles.  Larvae  of  the  Bramidae,  Kyphosidae,  Pentacer- 
otidae,  Priacanthidae,  some  anthiin  serranids,  the  sparid  Pagrus 
and  the  sparoid  family  Lethrinidae  possess  spinous  scales  in 
which  one  or  more  spines  project  outward  at  less  than  right 
angles  from  the  posterior  field  or  margin  of  imbricate  plates  that 
more  closely  resemble  scales  of  the  adults.  Among  non-percoid 
fishes,  spinous  larval  scales  occur  in  trachichthyids,  chiasmo- 
dontids,  acanthurids,  Xiphias,  Anttgoma  and  some  pleuronec- 
tiforms,  telraodontiforms,  scorpaeniforms  and  gasterostei- 
forms.  The  function  of  specialized  larval  scales  is  unknown,  but 
it  seems  likely  that  they  provide  some  defense  against  small 
biting  predators,  parasites  and/or  nematocysts. 

Head  spination.— The  simple  to  elaborate  spinous  ornamenta- 
tion of  various  bones  of  the  head  in  larvae  of  many  percoid 
fishes  is  an  area  ripe  for  future  detailed  investigations.  Nowhere 
is  the  potential  utility  of  larval  morphology  in  phylogenetic 
studies  more  evident,  for  it  is  in  this  feature  that  larval  percoids 
frequently  exhibit  far  more  complexity  and  diversity  than  adults. 
Although  more  work  is  needed  to  determine  if  patterns  of  head 
spination  will  prove  useful  in  studies  of  interfamilial  relation- 
ships, there  can  be  no  doubt  that  the  diversity  of  these  patterns 
within  some  well-defined  families  or  subfamilies  (e.g.,  anthiin 
serranids,  chaetodontids,  priacanthids,  malacanthids,  poma- 
canthids. haemulids.  etc.)  offer  critical  information  for  intra- 
familial  phylogenetic  analyses. 

Extensive  head  spination  appears  to  have  arisen  indepen- 
dently numerous  times  within  the  Percoidei.  Nevertheless,  an 
ordered  progression  of  increasing  complexity  is  evident  in  the 
sequence  in  which  ornamentation  is  added  to  various  bones. 
Most  families  are  characterized  by  a  single  level  of  complexity, 
but  some  are  more  diverse.  In  the  larvae  of  several  unrelated 
families  (e.g.,  Cheilodactylidae,  Echeneididae,  Menidae,  Mul- 
lidae,  Percichthyidae)  head  spines  are  completely  lacking.  A 


490 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


somewhat  larger,  equally  heterogeneous  assemblage  of  percoid 
groups  (including  the  Ambassidae,  Centracanthidae,  Centro- 
pomidae.  Cirrhitidae,  Moronidae,  Percidae,  Pomatomidae  and 
Pseudochromidae)  has  minimal  head  spination,  consisting  of 
only  a  few  small  spines  along  the  posterior,  and  usually  lateral, 
margins  of  the  preopercle.  In  most  instances,  these  spines  are 
so  small  and  isolated  that  it  is  difficult  to  imagine  that  they  serve 
any  useful  function. 

The  most  common  pattern  of  head  spination  among  larval 
percoids  is  one  in  which,  in  addition  to  small  to  moderate  pre- 
opercular  spines,  small  spines  may  also  occur  on  other  bones 
of  the  opercular  series  (interopercle,  subopercle  and  opercle)  and 
on  various  bones  of  the  pectoral  series  (cleithrum,  postcleith- 
rum,  supracleithrum,  posttemporal  and  tabulars).  This  pattern 
occurs  in  many  of  the  more  generalized  families  that  have  usu- 
ally been  considered  "basal"  percoids,  including  the  Acropo- 
matidae,  Gerreidae,  Girellidae,  Haemulidae,  Kyphosidae, 
Sciaenidae,  Scorpididae,  Sparidae  and  Teraponidae,  and  it  must 
be  primitive  for  at  least  some  large  subgroup  of  percoid  families. 

Two  additional  levels  of  complexity  in  this  artificial  hierarchy 
involve  modifications  of  cranial  bones  (frontal  and  supraoccip- 
ital)  in  addition  to  opercular  and  pectoral  series  spination.  Mod- 
ifications of  the  frontal  bones  occur  only  in  those  larvae  with 
opercular  and  pectoral  series  spination  and  encompass  several 
types  of  ornamentation.  Frontal  surface  rugosity  is  found  in  a 
few  apogonids,  bramids  and  serranids  as  well  as  in  Acantho- 
cepola,  Lobotes,  Hapalogenys,  Pseudopenaceros  and  Sphyrae- 
nops.  Johnson  and  Keener  (1984)  noted  this  condition  in  larval 
Alphestes.  but  it  was  not  previously  considered  in  descriptions 
of  percoid  larvae.  With  closer  examination,  cranial  rugosity  will 
undoubtedly  be  detected  in  larvae  of  other  percoid  and  non- 
percoid  groups.  It  probably  offers  an  efficient  way  to  strengthen 
the  neurocranium  during  early  development.  Frontal  spines  or 
serrations  are  most  frequently  borne  along  the  supraorbital  ridge. 
Coryphaena,  Rachycentron,  Lobotes.  and  some  carangids  have 
one  large,  broad-based  supraorbital  spine,  but  the  more  com- 
mon condition  is  a  series  of  supraorbital  spines  or  serrations. 
These  are  found  in  lutjanids,  malacanthids,  monodactylids,  po- 
macanthids,  Stereolepis.  some  acropomatids,  carangids,  hae- 
mulids,  sciaenids,  and  serranids  as  well  as  in  most  groups  with 
supraoccipital  modifications.  More  elaborate  ornamentation, 
consisting  of  a  series  of  parallel  serrated  ridges  on  the  dorsal 
surface  of  the  frontals,  characterizes  larval  malacanthids,  pria- 
canthids,  Synagrops  and  some  anthiin  serranids. 

The  most  extreme  example  of  frontal  spination  is  seen  in 
Symphysanodon  (Fig.  254A).  A  longitudinal  serrated  crest  above 
the  supraorbital  ridge  on  each  frontal  bone  continues  posteriorly 
as  a  long,  spike-like  serrated  spine  extending  to  about  the  middle 
of  the  spinous  dorsal  fin.  The  only  other  example  of  large  paired 
cranial  spines  among  larval  perciforms  is  found  in  istiophorids, 
where  the  spines  originate  from  the  pterotics.  This  "homed" 
effect  occurs  elsewhere  in  larvae  of  many  scorpaeniform  groups 
(e.g.,  Scorpaenidae  and  Triglidae)  and  in  the  beryciforms,  Di- 
retmus  and  Anoplogaster,  but  in  these  groups  the  large  paired 
spines  are  parietal  in  origin.  With  the  exception  of  occasional 
minute  spines  or  small  ridges,  larvae  of  perciform  fishes  never 
develop  parietal  ornamentation,  and  it  is  tempting  to  speculate 
that  the  presence  of  variously  developed  parietal  spines  among 
larvae  of  many  scorpaeniform  groups  offers  support  for  the  often 
questioned  monophyly  of  the  Scorpaeniformes.  In  any  case,  this 
uncommon  feature  should  be  examined  in  future  considerations 
of  higher  relationships  among  acanthopterygian  fishes.  The 


monophyly  of  the  Beryciformes  has  recently  been  questioned 
(Zehren,  1979),  and  it  is  interesting  to  note  that  although  Di- 
relmus.  Anoplogaster  and  at  least  some  trachicthyoids  share 
larval  parietal  spines  with  scorpaeniforms,  holocentrids  lack 
them,  instead  possessing  frontal,  supraoccipital  and  preoper- 
cular  spination  similar  to  that  seen  in  more  elaborately  orna- 
mented larval  percoids. 

Modifications  of  the  supraoccipital,  representing  the  last  cat- 
egory of  complexity  in  head  spination,  occur  in  those  larvae 
which  also  have  opercular  series,  pectoral  series  and  frontal 
ornamentation.  Simple  forms  of  supraoccipital  ornamentation 
include  a  small  peak-like  median  crest  (Chaetodipterus,  Pagrus, 
Polyprion,  Sphyraenops,  and  some  acropomatids,  apogonids, 
carangids  and  anthiin  serranids)  or  a  serrated,  ridge-like  crest 
(Synagrops.  some  sciaenids  and  anthiin  serranids).  The  more 
extreme  form  is  a  large,  vaulted,  variously  serrate  spine-like 
crest  that  projects  beyond  the  posterior  margin  of  the  cranium 
and  is  well-developed  in  preflexion  larvae  soon  after  hatching. 
This  type  of  crest  characterizes  larval  cepolids,  Hapalogenys, 
leiognathids,  lethrinids,  (lobotids?),  pentacerotids,  priacanthids 
and  Scoinbrops.  To  my  knowledge,  it  occurs  elsewhere  only  in 
the  larvae  of  holocentrid  beryciforms  and  the  caproid  Antigonia. 

The  so  called  "tholichthys"  larvae  of  the  Chaetodontidae  and 
Scatophagidae  (Fig.  262A-D)  perhaps  represent  the  ultimate  in 
head  bone  modification  among  larval  percoids.  The  cranial  bones 
and  many  of  the  other  exposed  bones  of  the  head  are  thickened 
and  rugose,  effecting  an  armor-like  protective  covering.  In  chae- 
todontids  the  posttemporal  and  supracleithrum  are  rugose  and 
expanded  posteriorly  as  large  laminar  plates.  The  preopercle  is 
similarly  expanded  anteriorly  and  posteriorly  and  at  its  angle 
bears  a  broad,  flattened  or  serrated,  terete  spine.  In  scatophagids 
the  preopercle  is  rugose  and  expanded,  but,  unlike  chaetodon- 
tids,  the  supracleithrum  is  unmodified.  The  posttemporal  is 
rugose,  its  dorsal  portion  is  somewhat  expanded,  and  its  ventral 
half  extends  posteriorly  as  a  very  blunt,  thick,  spine-like  pro- 
jection. Also  notable  is  a  large,  thick,  rugose  protuberance  cov- 
ering the  pterotic.  Although  not  identical,  the  larvae  of  chae- 
todontids  and  scatophagids  share  a  unique  physiognomy,  the 
details  of  which  should  be  investigated  in  relation  to  possible 
close  affinity  of  these  two  families. 

Spination  on  circumorbital,  nasal,  premaxillary  and  maxillary 
bones  is  generally  found  only  in  those  larval  percoids  with  cra- 
nial ornamentation,  and  it  is  almost  exclusively  in  these  larvae 
that  other  specializations,  such  as  elongate  serrate  fin  spines  and 
spinous  scales  occur.  In  addition,  opercular  and  pectoral  series 
spination  is  usually  more  extensive  and  almost  always  includes 
an  elongate  and/or  serrate  spine  at  the  angle  of  the  preopercle. 

In  summary,  there  seem  to  have  been  some  common  evo- 
lutionary constraints  on  the  order  in  which  morphological  com- 
plexity and  specialization  of  larval  percoids  has  progressed,  but 
a  simple  direct  relationship  between  this  ordered  progression 
and  phylogenetic  affinity  among  families  is  not  apparent.  In  fact, 
the  assemblages  of  taxa  that  characterize  the  various  levels  of 
complexity  discussed  above  are  quite  diverse  and  not  compat- 
ible with  what  little  we  do  understand  about  percoid  affinities 
based  on  adult  morphology.  Furthermore,  it  is  clear  that  elab- 
orately ornamented  larvae  have  arisen  independently  several 
times  within  monophyletic  groups  otherwise  characterized  by 
larvae  with  only  generalized  opercular  and  preopercular  spi- 
nation. Examples  include  the  haemulid  Conodon,  the  sparoid 
family  Lethrinidae  and  the  serranid  subfamily  Anthiinae.  Res- 
olution of  the  phylogenetic  significance  of  intricate  patterns  of 


JOHNSON:  PERCOIDEI 


491 


head  spination  among  larval  percoids  will  entail  more  precise 
study  than  has  characterized  much  previous  work.  Determi- 
nation of  homology  will  require  detailed  information  about  lo- 
cation, conformation  and  processes  of  development  of  head 
spines  prior  to  considering  the  question  of  compatibility  with 
adult  characters. 

Utility  of  Larval  Morphology  in 
Phylogenetic  Studies 

The  preceding  two  decades  have  seen  notable  advances  in  our 
understanding  of  the  evolutionary  relationships  of  teleost  fishes; 
however,  as  noted  above,  progress  in  elucidating  the  phylogeny 
of  the  Percoidei  has  not  kept  pace.  Many  families  are  poorly 
delineated  and  hypotheses  about  inter-  and  intrafamilial  rela- 
tionships are  few.  Lack  of  progress  is  chiefly  attributable  to  the 
size  and  diversity  of  the  Percoidei.  the  adaptive  malleability  and 
convergence  that  have  characterized  percoid  evolution  and  the 
paucity  of  conspicuous  morphological  specializations  that  can 
be  readily  identified  as  true  synapomorphies.  With  few  excep- 
tions (Burgess,  1974;  Dooley,  1978;  Kendall,  1979;  Johnson, 
1983),  previous  studies  of  percoid  phylogeny  and  classification 
have  failed  to  consider  early  life  history  stages,  even  though  it 
is  obvious  that  the  prodigious  variety  of  larval  form  and  spe- 
cialization among  percoids  offers  a  rich  suite  of  additional  char- 
acters. 

Within  many  families  there  is  a  complexity  of  larval  mor- 
phology or  diversity  of  larval  form  that  suggests  excellent  po- 
tential for  the  application  of  larval  characters  in  elucidating 
generic  interrelationships.  Particularly  promising  families  in  this 
regard  include  the  Acropomatidae,  Apogonidae,  Bramidae,  Ca- 
rangidae,  Cepolidae,  Chaetodontidae,  Haemulidae,  Lutjanidae, 
Malacanthidae,  Pentacerotidae,  Pomacanthidae,  Priacanthidae, 
Sciaenidae,  and  Serranidae.  The  intricate  bony  ornamentation 
of  the  larvae  of  anthiin  serranids,  for  instance,  is  considerably 
more  complex  than  that  of  the  adults,  and  preliminary  studies 
of  details  of  larval  head  spination  and  scale  development  among 
New  World  genera  indicate  that  the  current  generic  classifica- 
tion, based  exclusively  on  adult  morphology,  should  be  reex- 
amined (Carole  Baldwin,  Abstracts  of  1 983  ASIH  Annual  Meet- 
ing). Larvae  of  groups  like  the  apogonids  and  carangids  exhibit 
a  less  complex  morphology,  but  the  wide  range  of  form  and 
specialization  should  prove  useful  in  phylogenetic  analyses. 

Larval  morphology  will  undoubtedly  also  prove  useful  in  con- 
siderations of  higher  relationships  among  percoids.  At  the  fam- 
ily level,  a  rather  simplistic  approach  is  to  consider  that  larvae 
offer  independent  tests  of  hypotheses  of  monophyly.  In  other 
words,  do  the  larvae  of  each  percoid  family  share  one  or  more 
derived  features  that  corroborate  the  monophyly  of  that  family 
as  currently  defined  on  the  basis  of  adult  morphology?  The 
answer  to  this  question  appears  to  be  yes  for  many  groups,  but 
problems  stem  from  an  inadequate  understanding  of  character 
polarity  and  the  fact  that,  for  most  families,  larvae  of  many 
genera  and  most  species  remain  undescribed.  Nonetheless,  this 
is  a  useful  concept,  and  the  validity  and  power  of  such  a  test 
will  increase  as  we  gain  more  knowledge  of  the  larvae  of  various 
percoid  groups. 

Consider,  for  example,  the  bearing  of  larval  morphology  on 
several  hypotheses  of  relationship  resulting  from  the  recent  re- 
definition of  Schultz's  (1945)  Emmelichthyidae,  a  polyphyletic 
assemblage  of  planktivorous  fishes.  Heemstra  and  Randall  (1977) 
transferred  Diptcrygonolus  to  the  Caesionidae  and  Johnson 
(1980)  hypothesized  that  caesionids  are  lutjanoid  fishes  most 


closely  related  to  the  lutjanid  subfamily  Lutjaninae.  Caesionids 
are  quite  distinctive  in  body  form  and  upper  jaw  configuration, 
but  share  with  the  lutjanines  a  number  of  osteological  features 
and  a  specialized  adductor  mandibulae  (similar  to  that  of  most 
carangids)  in  which  a  separate  division  of  A,  originates  on  the 
subocular  shelf  Subsequent  descriptions  of  larval  lutjanines  and 
caesionids  (see  Table  122)  show  that  they  share  a  distinctive 
body  form,  pattern  of  head  spination,  precocious  first  dorsal 
and  pelvic  fins  with  elongate  spines  and  soft  rays,  and  sparse 
pigmentation  (Fig.  256A,  B).  The  hypothesized  sister  group  re- 
lationship is  thus  corroborated  by  larval  morphology. 

The  Centracanthidae  were  also  removed  from  the  Emme- 
lichthyidae and  hypothesized  to  be  most  closely  related  to  the 
Sparidae  (Heemstra  and  Randall,  1977;  Johnson,  1980)  based 
on  adult  morphology.  Although  the  larvae  of  these  two  groups 
share  no  obvious  specializations,  they  are  quite  similar  (Fig. 
2581,  J),  and  are  distinguishable  from  those  of  the  Emme- 
lichthyidae (Fig.  2591)  and  the  other  reassigned  groups.  Labra- 
coglossa,  placed  in  a  separate  family  by  Heemstra  and  Randall 
(1977)  is  here  placed  in  the  family  Scorpididae  (see  section  on 
classification),  and  the  larval  form  corroborates  this  placement 
(Fig.  258A,  B).  The  larvae  of  inermiids,  Inermia  and  Emme- 
lichthyops.  also  removed  from  the  Emmelichthyidae,  remain 
undescribed,  but  their  identification  can  provide  a  test  of  the 
hypothesis  that  they  are  most  closely  related  to  the  Haemulidae 
(Johnson,  1980). 

These  examples  and  those  that  follow  demonstrate  that  early 
life  history  stages  offer  important  information  that  can  be  used 
to  test  previous  phylogenetic  hypotheses  or  incorporated  with 
adult  characters  into  new  phylogenetic  analyses.  Additional  ex- 
amples are  mentioned  in  the  discussion  of  familial  classification. 
Where  the  larvae  are  known,  failure  to  consider  their  mor- 
phology in  studies  of  percoid  phylogeny  seems  hardly  justifiable, 
and  may  inhibit  progress  or  lead  to  false  conclusions.  This  point 
is  well-illustrated  in  the  two  examples  discussed  below,  in  which 
details  of  larval  morphology  provide  critical  evidence  in  support 
of  new  or  previously  rejected  phylogenetic  hypotheses. 

The  families  Branchiostegidae  (=Latilidae)  and  Malacanthi- 
dae have  been  variously  united  and  separated  in  past  classifi- 
cations. In  the  most  recent  revision,  Dooley  (1978)  concluded 
that  "the  branchiostegids  and  malacanthids  have  few  characters 
in  common  that  might  be  used  to  justify  their  consolidation 
into  a  single  family"  and  noted  that  they  "could  as  easily  be 
aligned  with  several  other  percoid  families  as  with  each  other." 
He  suggested  that  the  malacanthids  are  possibly  "a  branch  of 
the  labrid-scarid  lineage,  while  the  branchiostegids  show  closer 
affinities  to  the  serranid-percid  line  of  perciform  evolution."  In 
contrast,  Robins  et  al.  ( 1 980)  recognized  a  close  affinity  between 
the  two  groups  by  treating  them  as  subfamilies  of  the  Malacan- 
thidae. Marino  and  Dooley  (1982)  took  issue  with  this  classi- 
fication and  stated  that  there  are  "several  more  myological  (dif- 
ferences) why  the  families  are  distinct."  Actually,  Marino  and 
Dooley  listed  only  one  myological  difference,  the  absence  of 
adductor  mandibulae  section  A,,,.  This  difference  and  the  other 
1 3  listed  by  Dooley  ( 1 978.  Table  1 ),  including  body  depth,  body 
shape,  and  skull  contour,  have  little  relevance  to  the  phyloge- 
netic affinity  of  these  two  groups.  As  for  features  common  to 
the  malacanthids  and  branchiostegids,  Dooley  found  only  three: 
dorsal  and  anal  fins  relatively  long  and  continuous,  a  single 
opercular  spine,  and  "grossly  similar  larval  stages."  Dooley  cor- 
rectly noted  that  the  first  two  of  these  are  not  particularly  mean- 
ingful because  they  are  fairly  common  percoid  features,  but  he 


492 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  122.    References  to  Larval  Percoidei. 


Eggs 


Poslflexion 


Acanthoclinidae 

Acropomatidae 
Ambassidae 

Apogonidae 


Bramidae 

Caesionidae 
Callanthiidae 


Carangidae 
Caristiidae 
Centracanthidae 


Centrarchidae 

Centropomidae 

Cepolidae 


Chaetodontidae 


Cheilodactylidae 


Cirrhitidae 


Congrogadidae 

Coracinidae 

Coryphaenidae 


Jillett,  1968 


Breder  and  Rosen,  1966 

Eng,  1969 

Nair,  1958 

Breder  and  Rosen,  1966 

Leis  and  Rennis,  1983 

Allen,  1975b 

Bertolini,  1933a 


Jillett,  1968 


Eng,  1969 


Jillett,  1968 


Leis  and  Rennis,  1983 
Miller  etal.,  1979 
Allen,  1975b 
Bertolini,  1933a 


Leis  and  Rennis,  1983 
Miller  etal.,  1979 
Allen,  1975b 
De  Gaetani,  1937 


Johnson,  1978 
Mead,  1972 


Johnson,  1978 
Mead,  1972 


—  —  Leis  and  Rennis,  1983 

Bertolini,  1933b 
Page,  1918 

Laroche  et  a!.,  this  volume 

Brownell,  1979 

Thomopoulos,  1954 

Aboussouan,  1964 

Montalenti,  1933 

Sanzo,  1939c 

Numerous  references,  see  Breder  and  Rosen,  1966;  Hardy,  1978b;  and  Auer,  1982 


Brownell,  1979 
Sanzo,  1939c 


Brownell,  1979 
Sanzo,  1939c 


Lauand  Shafland,  1982 
Breder  and  Rosen,  1966 
Russell,  1976 
Holt,  1891 
Montalenti,  1937b 


Leis  and  Rennis,  1983 
Burgess,  1978 
Suzuki  etal.,  1980 


Brownell,  1979 
Mito,  1963 
Robertson,  1978 
Gilchrist  and  Hunter, 

1919 
Barnard,  1927 


Lauand  Shafland,  1982 


Leis  and  Rennis,  1983 
Suzuki  etal.,  1980 


Brownell,  1979 
Robertson,  1978 


Johnson,  1978 
Miller  etal.,  1979 
Mito,  1960 


Johnson,  1978 
Miller  etal.,  1979 
Mito,  1960 


Lau  and  Shafland,  1982 
Russell,  1976 
Page,  1918 
Montalenti,  1937b 
Okiyama,  1982b 


Leis  and  Rennis,  1983 
Suzuki  etal.,  1980 


Brownell,  1979 

Gilchnst  and  Hunter,  1919 

Hatton,  1964 


Leis  and  Rennis,  1983 


Johnson,  1978 
Miller  etal.,  1979 
Mito,  1960 
Potthoff,  1980 


Crossland,  1981 
Crossland,  1982 
Jillett,  1968 
Pourmanoir,  1976 
Okiyama,  1982b 
Nair,  1952b 
Gopinath,  1946 
Nair,  1958 

Leis  and  Rennis,  1983 
Miller  etal.,  1979 
Allen,  1975b 
Pourmanoir,  1976 
Okiyama,  1982b 
Bertolini,  1933a 
Pahay,  1975 
Whitley,  1926 
Vatanachi,  1972 
De  Gaetani,  1937 
Johnson,  1978 
Mead,  1972 
Pahay,  1983 
Leis  and  Rennis,  1983 
Leis  and  Rennis,  1983 
Pourmanoir,  1976 
Bertolini,  1933b 
Page.  1918 

Belyanina,  1982b 
Brownell,  1979 
Page,  1918 
Montalenti,  1933 


Lau  and  Shafland,  1982 
Russell,  1976 
Pourmanoir,  1976 
Clark,  1920 
Page,  1918 
Montalenti,  1937 
Pourmanoir,  1973 

Leis  and  Rennis,  1983 
Burgess,  1978 
Pourmanoir,  1976 
Kendall  and  Goldsborough, 

1911 
Burgess,  1974 
Brownell,  1979 
Dudnik,  1977 
Vooren,  1972 
Tong  and  Saito,  1977 
Nielsen,  1963a 
Hatton,  1964 
Leis  and  Rennis,  1983 
Pourmanoir,  1973 
Pourmanoir,  1971a 
Whitley,  1926 
Smith,  1938 
Johnson,  1978 
Miller  et  al.,  1979 
Gibbs  and  Collette.  1959 
Aboussouan.  1969 
Potthotf,  1980 


JOHNSON:  PERCOIDEI 


493 


Table  122.    Continued. 


Eggs 


Yolk-sac 


Preflcxion 


Poslflexion 


Echeneididae 


Emmelichthyidae 
Ephippididae 


Epigonrdae 
Gerreidae 


Girellidae 


Haemulidae 


Hapalogenys 

Howella 
Kyphosidae 


Lactariidae 
Laleolabrax 

Leiognathidae 

Lethrinidae 

Lobotidae 
Lutjanidae 


John,  1950 

Sanzo,  1930a 

Martin  and  Drewry,  1978 

Sanzo.  1928 

Akazaki  et  al.,  1976 

Breder  and  Rosen    1966 
Johnson.  1978 
Ryder.  1887 


Leis  and  Rennis,  1983 
Rass.  1972 

Breder  and  Rosen,  1966 
Uchidaelal.,  1958 
Mito,  1957a 

Breder  and  Rosen,  1966 
Leis  and  Rennis.  1983 
Johnson.  1978 
Mito.  1966 
Podosinnikov.  1977 
Saksena  and  Richards. 

1975 
Hildebrand  and  Cable. 

1930 
Fahay.  1983 
Suzuki  et  al..  1983 


Leis  and  Rennis.  1983 
Miller  et  al..  1979 
Watson  and  Leis.  1974 

Breder  and  Rosen.  1966 
Chacko.  1944 
Breder  and  Rosen.  1966 
Mito,  1957b 
Uchidaet  al.,  1958 
Breder  and  Rosen,  1966 
Fujita,  1960 

Leis  and  Rennis,  1983 
Suzuki  and  Hioki,  1978 
Renzhai  and  Suif'en, 

1980a 
Mito,  1956a 
Hardy,  1978b 
Gudger,  1931 

Leis  and  Rennis,  1983 
Suzuki  and  Hioki,  1979b 
Rabalaiset  al.,  1980 
Stark,  1971 
Mon,  1984 


John,  1950 

Sanzo,  1930a 

Martin  and  Drewry.  1978 

Sanzo,  1928 

Akazaki  et  al..  1976 

Johnson,  1978 
Ryder,  1887 


Leis  and  Rennis,  1983 


Uchidaet  al.,  1958 
Mito,  1957a 


Leis  and  Rennis,  1983 
Johnson,  1978 
Mito,  1966 
Podosinnikov,  1977 
Saksena  and  Richards, 

1975 
Hildebrand  and  Cable, 

1930 
Fahay.  1983 

Suzuki  et  al.,  1983 


Leis  and  Rennis,  1983 
Miller  etal..  1979 


Mito.  1957b 
Uchida  et  al.. 

Fujita.  1960 


1958 


Leis  and  Rennis.  1983 
Suzuki  and  Hioki.  1978 
Renzhai  and  Suifen. 

1980a 
Mito.  1956a 


Leis  and  Rennis.  1983 
Suzuki  and  Hioki.  1979b 
Rabalais  et  al.,  1980 
Mori,  1984 


John.  1950 

Martin  and  Drewry,  1978 

Sanzo.  1928 

Akazaki  et  al..  1976 


Johnson.  1978 
Hildebrand  and  Cable, 

1938 
Fahay,  1983 

Leis  and  Rennis,  1983 


Uchidaet  al.,  1958 
Mho,  1957a 


Leis  and  Rennis,  1983 
Johnson,  1978 
Saksena  and  Richards, 

1975 
Hildebrand  and  Cable, 

1930 
Fahay,  1983 


Suzuki  et  al.,  1983 

Gonzales.  1946 

Leis  and  Rennis,  1983 

Miller  etal.,  1979 


Mito,  1957b 
Uchidaet  al.,  1958 


Malacanthidae 


Breder  and  Rosen,  1966 
Fischer.  1958 


Fischer,  1958a 
Fahay,  1983 


Fujita,  1960 


Leis  and  Rennis,  1983 


Hardy,  1978b 
Uchidaet  al.,  1958 

Leis  and  Rennis,  1983 
Richards  and  Saksena. 

1980 
Collins  et  al..  1980 
Laroche.  1977 
Mon.  1984 


Fischer.  1958a 
Okiyama,  1964 


Gudger,  1926 
Gudger,  1928 
Akazaki  et  al.,  1976 


Nakahara,  1962 
Johnson,  1978 
Hildebrand  and  Cable.  1938 
Fahay.  1983 

Mayer.  1972 
Leis  and  Rennis.  1983 
Nair,  1952b 
Uchidaet  al.,  1958 
Kobayashi  and  Igarashi, 

1961 
Munro,  1945 
Uchidaet  al.,  1958 
Leis  and  Rennis,  1983 
Johnson,  1978 
Saksena  and  Richards,  1975 
Hildebrand  and  Cable,  1930 
Nellen,  1973b 
Fahay,  1983 
Heemstra,  1974 


Okiyama,  1982b 
Suzuki  etal.,  1983 
Gonzales,  1946 
Leis  and  Rennis,  1983 
Moore,  1962 
Johnson,  1978 
Uchidaet  al.,  1958 
Nair,  1952b 

Okiyama,  1982b 
Mito,  1957b 
Uchida  etal.,  1958 
Nair,  1952b 
Vatanachi,  1972 
Gopinath,  1946 
Leis  and  Rennis,  1983 


Hardy,  1978b 
Okiyama,  1982b 
Uchidaet  al.,  1958 
Leis  and  Rennis,  1983 
Fourmanoir,  1976 
Okiyama,  1982b 
Richards  and  Saksena,  1980 
Collins  et  al.,  1980 
Fahay,  1975 
Heemstra,  1974 
Vatanachi,  1972 
Stark,  1971 

Musiy  and  Sergiyenko,  1977 
Laroche,  1977;  Mori,  1984 
Fourmanoir,  1970,  1976 
Dooley,  1978 


494 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  122.    Continued. 


Eggs 


Preflexion 


Fahay,  1983 


Microcanthidae 
Monodactylidae 
Moronidae 


Mullidae 


Nemipteridae 


Opistognathidae 
Oplegnathidae 


Pempheridae 
Pentacerotidae 

Percichthyidae 


Percidae 
Plesiopodae 

Polyprion 


Pomacanthidae 


Pomatomidae 


Akatsuet  al.,  1977 

Breder  and  Rosen.  1966 
Hardy,  1978b 
Mansueti,  1964 
Ryder,  1887 
Mansueti,  1958 
Pearson,  1938 


Breder  and  Rosen,  1966 
Leis  and  Rennis,  1983 
Russell,  1976 
Miller  et  al.,  1979 
Marinaro,  1971 
Raffaele,  1888 
Heincke  and  Ehrenbaum, 

1900 
Leis  and  Rennis,  1983 
Aoyama  and  Sotogaki, 

1955 
Renzhai  and  Suifen, 

1980b 

Breder  and  Rosen,  1966 
Mito,  1956b 
Uchidaet  al.,  1958 
Leis  and  Rennis,  1983 


Breder  and  Rosen,  1 966 
Dakin  and  Kesteven, 

1938 
Llewellyn,  1974 
Lake,  1967 
Jackson,  1978 
Fuster  de  Plaza  and  Plaza, 

1955 


Akatsuet  al.,  1977 

Hardy,  1978b 
Mansueti,  1964 
Ryder,  1887 
Mansueti,  1958 
Pearson,  1938 
Doroshev,  1970 


Leis  and  Rennis.  1983 
Russell,  1976 
Marinaro,  1971 
Raffaele,  1888 
Heincke  and  Ehrenbaum, 
1900 


Leis  and  Rennis,  1983 
Aoyama  and  Sotogaki, 

1955 
Renzhai  and  Suifen, 

1980b 

Fukuharaand  Ito,  1978 
Mito,  1956b 
Uchidaet  al.,  1958 
Leis  and  Rennis,  1983 


Dakin  and  Kesteven, 

1938 
Llewellyn,  1974 
Lake.  1967 
Jackson,  1978 


Fahay,  1983 


Leis  and  Rennis,  1983 
Uchidaet  al.,  1958 
Akatsuet  al.,  1977 

Hardy,  1978b 
Mansueti,  1964 
Ryder,  1887 
Mansueti,  1958 
Pearson,  1938 
Doroshev,  1970 
Fritzsche  and  Johnson, 

1980 
Leis  and  Rennis,  1983 
Russell,  1976 
Miller  etal.,  1979 
Heincke  and  Ehrenbaum, 

1900 
Montalenti,  1937 
Uchidaet  al.,  1958 
Lo  Bianco,  1908b 
Leis  and  Rennis,  1983 


Fukuhara  and  Ito,  1978 
Uchidaet  al.,  1958 

Leis  and  Rennis,  1983 


Dakin  and  Kesteven,  1938 
Llewellyn,  1974 
Lake,  1967 
Jackson,  1978 


Numerous  references,  see  Breder  and  Rosen,  1966;  Hardy,  1978b;  and  Auer,  1982 


Breder  and  Rosen,  1 966 
Mito,  1955 
Hardy,  1978b 
Sparta,  1939a 
Thomson  and  Anderton, 

1921 
Leis  and  Rennis,  1983 
Suzuki  etal..  1979 
Fujita  and  Mito,  1960 

Hardy,  1978b 
Deuel  etal.,  1966 
Dekhnik,  1973 
Salekhova,  1959 
Sparta,  1962 
Fahay,  1983 


Mito,  1955 

Hardy,  1978b 
Sparta,  1939a 


Leis  and  Rennis,  1983 
Suzuki  etal.,  1979 
Fujita  and  Mito,  1960 

Hardy,  1978b 
Deuel  et  al.,  1966 
Dikhnik,  1973 
Salekhova,  1959 
Sparta,  1962 
Fahay,  1983 


Hardy,  1978b 
Sparta,  1939a 


Leis  and  Rennis,  1983 
Burgess,  1974 


Hardy,  1978b 
Deuei  et  al.,  1966 
Dekhnik,  1973 
Salekhova,  1959 
Sparta,  1962 
Norcross  et  al.,  1974 
Pearson,  1941 
Fahay,  1983 


Moser,  1981 

Okiyama,  1964 

Okiyama,  1982b 

Fahay,  1983 

Berry,  1958 

Hubbs,  1958 

Leis  and  Rennis,  1983 

Uchidaetal.,  1958 

Akatsu  et  al.,  1977 

Ogasawara  et  al.,  1978 

Hardy,  1978b 

Mansueti,  1964 

Mansueti,  1958 

Pearson,  1938 

Doroshev,  1970 

Okiyama,  1982b 

Fritzsche  and  Johnson,  1980 

Leis  and  Rennis,  1983 
Johnson,  1978 
Russell.  1976 
Miller  etal.,  1979 
Uchidaetal.,  1958 
Vatanachi,  1972 
M.  C.  Caldwell,  1962 
Lo  Bianco,  1908b 
Leis  and  Rennis,  1983 


Vatanachi,  1972 
Fukuharaand  Ito,  1978 
Fuskusho,  1975 

Leis  and  Rennis,  1983 

Zama  el  al..  1977 

Hardy,  1982 

Dakin  and  Kesteven,  1938 

Lake,  1967 

Jackson,  1978 


Hardy,  1978b 
Sparta,  1939a 
Bertolini,  1933b 

Leis  and  Rennis,  1983 
Burgess,  1978 
Fourmanoir,  1976 
Burgess,  1974 
Hardy,  1978b 
Dekhnik,  1973 
Salekhova,  1959 
Norcross  et  al.,  1974 
Pearson,  1941 
Fahay,  1983 
Silverman,  1975 


JOHNSON:  PERCOIDEI 


495 


Table  122.    Continued. 


Eggs 


Prcficxion 


Postflexion 


Priacanthidae 

Pseudochromidae 

Rachycentridae 

Scatophagidae 

Sciaenidae 
Scorpididae 
Serranidae 
Sillaginidae 

Siniperca 
Sparidae 


Stereolepis 

Symphysanodon 

Terapondiae 


Leis  and  Rennis,  1983 
Suzuki  et  al.,  1980 


Leis  and  Rennis,  1983 
Lubbock.  1975 
Hardy,  1978b 


Leis  and  Rennis, 
Lubbock,  1975 


1983 


Leis  and  Rennis,  1983 
Hardy,  1978b 
D.  K.  Caldwell,  1962 
Aboussouan,  1969 

Leis  and  Rennis,  1983 


—  Weber  and  de  Beaufort,  — 

1936 
Numerous  references,  see  Breder  and  Rosen,  1966;  Hardy,  1978b;  and  Auer,  1982 

-  -  Hattori.  1964 
Kendall,  this  volume 

Breder  and  Rosen,  1966 
Ueno  and  Fujita,  1954 
Uchidaet  al..  1958 


Ueno  and  Fujita,  1954 
Uchidaet  al.,  1958 


Munro,  1945 
Uchidaet  al..  1958 


Imai  and  Nakahara,  1957 
Chyung,  1977 

Breder  and  Rosen,  1966 
Johnson,  1978 
Russell,  1976 
Ranzi,  1933 
Rathbun,  1893 
Cardeilhac,  1976 
Kuntzand  Radcliffe,  1917 
Houde  and  Potthoff,  1976 
Uchidaet  al.,  1958 
Fahay,  1983 
Hussain  et  al.,  1981 


Breder  and  Rosen,  1966 
Llewellyn,  1973 
Zvjagina,  1965b 
Lake,  1967 


Imai  and  Nakahara, 
Chyung,  1977 


1957 


Imai  and  Nakahara, 
Chyung,  1977 


1957 


Johnson,  1978 
Russell,  1976 
Ranzi,  1933 

Kuntzand  Radcliffe,  1917 
Houde  and  Potthoff,  1976 
Uchidaet  al.,  1958 
Fahay,  1983 
Kohnoet  al..  1983 
Hussain  et  al.,  1981 


Llewellyn,  1973 
Uke.  1967 


Johnson,  1978 
Russell.  1976 
Ranzi,  1933 
Hildebrand  and  Cable, 

1930 
Kuntzand  Radcliffe,  1917 
Houde  and  Potthoff,  1976 
Fahay.  1983 
Kohnoet  al..  1983 
Hussam  et  al.,  1981 


Llewellyn,  1973 
Zvjagina,  1965b 
Uke,  1967 


Leis  and  Rennis,  1983 
Hardy,  1978b 
D.  K.  Caldwell,  1962 
Fourmanoir,  1976 
Okiyama,  1982b 
Leis  and  Rennis,  1983 

Hardy,  1978b 

Dawson,  1971a 

Nair,  1952b 

Weber  and  de  Beaufort,  1936 

Hattori,  1964 

Okiyama,  1982b 

Munro,  1945 

Uchidaet  al.,  1958 

Gopinath.  1946 

Okiyama,  1982b 

Imai  and  Nakahara,  1957 

Chyung,  1977 

Johnson,  1978 

Russell,  1976 

Ranzi,  1933 

Hildebrand  and  Cable,  1930 

Kuntz  and  Radcliffe,  1917 

Okiyama,  1982b 

Munro,  1945 

Houde  and  Potthoff,  1976 

Uchidaet  al.,  1958 

Fahay,  1983 

Kohnoet  al.,  1983 

Hussain  et  al.,  1981 

Okiyama,  1982b 

Fourmanoir,  1973 

Llewellyn,  1973 

Nair,  1952b 

Munro.  1945 

Zvjagina.  1965b 

Lake,  1967 

Vatanachi,  1972 


incorrectly  dismissed  the  significance  of  the  larvae,  which,  as 
Okiyama  (1982b)  pointed  out,  are  remarkably  similar  and  dis- 
tinctive among  the  percoids.  I  believe  the  larval  morphology  of 
these  two  groups  offers  conclusive  evidence  for  a  sister-group 
relationship  between  them,  including  a  synapomorphy  unique 
among  percoids,  and  perhaps  all  teleosts. 

Larval  malacanthids  and  branchiostegids  (Fig.  256E,  ¥),  are 
among  the  most  elaborately  ornamented  in  the  Percoidei.  They 
share  early  developing  spinous  scales,  a  series  of  serrate  ridges 
on  the  frontals,  and  have  very  similar  configurations  of  spines 
and  serrate  ridges  on  many  of  the  exposed  bones  of  the  head. 
The  most  distinctive  feature  is  a  median  rostral  bony  structure, 
forming  a  blunt,  serrate-ridged  projection  in  Caulolatilus.  Lo- 
pholattlus  and  Branchiostegus.  a  smooth  anchor-shaped  projec- 
tion in  Malacanthus  and  a  long  spike-like  spine  with  serrate 
ridges  in  Hoplolatilus.  Dooley  (1978)  stated  that  larvae  with 
similar  rostra  and  head  spination  occur  among  holocentrids, 
lutjanids,  serranids  and  istiophorids  and  thai  the  similarity  "could 


be  considered  as  convergence  or  perhaps  a  relict  characteristic 
carried  over  from  a  common  beryciform  ancestor."  In  fact,  the 
larvae  of  these  groups  are  quite  different  morphologically,  and 
misconceptions  about  their  similarity  apparently  result  from 
superficial  considerations  that  have  often  characterized  earlier 
larval  descriptions.  Neither  larval  lutjanids  nor  serranids  have 
rostral  projections  or  (with  the  exception  of  some  anthiin  ser- 
ranids) particularly  elaborate  head  spination.  The  rostral  pro- 
jection of  istiophorids  is  a  premaxillary  beak  or  bill,  supported 
internally  by  a  fixed,  horizontally-oriented  rostral  cartilage  and 
is  structurally  homologous  to  that  of  larval  Xiphias  and  scom- 
brids  (except  Scombrini).  Although  the  spinous  rostrum  of  hol- 
ocentrids bears  a  strong  resemblance  to  that  of  Hoplolatilus.  it 
is  an  entirely  different  structure,  formed  by  enlargement  of  the 
supraethmoid  and  supported  by  a  greatly  enlarged  ethmoid  car- 
tilage. The  median  rostral  projection  of  malacanthids  and  bran- 
chiostegids has  been  described  as  an  ethmoid  spine  (Okiyama, 
1964,  1982b),  but  it  actually  originates  from  a  modification  of 


496 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  263.     Scanning  electromicrographs  of  epithelium  of  juvenile  dolphins  and  cobia  at  various  magnifications.  (A)  Coryphaena  hippurus,  28 
mm  SL,  15  x;  (B)  C.  hippurus.  28  mm  SL,  360 x;  (C)  Rachycentron  canadum.  30  mm  SL,  15  x;  and  (D)  R.  canadum.  80  mm  SL,  360 x. 


the  nasal  bones.  The  nasal  bones  first  appear  as  separate  struc- 
tures, but  prior  to  or  during  flexion,  they  become  fused  anteriorly 
by  a  median  bony  bridge.  This  modified  nasal  structure  then 
develops  the  various  ornamentations  that  characterize  mala- 
canthid  and  branchiostegid  larvae.  At  transformation,  the  bony 
bridge  begins  to  fragment  and  is  eventually  entirely  resorbed, 
so  that  the  nasal  bones  once  again  become  completely  separate. 
I  know  of  no  other  example  in  fishes  of  transient  ontogenetic 
fusion  of  nasal  bones.  This  unique  synapomorphy,  in  conjunc- 
tion with  the  other  shared  larval  specializations,  cogently  sup- 
ports the  hypothesis  that  malacanthids  and  branchiostegids  are 
sister  groups.  Classification  of  the  two  lineages  of  tilefishes  as 
subfamilies  of  the  Malacanthidae  seems  an  appropriate  way  to 
express  this  relationship. 

The  evolutionary  relationships  of  the  dolphins,  Coryphaen- 
idae,  have  remained  uncertain,  but  the  family  has  usually  been 
placed  close  to  the  Carangidae  as  have  the  Echeneididae  and 
the  monotypic  Rachycentridae.  Examination  of  the  larvae  of 
these  groups  during  this  investigation  and  subsequent  consid- 
erations of  adult  morphology  have  led  to  further  resolution  of 
the  interrelationships  of  these  families  (Johnson,  Abstracts  of 
1983  ASIH  Annual  Meeting).  This  final  example  provides  the 


most  convincing  illustration  of  the  importance  of  larval  char- 
acters to  studies  of  phylogeny  among  percoids.  Consequently  I 
discuss  it  in  considerable  detail. 

Freihofer  (1978)  noted  that  the  Nematistiidae,  Carangidae, 
Coryphaenidae,  Rachycentridae  and  Echeneididae  share  a  unique 
specialization  in  the  lateralis  system  on  the  snout— an  anterior 
extension  of  the  nasal  canal  consisting  of  one  (Nematistiidae) 
or  two  prenasal  canal  units,  with  one  (Nematistiidae  and  Ca- 
rangidae) or  both  (remaining  three  families)  surrounded  by  tu- 
bular ossifications.  In  addition,  they  share  small,  adherent  cy- 
cloid scales.  Based  on  two  presumed  synapomorphies,  then, 
these  five  families  constitute  a  monophyletic  group,  hereafter 
referred  to  as  the  carangoids. 

Three  synapomorphies  unite  the  Carangidae,  Coryphaenidae, 
Rachycentridae  and  Echeneididae  as  a  monophyletic  group. 
These  four  families  lack  the  bony  stay  (Potthoff,  1975)  posterior 
to  the  ultimate  dorsal  and  anal  pterygiophores  found  in  almost 
all  other  percoids  (see  Table  1 20),  have  two  prenasal  canal  units 
and  have  a  lamellar  expansion  along  the  anterior  margin  of  the 
coracoid.  Nematisttus,  placed  in  separate  family  by  Rosenblatt 
and  Bell  (1976),  is  apparently  the  sister  group  of  these  four 
families  (see  cladogram.  Fig.  276,  in  Smith-Vaniz,  this  volume). 


JOHNSON:  PERCOIDEI 


497 


Fig.  264.     Scjiiaiiig  clcctromicrographs  of  epillicliuiu  ol  laival  Uolpliin  and  amberjack  at  various  magnifications.  (A)  Coryphaena  hippurus, 
17.0  mm  SL,  55x;(B)  C.  hippurus.  17.0  mm  SL,  400x;(C)  Seriola  sp.,  11.2  mm  SL,  55x;and(D)  5.  sp.,  11.2  mm  SL,  2,000  x. 


It  has  a  well  developed  bony  stay,  a  single,  partly  ossified  pre- 
nasal  canal  unit  and  an  unmodified  coracoid. 

Within  the  carangoids,  the  Coryphaenidae,  Rachycentridae 
and  Echeneididae  form  a  monophyletic  group,  here  referred  to 
as  the  echeneoids.  Adult  echeneoids  are  specialized  with  respect 
to  the  Carangidae  in  the  following  features:  absence  of  predorsal 
bones;  anterior  shift  of  the  first  dorsal  pterygiophore  forward  of 
the  third  intemeural  space;  presence  of  several  anal  pterygio- 
phores  anterior  to  the  first  haemal  spine  (vs.  one  in  carangids 
and  most  other  percoids);  loss  of  the  so-called  beryciform  fo- 
ramen in  the  anterior  ceratohyal;  and  tubular  ossifications  sur- 
rounding both  prenasal  canal  units.  Larval  echeneoids  are  also 
specialized  with  respect  to  carangids  (larvae  of  Ncmatistius  are 
unknown).  Whereas  larval  carangids  are  moderate  to  deep-bod- 
ied, hatch  at  small  sizes  (1-3.5  mm)  and  complete  dorsal  fin 
and  anal  fin  rays  in  conjunction  with  or  soon  after  flexion, 
echeneoid  larvae  (Fig.  261 A-C)  are  very  elongate,  hatch  at  large 
sizes  and  complete  dorsal  fin  rays  at  two  to  three  times  the  size 
at  flexion  (sec  Table  121).  Larval  morphology  thereby  corrob- 
orates the  hypothesized  monophyly  of  the  echeneoids. 


Although  a  sister-group  relationship  between  the  Coryphaen- 
idae and  either  the  Rachycentridae  or  the  Echeneididae  has  not 
been  previously  proposed,  it  has  often  been  suggested  that 
Rachycentron  and  the  echeneidids  are  sister  groups.  This  hy- 
pothesis was  based  on  general  external  similarity  including  the 
remarkable  resemblance  in  body  form,  color  pattern  and  caudal 
fin  shape  between  juveniles  of  Rachycentron  and  Echeneis  nau- 
aa?«(B6hlke  and  Chaplin,  1968).  Because  the  juvenile  features 
of  Rachycentron  are  shared  by  only  one  species  of  echeneidid, 
they  do  not  provide  evidence  for  a  sister-group  relationship 
between  the  Rachycentridae  and  the  Echeneididae,  nor  does  a 
detailed  osteological  comparison  of  the  two  groups.  The  eche- 
neidids are  highly  modified  in  almost  every  aspect  of  their  os- 
teology compared  to  both  Rachycentron  and  Coryphaena,  and 
with  two  exceptions  (absence  of  a  median  cranial  crest  and 
fusion  of  the  prenasal  ossifications),  the  only  specializations 
shared  by  both  Rachycentron  and  the  echeneidids  are  also  shared 
by  Coryphaena.  The  following  are  autapomorphies  of  the  Eche- 
neididae: spinous  dorsal  fin  modified  as  an  attachment  disc 
covenng  the  dorsal  surface  of  the  cranium;  first  neural  arch  fused 


498 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


to  its  centrum,  spine  absent;  endopterygoid  absent;  quadrate 
with  a  lateral  shelf;  palatine  and  upper  jaw  bones  distinctively 
modified;  postcleithra  absent;  supracleithrum  extremely  re- 
duced; medial  tabular  bones  absent;  posttemporal  modified  in 
shape  and  angle  of  articulation  with  supracleithrum;  pelvic  gir- 
dle broad  and  short,  with  two  distinct  anterior  processes;  caudal 
skeleton  with  a  full  neural  spine  on  the  second  preural  centrum; 
branchial  skeleton  with  main  arm  of  first  epibranchial  reduced 
to  a  nubbin,  uncinate  process  enlarged  and  articulating  directly 
with  second  pharyngobranchial,  and  interarcual  cartilage  absent. 
None  of  these  extreme  modifications  (those  of  the  caudal  and 
branchial  skeletons  being  unique  among  percoids)  are  even  fore- 
shadowed in  the  skeleton  of  Rachycentron,  which  is  instead 
remarkably  similar  to  that  oi  Coryphaena.  except  in  the  anterior 
portion  of  the  dorsal  fin  and  the  neurocranium. 

In  Coryphaena,  the  dorsal  fin  is  elaborated  anteriorly  and 
extended  into  the  first  intemeural  space  (second  in  Rachycen- 
tron) and  there  is  an  extreme  supraoccipito-frontal  crest  on  the 
neurocranium.  The  dorsal  fin  modification  is  autapomorphic 
for  Coryphaena.  but  the  median  cranial  crest  is  probably  prim- 
itive for  echeneoids  since  it  is  variously  developed  in  all  caran- 
gids  and  well-developed  in  Nematislius.  The  absence  of  this 
crest  in  Rachycentron,  associated  with  a  slight  flattening  of  the 
neurocranium,  is  the  only  specialization  shared  with  the  Eche- 
neididae.  Here  again,  however,  there  is  little  similarity  between 
the  slightly  flattened  neurocranium  of  Rachycentron  and  the 
extremely  flattened  and  restructured  neurocranium  of  the  eche- 
neidids,  in  which,  for  instance,  the  supraethmoid  and  vomer 
have  become  flat  plates  and  the  orbit  is  completely  occluded  by 
enlargement  and  anterior  extension  of  the  pterosphenoids.  This 
extreme  restructuring  of  most  cranial  bones  is  evident  even  in 
larval  echeneidids  at  the  earliest  development  of  the  neuro- 
cranium, whereas  the  neurocrania  of  Rachycentron  and  Cory- 
phaena exhibit  a  generalized  development  similar  to  that  of 
carangids.  Prior  to  development  of  the  median  crest  in  Cory- 
phaena(>  100  mm),  the  neurocrania  of  cobia  and  dolphin  differ 
mainly  in  relative  depth.  Echeneidids  also  have  an  exceptionally 
modified  adductor  mandibulae  in  which  A,  is  absent  and  A, 
and  A„  are  distinctively  subdivided.  Coryphaena  and  Rachy- 
centron share  a  relatively  generalized  adductor  mandibulae,  spe- 
cialized with  respect  to  the  primitive  carangids  (see  section  on 
Carangidae)  in  having  A,  somewhat  reduced  and  inserting  nar- 
rowly on  the  maxillo-mandibular  ligament. 

The  pronounced  similarities  between  Coryphaena  and  Rachy- 
centron in  the  adductor  manidbulae  and  most  osteological  fea- 
tures merely  serve  to  reiterate  the  lack  of  evidence  for  the  fre- 
quently proposed  sister-group  relationship  between  Rachycentron 
and  the  echeneidids.  Further  comparison  with  character  states 
throughout  the  Carangidae  will  be  required  to  define  these  adult 
similarities  as  primitive  or  derived  features.  The  most  com- 
pelling evidence  for  a  sister-group  relationship  between  Cory- 
phaena and  Rachycentron  is  found  in  the  morphology  of  their 
larvae.  As  noted  above  all  echeneoid  larvae  have  a  similar  body 
form  and  pattern  of  development,  but  the  elongate,  flattened 
head  of  larval  echeneidids  lacks  ornamentation.  In  contrast, 
larval  dolphin  and  cobia  share  identical  patterns  of  head  spi- 
nation:  a  small  posttemporal  spine;  several  spines  on  the  pos- 
terior and  lateral  margin  of  the  preopercle,  including  one  en- 
larged spine  on  either  side  of  its  angle;  and  a  very  large, 
posterolaterally  directed  spine  on  the  supraorbital  ridge  of  each 
frontal  bone.  Another  obvious  feature  is  the  presence  of  laterally 
swollen  pterotics,  previously  described  in  Coryphaena  as  blunt 


sphenotic  spines  (Gibbs  and  Collette,  1959).  This  specific  pat- 
tern of  head  spines  is  distinctive,  but  similar  features  occur  in 
various  combinations  among  carangid  larvae,  and  it  is  pre- 
mature to  interpret  this  configuration  as  synapomorphic  for 
Coryphaena  and  Rachycentron  until  detailed  comparisons  with 
carangids  have  been  made. 

A  specialization  clearly  unique  to  the  larvae  of  dolphin  and 
cobia,  however,  is  a  modified  epithelial  cuticle  in  which  are 
borne  minute  crown-shaped  spicules  (Figs.  263A-D,  264A,  B). 
The  cuticle  itself  is  composed  of  large,  multinucleate  "cells," 
40-100  m  in  diameter,  that  appear  to  continually  produce  and 
slough-off"  the  thorny  spicules.  Each  epithelial  "cell"  produces 
one  spicule,  so  that  these  extraordinary  structures  cover  all  ex- 
posed body  surfaces,  excluding  the  pupil  of  the  eye,  giving  the 
integument  a  bristly  appearance  under  magnification  (Fig.  264A). 
They  first  appear  at  about  8mm  and  are  present  in  some  indi- 
viduals as  large  as  100  mm.  Further  histological  work  and  elec- 
tron microscopy  will  be  necessary  to  determine  the  composition 
of  the  spicules,  which  may  be  keratinous.  It  is  clear,  however, 
that  they  are  neither  bony  nor  cartilaginous.  Their  function  is 
unknown,  but  as  with  spinous  scales,  it  seems  likely  that  they 
are  defensive. 

The  surface  and  cellular  composition  of  the  epithelium  of 
larval  echeneidids  appear  normal,  but  some  modification  of  the 
larval  epithelium  may  actually  be  a  primitive  feature  of  car- 
angoids.  In  larvae  of  trachinotine  and  naucratine  carangids  ex- 
amined thus  far  (Trachinotus,  Naucrates,  Seriola)  the  epithelial 
cells  are  of  normal  size  ( ~  8- 1 2  m),  but  their  surfaces  bear  clusters 
of  bumplike  structures,  seemingly  the  result  of  keratinization 
(Fig.  264C,  D).  Absence  of  these  modified  epithelial  cells  in 
larvae  of  carangine  carangids  is  parsimoniously  interpreted  as 
secondary  (see  Laroche  et  al.,  this  volume).  Their  presence  in 
the  larvae  of  Neinalistius  (curtently  unknown)  would  corrobo- 
rate the  hypothesis  that  modified  larval  epithelium  is  primitive 
for  carangoids  and  thus  also  for  echeneioids,  suggesting  that  it 
has  been  lost  in  carangines  and  echeneidids. 

The  multinucleate  epithelial  cells  and  enlarged,  thorny  spic- 
ules of  larval  Coryphaena  and  Rachycentron  represent  a  com- 
plex, shared  specialization,  unique  among  percoids.  The  phy- 
logenetic  significance  of  this  synapomorphy  is  lessened  only  by 
the  unlikely  possibility  that  loss  of  a  modified  epithelium  in 
echeneidids  occurted  after  development  of  multinucleate  cells 
and  spicules.  Available  evidence  strongly  points  to  a  Cory- 
phaena-Rachycentron  sister-group  relationship,  and  it  should 
be  clear  that  further  investigations  testing  this  hypothesis  must 
integrate  larval,  adult  and  developmental  characters. 

In  conclusion,  the  study  of  early  life  history  stages  of  fishes 
has  traditionally  been  treated  as  a  discipline  somewhat  removed 
from  the  mainstream  of  systematic  ichthyology.  As  a  result, 
larval  morphology  has  rarely  beeen  incorporated  into  studies  of 
evolutionary  relationships  of  fishes.  It  is  evident  that  the  larvae 
of  percoid  fishes  exhibit  a  prodigious  array  of  complexity  and 
diversity  that  offers  exceptional  potential  applicability  to  phy- 
logenetic  studies.  Recognition  and  application  of  this  potential 
will  be  an  important  step  in  understanding  the  complex  evo- 
lutionary history  of  the  Percoidei. 

South  Carolina  Wildlife  and  Marine  Resources  De- 
partment, Post  Office  Box  12559,  Charleston,  South 
Carolina  29412.  Present  Address:  Fish  Division,  Na- 
tional Museum  of  Natural  History,  Washington, 
District  of  Columbia  20560. 


Serranidae:  Development  and  Relationships 
A.  W.  Kendall,  jr. 


THE  percoid  family  Serranidae  is  defined  by  the  presence  of 
three  spines  on  the  opercle  (Goshne,  1966)  and  three  re- 
ductive specializations  (absence  of  the  posterior  uroneural,  pro- 
current  spur,  and  third  preural  radial  cartilage)  that  separate  it 
from  the  Percichthyidae  (Johnson,  1983).  These  are  primarily 
tropical  to  temperate  marine  fishes  that  vary  in  size  from  <  10 
cm  to  >300  cm.  It  is  a  speciose  family  with  nearly  400  species 
(Nelson,  1976)  that  has  had  a  history  of  being  hard  to  charac- 
terize and  subdivide.  The  serranids  are  continuing  objects  of 
taxonomic  studies  from  the  species  to  subfamily  levels  and  sev- 
eral new  species  are  described  each  year,  primarily  anthiines 
whose  deep-water  reef  habitat  has  made  collecting  difficult.  As 
presently  understood  (Johnson,  1983).  the  family  is  composed 
of  3  subfamilies  (Serraninae.  Anthiinae.  and  Epinephelinae), 
although  Katayama  (1960)  recognized  15  subfamilies.  Various 
authors  have  included  other  groups  (e.g..  Callanthias)  in  the 
Serranidae,  and  others  have  raised  parts  of  the  family  to  familial 
status  (e.g.,  Anthiinae  and  Grammistinae).  Such  problems  will 
probably  not  be  resolved  without  a  worldwide  revision  of  the 
family,  which  is  not  forthcoming. 

Development 

The  eggs  of  all  but  a  few  serranids  are  unknown.  Often,  Wil- 
son's (1891)  classic  work  on  the  development  of  Centropnsiis 
striata  eggs  has  been  cited  as  the  example  of  teleost  embrjology 
in  texts  (e.g.,  Nelsen,  1953).  Serranid  eggs  described  to  date  are 
typical  of  the  majority  of  pelagic  marine  teleost  eggs:  they  are 
spherical,  about  1  mm  in  diameter,  have  a  single  oil  globule,  a 
narrow  perivitelline  space,  and  a  smooth  egg  envelope.  Several 
species  of  Epincphelus  (e.g.,  Guitart  Manday  and  Juarez  Fer- 
nandez, 1966:  Hussain  and  Higuchi,  1980),  fara/aira.v  (Butler 
et  al.,  1982),  and  several  anthiines  (e.g..  Suzuki  et  al.,  1974, 
1978)  have  been  reared.  There  seems  to  be  a  difference  in  oil 
globule  placement  in  yolk-sac  larvae  among  the  subfamilies  (Fig. 
265).  Larvae  of  representatives  of  all  the  subfamilies,  most  of 
the  tribes,  and  about  a  third  of  the  genera  of  serranids  have 
been  described.  Serranid  larvae  fall  into  one  of  four  types,  which 
correspond  to  two  of  the  subfamilies  and  two  of  the  tribes  within 
the  Epinephelinae.  These  larval  types  can  be  characterized  based 
on  the  taxa  for  which  larvae  are  known  as  follows  (based  on 
Kendall,  1979). 

Serraninae.  — hody  proportions  show  rather  direct  develop- 
ment. There  are  no  elongate  spines  in  the  opercular  region, 
rather  a  series  of  blunt  points.  The  fin  spines  are  thin  and  only 
slightly  elongated  in  some.  Most  larval  pigment  consists  of  me- 
lanophores  in  characteristic  positions  along  the  ventral  midline. 

Anthiinae.— These  deep-bodied  larvae  have  produced  spines  on 
several  bones  in  the  opercular  region,  some  of  which  may  be 
serrated.  There  is  a  tendency  to  develop  armature  on  the  head, 
and  the  interopercular  has  a  characteristic  long  posteriorly  di- 
rected spine  that  is  overlaid  by  an  even  larger,  similar  spine  on 
the  preopercular.  The  pelvic  and  some  dorsal  fin  spines  are 
strong,  serrate  in  some,  and  not  very  elongate.  Pigment  consists 


mainly  of  large  blotches  and  dashes  in  characteristic  positions 
on  the  trunk. 

Epinephelini.  —  Knovfn  larvae  of  members  of  this  tribe  are  all 
quite  similar  and  generally  difficult  to  assign  to  a  genus  on  the 
basis  of  larval  characters.  These  are  among  the  most  spectacular 
offish  larvae,  with  stout,  elongate,  serrate,  and  pigmented  dorsal 
and  pelvic  fin  spines.  Usually  the  second  dorsal  spine  is  much 
longer  than  the  others  and  it,  as  well  as  the  pelvic  spines,  are 
as  long  as  the  body.  The  dorsal  spine  is  often  "locked"  in  an 
upright  position— presumably  possible  because  of  a  unique  pte- 


Fig.  265.  Newly  hatched  yolk-sac  larvae  of  serranids:  (A)  Serraninae: 
Paralabrax  clathratus.  from  Butler  et  al.  (1982);  (B)  Anthiinae:  Sacura 
marganlacea.  from  Suzuki  et  al.  (1974);  and  (C)  Epinephelinae:  Epi- 
nephelus  akaara.  from  Ukawa  et  al.  (1966). 


499 


500 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  123.    Serranid  Taxa  (Subfamilies  through  Subgenera).  Their  general  distribution  and  references  to  early  life  history  (ELH)  descriptions. 

A-A:  Atlanto-American.  I-P;  Indopacific.  Stages  described:  E  (eggs),  Y  (yolk-sac  larvae),  L  (larvae  — yolk-sac  through  post  flexion),  Pr  (preflexion 

larvae),  F  (flexion  larvae),  Po  (postflexion  larvae),  T  (transforming  larvae)  and  J  (juveniles). 


Subfamily 

Genus 
subgenus 

Distn 

bution 

Iribe 

A-a 

IP 

Serraninae 

Acanlhislius 

+ 

Cenlropnstis 

+ 

Chelidoperca 

+ 

Crattnus 

+ 

Diplectrum 

-1- 

Dules 

-1- 

Hypoplectrus 

+ 

Paratabrax 

+ 

Schultzea 

+ 

Senamculus 

-1- 

Serramis 

+ 

Serranus 

Paracenlropnstis 

Anthiinae 

Anthias 
Microlahrichthys 
Nemanthias 
Pseudanlhias 

+ 

+ 

Caesioperca 

+ 

Caprodon 

-1- 

-1- 

Daciylanthias 

+ 

Ellerketdia 

-t- 

Franzia 

+ 

Giganthias 

+ 

Hemanlhias 

+ 

Holanthias 

+ 

+ 

Luzonichlhys 

+ 

Ocyanthias 

+ 

Odontanlhias 

+ 

Pleclranthias^ 

-1- 

+ 

Pronotogrammus 

+ 

Sacura 

+ 

Selenanthias 

+ 

Senanocirrhitus 

+ 

Tosana 

+ 

Tosanoides 

+ 

Epinephelinae'' 

Niphonini 

Niphon 

+ 

Epinephelini 

A  nyperodon 

+ 

Cromileplcs 

+ 

Epmephelus^ 

-1- 

+ 

ELH  descnplions 


Alphestes 


Cephalopholis 

Dermalolepis 

Epinephelus 

Promicrops 

Gonioplectrus 

+ 

Gracilia 

+ 

Mvcleroperccf 

+ 

Paranthias 

+ 

+ 

Pleclropomus 

-1- 

Trisolropis 

+ 

Variola 

+ 

Diploprionini 

Aulacocephalus 

+ 

Belonoperca 

+ 

Diploprion 

+ 

Liopropomini 

Jehoehlkia 

-1- 

Liopropoma 

+ 

+ 

Ryder  (1888)-E-Y,  Wilson  (189I)-E-Y,  Hofr(1970)-E-Y,  Kendall  (1972)-Pr-J. 
Kendall  (1977,  l979)-Pr-Po 


Kendall  (1977,  l979)-Pr-Po 

Kendall  (1977,  l979)-Pr-Po 

Kendall  (1977,  1979)-Pr-Po,  Butler  et  al.  (1982)-E-J 

Kendall  (1977,  l979)-Pr-Po 

Rafl!"aele  (1888)-E,Y,  Page  (1918)-Pr-J,  Roule  and  Angel  (I930)-L,  Bertolini 

(1933b)-E-J,  Vodyanitsky  and  Kazanova  (1954)-E-Po,  Aboussouan  (1972b)-Pr- 

Po,  Kendall  (1977,  1979)-Pr-Po 


Roule  and  Angel  (I930)-Pr-T,  Sparta  (1932)- Pr-T,  Bertolini  (1933b)-Pr-J,  Abous- 
souan (1972b)-Pr-Po,  Fourmanoir  (l976)-Po,  Kendall  (1977,  1979)-Pr-Po,  Su- 
zuki et  al.  (in  press),  Leis  and  Rennis  (1983)— Pr-T 

Fourmanoir  (1976)— Po 


Suzuki  el  al.  (1978)-E-Y 
Kendall  (1977,  1979)-Pr-Po 
Fourmanoir  (1976)  — T 


Kendall  (1977,  1979)-Po 

Kendall  (1977,  l979)-Pr-Po 

Suzuki  et  al.  (I974)-E-Pr,  Fourmanoir  (1976)- Po 

Fourmanoir  (1973)  — T 


Raffaele  (I888)-E,  Page  (19l8)-Po,  Bertolini  (1933b)-F,  Sparta  (1935)-E-T,  Fow- 
ler (1944)— T,  Vodyanitsky  and  Kazanova  (1954)  — E-Po,  Guitart  Manday  and 
Juarez  Fernandez  (1966)-E-Y,  Ukawa  et  al.  (1966)-E-Pr,  Mito  et  al.  (1967)-Pr-J 

Presley  (1970)-F-Po,  Smith  (1971)-Po,  Aboussouan  (1972b)-Pr-Po,  Fourmanoir 
(1976)-L,  Chenet  al.  (1977)-E,  Kendall  (1977.  l979)-Po,  Hussain  and  Higuchi 
(I980)-Y-J 

Johnson  and  Ashe  (1984)- Po-J,  Leis  and  Rennis  (1983)-Pr-Po 


Kendall  and  Fahay  (1979)-Po,  Johnson  and  Ashe  (l984)-Po 

Kendall  (1977,  1979)-Pr-J,  Johnson  and  Ashe  (1984)- Po-T 
Kendall  (1977,  l979)-Po,  Johnson  and  Ashe  (1 984)- Po-T 


HubbsandChu  (1934)-T 

Kotthaus  (1970)-L,  Fourmanoir  (197 la)- Po,  Fourmanoir  (1976)-Po,  Kendall 
(1977.  1979)-Pr-Po 


KENDALL:  SERRANIDAE 


501 


Table  123.    Continued. 


Subfamily 

Genus 
subgenus 

Distnbution 

inbc 

A-A         1-P 

Pikea 

+ 

Rainfordia 

+ 

Grammistini 

Aporops 

+ 

Grammistes 

+ 

Grammistops 

+ 

Pogonoperca 

+ 

Pseudogramma 

+           + 

Rypticus 

+            + 

Suttonia 

+ 

ELH  descnptions 


Fourmanoir  (1976)— Po 
Fourmanoir  (1976)— T 

Kendall  (1977,  I979)-Pr-Po.  U-is  and  Rennis  (1983)-Pr-Po 
Aboussouan  (1972b)-Po,  Kendall  (1977,  1979)-Po 


'  Randall  (1980)  includes  in  Plectranthias:  Sayanura,  hobuna.  Xenanlhias.  Pleranthias,  Zatanthias.  Serranops.  Peionlnis.  and  Zacallanlhias 
*•  Subdivisions  follow  Johnson  (1983). 

'  Tortonese  (1973)  states  thai  Bertolini  (1933b)  and  Sparta  (1935)  described  Mycteroperca  ruba  larvae  as  Ephmephetus  alexandnnus  and  that  this  mistake  has  been  continued  in  more  recent 
literature 


rygiophore  arrangement  (Johnson,  1983).  The  first  and  third 
dorsal  spines  and  the  anal  spines  are  also  stout  and  may  bear 
serrations.  The  spine  at  the  angle  of  the  preopercular  is  elongate 
and  serrate;  there  are  two  smaller  spmes  dorsal  and  ventral  to 
the  one  at  the  angle,  and  these  may  also  bear  serrations.  There 
is  a  serrate  spine  on  the  supracleithrum.  The  body  is  "kite- 
shaped";  pigment  lines  the  body  cavity  and  there  is  a  large, 
conspicuous  spot  on  the  caudal  peduncle  that  migrates  from  the 
ventral  midline  to  a  midlateral  position  during  flexion. 

Grammistini-Liopropomini.  —  The  body  is  roughly  tubular  with 
a  deep  caudal  peduncle.  Among  the  bones  in  the  opercular  series 
the  preopercular  is  armed  with  about  five  elongated,  simple 
spines.  One  or  two  dorsal  fin  spines  become  quite  elongate,  and 
are  thin  and  flexible  with  pigmented  membranous  sheaths  around 
them.  Bodies  of  the  larvae  are  practically  devoid  of  pigment 
throughout  development. 

The  following  is  a  summary  of  the  current  status  of  the  sys- 
tematics  and  knowledge  of  larval  morphology  of  each  of  the 
subfamilies  of  serranids  (Table  123). 

Serraninae 

There  has  been  no  revision  of  this  primarily  Atlanto-Amer- 
ican  subfamily,  and  little  work  on  relationships  among  species 
in  the  various  genera  (Bortone,  1977).  These  are  considered  the 
least  specialized  of  the  serranids  and  are  riiainly  united  by  shared 
possession  of  basal  percoid  characters  rather  than  unique  spe- 
cializations, which  would  allow  a  definitive  statement  about 
monophyly.  They  possess  the  four  serranid  specializations  as 
mentioned  by  Johnson  (1983),  are  hermaphroditic  or  second- 
arily gonochoristic  (see  Kendall,  1977).  have  a  common  pre- 
dorsal  bone  pattern  (0/0/0/2),  and  a  fairly  coherent  larval  mor- 
phology (Fig.  266). 

The  larvae  of  Schulcea.  Dules.  Acanihistius,  and  Crulmus 
are  unknown.  The  following  summary  of  what  is  known  of  the 
larval  morphology  of  the  rest  of  the  serranines  is  based  primarily 
on  Kendall  (1977,  1979).  The  only  more  recent  contributions 
to  serranine  larval  knowledge  are  the  descriptions  of  Paralabra.x 
(Butler  et  al.,  1982). 

Centropnstis.— Only  one  larval  type  is  known,  although  four 
species  are  named.  The  eggs  and  yolk-sac  larvae  have  been 
described  from  reared  specimens.  Development  is  typical  of 
serranines  with  small  simple  spines  on  the  preopercular.  The 
first  and  second  dorsal  fins  develop  at  about  the  same  rate;  there 


are  no  elongate  or  armed  fin  spines.  Most  pigment  is  in  blotches 
in  characteristic  serranine  positions.  The  body  gradually  as- 
sumes the  adult  shape. 

Paralabra.x— Biil\eT  et  al.  ( 1 982)  reared  from  eggs  and  described 
development  of  the  three  species  found  off  California.  These 
larvae  vary  from  the  general  serranine  pattern  of  development, 
primarily  in  having  pigmented  membranes  of  the  pectoral,  pel- 
vic, first  dorsal,  and  anal  fins  variously  developed  among  the 
species.  Pigment  is  also  variously  present  on  the  body  ventral 
to  the  first  dorsal  fin. 

Serraniculus.  —  Larvae  of  the  only  species  (S.  pumilio)  are  deeper 
bodied  and  have  more  lateral  pigment  than  other  serranines. 
The  flank  pigment  is  composed  of  three  series  of  dashes  (one 
along  the  midlateral  septum  and  one  along  the  base  of  the  dorsal 
and  anal  fins)  and  superficial  small  spots  over  much  of  the  trunk. 
The  ventral  midline  spots  are  small  and  rather  uniform  in  size. 
The  first  dorsal  fin  develops  concurrently  with  the  second  dorsal, 
and  the  spines  are  no  longer  than  the  rays. 

Diplectrum— Two  distinct  types  of  larvae  with  the  meristic 
characters  of  this  genus  were  found  in  both  Atlantic  and  Pacific 
material.  One  type  (Type  1 )  closely  follows  the  serranine  pattern 
of  development,  the  main  difference  being  in  the  early  devel- 
opment of  the  spinous  dorsal  and  pelvic  fins.  The  ventral  mid- 
line pigment  spots  seem  more  uniform  in  size  than  in  other 
genera,  and  there  is  pigment  on  some  of  the  fin  membranes. 
Larvae  of  the  other  type  (Type  2)  are  quite  different  from  other 
serranine  larvae  in  that  the  pectoral  and  pelvic  fins  develop 
early  and  are  enlarged  and  pigmented  on  their  distal  thirds.  The 
body  is  practically  devoid  of  pigment  except  for  two  spots  on 
the  caudal  peduncle— one  dorsal  and  one  ventral,  and  an  inter- 
nal diffuse  area  of  pigment  lateral  to  the  anterior  part  of  the 
anal  fin  that  develops  after  the  fin  rays  are  formed. 

Serranus.  —  Larvae  of  this  genus  from  both  sides  of  the  Atlantic 
have  been  described,  and  reared  eggs  and  yolk-sac  larvae  were 
among  the  first  serranids  descnbed  (Raffaele,  1888).  These  lar- 
vae differ  from  the  serranid  pattern  of  development  in  having 
eariy-forming  elongate  dorsal  spines  and  a  deeper  body.  In  5. 
cabrilla  the  pelvic  spines  develop  before  any  other  fin  rays  and 
they  and  the  third  dorsal  spine  become  quite  elongate.  Some  of 
the  smaller  ventral  melanophores  seen  in  other  serranines  are 
absent  from  Serranus  larvae,  while  some  of  the  larger  spots  are 


502 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


B 


KENDALL:  SERRANIDAE 


503 


more  intense.  Pigment  develops  variously  at  the  base  of  the 
dorsal  fin  and  in  the  membranes  of  the  first  dorsal,  pelvic,  and 
anal  fins.  S.  cabrilla  has  large  opposing  spots  on  the  caudal 
peduncle. 

Hypoplectrus— Reared  larvae  of  this  genus  are  quite  different 
from  other  serranines.  The  first  dorsal  and  pelvic  fins  develop 
early  and  are  heavily  pigmented.  The  head  and  fin  membranes 
are  fleshier  than  in  other  serranines,  and  these  larvae  do  not 
possess  the  characteristic  ventral  pigment  pattern.  Rather,  there 
are  a  few  spots  ventral  to  the  base  of  the  first  dorsal  fin,  and  a 
few  blotches  ventrally  at  the  base  of  the  pelvic  fins,  at  the  anus, 
along  the  base  of  the  anal  fin,  on  the  caudal  peduncle,  and  at 
the  base  of  the  middle  of  the  caudal  fin. 

Anthiinae 

This  is  a  cohesive  group  of  fishes  that  share  several  special- 
izations in  addition  to  those  they  hold  in  common  with  other 
serranids.  These  specializations  include  large  scales,  a  highly 
arched  lateral  line,  deep  bodies  and  large  heads,  mainly  10  + 
1 6  vertebrae,  and  a  predorsal  pattern  of  0/00/2  or  0/0/2.  They 
are  generally  small,  brightly  colored  reef  fishes.  The  generic 
alignments  of  many  species  are  dubious,  and  a  revision  of  the 
group  is  badly  needed.  Most  recent  work,  however,  has  focused 
on  describing  new  species,  faunal  studies,  and  some  generic 
revisions. 

Recent  and  ongoing  work  (Fitch,  1982;  Baldwin,  pers.  comm.) 
has  brought  out  several  incongruencies  in  generic  assignments 
of  Kendall  (1977,  1979).  In  the  following  summary  of  what  is 
known  of  anthiine  larval  morphology,  generic  larval  types  will 
be  described,  with  the  understanding  that  some  of  the  variation 
within  these  may  be  due  to  species  that  are  assigned  to  the  genus 
incorrectly.  Alternate  generic  placements  of  species  will  be  noted 
as  appropriate  (Table  1 24).  Better  definitions  of  the  genera  must 
await  a  worldwide  revision  that  will  include  information  on 
early  life  history  stages.  Larvae  of  10  of  the  19  currently  rec- 
ognized anthiine  genera  are  known  to  some  extent  (Fig.  267  and 
Table  123). 

Plectranthias  (Fig.  267a).  — Randall  (1980)  included  eight  nom- 
inal genera  in  this  genus,  but  the  monophyly  of  the  included 
species  is  not  resolved  (W.  D.  Anderson,  Jr.,  pers.  comm.,  Jan. 
1983).  Kendall  (1977,  1979)  described  larvae  of  the  American 
species  (P.  garupellus)  as  having  an  elongate  third  dorsal  spine, 
opposing  caudal  peduncle  pigment  blotches  as  well  as  a  blotch 
below  the  center  of  the  first  dorsal  fin,  and  no  serrated  head  or 
fin  armature  (rather  the  characteristic  anthiine  spines  are  thin 
and  weakly  developed).  The  larvae  showed  the  least  develop- 
ment of  anthiine  larval  characters  among  American  genera. 

Anthias  (¥\%.  267c).  — (includes  Pronotogrammus  multifasciatus 
(see  Fitch,  1982))  This  is  a  speciose  circumtropical  genus  that 
has  provisionally  been  subdivided  into  three  subgenera  (Randall 
and  Lubbock,  198  1).  Larvae  of  several  species  from  around  the 
world  have  been  described.  They  share  a  number  of  larval  char- 


Table  124.  ReassionmentofSome  Anthiine  Larvae.  Those  of  Ken- 
dall (1977,  1979)  reassigned  by  Baldwin  (pers.  comm.)  and  Kendall, 
based  on  work  on  adults  from  the  eastern  Pacific  by  Fitch  (1982)  and 
from  the  western  Atlantic  by  Anderson  and  Heemstra  (1980)  and  W. 
D.  Anderson  (pers.  comm..  unpublished  data).  Letters  after  most  likely 
species  names  refer  to  Baldwin  (B)  and  Kendall  (K)  who  recognized 
these  reassignments. 


Kendall.  1977,  1979 


Most  likely  species 


Figure 


Pronotogrammus 

aureoruhens 
Pronologrammus  eos 
Anihias  gordensis 

Amhias  sp.  Type  2 
Hemanthias  peruanus 


Hemanthias  leptus—B  267f 

Hemanthias  signifer—  B 
Pronotgrammiis  multi- 

fascialus—K. 
Holanthias  martinicensis—B  267d 

Pronotogrammus  eos—  K 


acters,  but  there  are  some  notable  differences  among  the  species. 
The  second  or  third  dorsal  spine  is  elongate  and  thin  (the  first 
may  be  late  forming,  so  the  elongate  spine  may  always  be  the 
third);  the  first  few  dorsal  spines  and  the  pelvic  spine  are  early 
forming;  the  elongate  dorsal  spine  has  a  pigmented  sheath;  the 
preopercular  and  interopercular  have  long  serrate  spines;  and 
there  are  generally  two  pigment  spots  ventrally  on  the  caudal 
peduncle.  There  is  a  simple  supraoccipital  spine  in  some  species 
and  a  variable  number  of  spines  on  a  ridge  above  the  eye. 
Pigment,  in  addition  to  that  mentioned  above,  varies  among 
the  species  and  some  species  become  fully  scaled  during  the 
larval  stage.  Whether  these  diflTerences  in  larval  characters  can 
be  related  to  the  subgeneric  alignment  of  species  must  await 
further  larval  descriptions.  Fitch  (1982)  synonymized  the  Pacific 
Anthias  (A.  gordensis),  whose  larvae  Kendall  (1977,  1979)  de- 
scribed, with  Pronotogrammus  multifasciatus. 

Franzia.  — Eggs  and  yolk-sac  larvae  of  F.  squamipmnis  have 
been  described  (Suzuki  et  al.,  1978)  but  later  larval  stages  are 
unknown. 

Caesioperca.  —  Vo\xrmano\r(\916)  illustrated  the  head  and  brief- 
ly described  a  transforming  specimen  thought  to  belong  to  this 
genus.  It  has  a  smooth  supraoccipital  region  and  no  spiny  ridge 
above  the  eye,  but  has  simple  stout  spines  in  the  characteristic 
position  on  the  preopercular  and  interopercular.  The  informa- 
tion presented  is  too  brief  for  further  evaluation  of  anthiine 
larval  characters. 

Luzonichthys.  —  Fourmanoir  (1976)  illustrated  the  anterior  por- 
tion and  briefly  described  two  transforming  specimens  of  this 
genus.  These  have  probably  lost  some  of  their  larval  characters, 
since  the  mouth  is  already  subterminal  and  the  body  covered 
with  scales.  The  spines  on  the  preopercular  are  not  especially 
elongate,  but  one  on  the  interopercular  is  pronounced,  simple, 
and  stout.  Anterior  dorsal  fin  spines  appear  thin  and  not  pro- 
duced. 


Fig.  266.  Examples  of  serranine  larvae:  (A)  Centropristis  striata,  8.3  mm,  from  Kendall  ( 1 979);  (B)  Paralabrax  clathratus.  7.4  mm,  from  Butler 
et  al.  (1982);  (C)  Serranicutus  pumilio.  5.8  mm,  from  Kendall  (1979);  (D)  Diplectrum  sp.,  6.1  mm,  from  Kendall  (1979);  and  (E)  Serranus  sp., 
5.5  mm,  from  Kendall  (1979). 


Fig.  267.  Examples  of  anthiine  larvae.  (A)  Plearanthias  garupelhis.  5.5  mm,  from  Kendall  (1979);  (B)  Pronotogramumus  aureoruhens,  9.8 
mm,  original  illustration;  (C)  Anihias  sp.,  5.3  mm.  from  Kendall  (1979);  (D)  Holanthias  manmicensis.  8.4  mm,  from  Kendall  (1979),  labelled 
Anlhias  sp.  Type  2;  (E)  Hemanthias  vivanus.  6.8  mm,  from  Kendall  (1979);  and  (F)  Hcmanthias  leptus.  6.0  mm,  from  Kendall  (1979)  labelled 
Pronotogrammus  aureoruhens. 


506 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


KENDALL:  SERRANIDAE 


507 


5acMra.  —  Reared  eggs  and  yolk-sac  larvae  were  described  by 
Suzuki  et  al.  ( 1 974)  and  a  postflexion  larva,  illustrated  and  brief- 
ly described  by  Fourmanoir(1976).  shows  characters  of  anthiine 
larval  development.  The  latter  specimen  has  the  third  dorsal 
and  pelvic  spines  extremely  elongate  and  with  a  pigmented 
sheath;  the  opercular  and  interopercular  are  armed  with  stout 
serrate  spines,  and  there  is  a  similar  more  ventrally-directed 
spine  anterior  to  these;  the  anal  spines  are  stout  and  serrate; 
there  is  a  serrate  ridge  above  the  eye,  and  a  midlateral  pigment 
dash  on  the  caudal  peduncle. 

Pronotogrammus-Hemanthias.— As  presently  understood,  two 
species  assigned  to  each  of  these  genera  occur  in  the  eastern 
Pacific  (P.  COS.  P.  multifasciatus,  H.  signifer,  and  H.  peruanus, 
see  Fitch  ( 1 982)),  and  there  are  two  Hemanthias  and  one  Prono- 
togrammus  in  the  western  Atlantic  (H.  leptus.  H.  vivanus,  and 
P.  aiircoruhens). 

Kendall  (1977,  1979)  assigned  larval  types  from  both  oceans 
to  these  genera.  More  recently,  Baldwin  (pers.  comm.)  has  es- 
tablished alternate  generic  assignments  for  some  of  Kendall's 
(1977,  1979)  types  based  on  more  complete  meristic  data  and 
has  assigned  a  previously  undescribed  type  to  Pronotogrammus 
aureorubens.  Thus,  present  generic  assignments,  do  not  coincide 
with  the  larval  types  described  by  Kendall  (1977,  1979).  In  the 
following,  the  morphology  of  the  larval  types  of  Kendall  (1977, 
1979)  will  be  summarized  under  the  species  whose  larvae  are 
represented  by  these  types. 

Hemanthias  signifer,  Hemanthias  leptus  (Kendall's  Pronoto- 
grammus eos  and  P.  aureorubens)  (see  Fig.  2670.  — These  larvae 
are  characterized  by  serrate,  spiny  armature  in  the  opercular 
region,  supraoccipital  crest  simple  or  absent,  first  spines  of  the 
dorsal  fin  and  the  pelvic  fin  early  developing  but  not  becoming 
elongate  or  serrate,  and  midlateral  trunk  pigment. 

Pronotogrammus  eos.  Hemanthias  vivanus  (Kendall's  Heman- 
thias peruanus  and  H.  vivanus)  (see  Fig.  267e).— These  larvae 
develop  a  complex  "cockscomb"  ridge  on  the  supraoccipital,  a 
serrate  ridge  above  the  eye,  some  serrate  spines  on  the  pre- 
opercular  and  interopercular,  some  serrate  fin  spines  (in  all  spiny 
rayed  fins  in  H.  vivanus,  only  in  the  pelvic  of  P.  eos),  and  spiny 
scales. 

Pronotogrammus  aureorubens  (Fig.  267b).  — Baldwin  (pers. 
comm.)  has  found  larvae  from  the  western  Atlantic  that  are 
heavily  spined  and  possess  the  meristic  characters  of  P.  aureo- 
rubens. These  larvae  are  completely  scaled,  have  serrations  on 
spines  of  all  spinous  fins  which  are  also  quite  stout,  and  have 
heavy  serrate  spines  in  the  opercular  region.  The  dorsal  aspect 
of  the  head  is  covered  with  spinous  ridges  including  a  complex 
cockscomb  spine  on  the  supraoccipital.  There  are  four  blotches 
of  pigment  dorsally  on  the  body:  two  ventral  to  the  first  dorsal 
fin,  one  ventral  to  the  second  dorsal  fin,  and  one  on  the  caudal 
peduncle. 

Holanthias  (Fig.  267rfA-Kendall  (1977,  1979)  illustrated  and 
briefly  mentioned  an  anthiine  larva  he  called  .4nthias  sp.  Type 
2  which  has  been  shown  to  be  Holanthias  martinicensis  (Bald- 


win, pers.  comm.).  These  larvae  are  deep-bodied  with  large 
heads  and  mouths.  They  develop  serrate  spines  in  the  opercular 
region,  and  a  simple  supraoccipital  spine  in  post-flexion  larvae. 
They  have  several  spines  above  the  eye  and  develop  scales  dur- 
ing the  larval  stage.  They  have  some  pigment  in  the  membrane 
of  the  first  dorsal  fin  as  well  as  a  line  on  the  body  ventral  to  the 
second  dorsal  fin.  Baldwin  (pers.  comm.)  has  pointed  out  the 
similarities  between  Holanthias  martinicensis  larvae  and  those 
Kendall  (1977,  1979)  described  as  .4nthias  gordensis,  including 
the  early  appearance  of  scales,  not  noted  by  Kendall  (1977, 
1979). 

Selenanlhias.—A  transforming  specimen  illustrated  and  briefly 
described  by  Fourmanoir  (1973)  is  deep-bodied  but  has  no  elon- 
gate fin  spines.  It  appears  to  be  fully  scaled  and  has  stout,  pos- 
sibly serrate  preopercular  and  interopercular  spines. 

Epinephelinae 

Johnson  (1983)  has  dealt  with  the  systematics  of  several  gen- 
era that  had  been  thought  variously  related  to  each  other.  These 
are  mainly  genera  in  the  epinepheline-grammistine  lineage  of 
Kendall  (1976).  On  the  basis  of  several  characters,  Johnson 
proposed  that  these  genera  form  a  monophyletic  lineage 
(subfamily  Epinephelinae)  that  is  composed  of  five  tribes  (Ni- 
phonini,  Epinephelini,  Diploprionini,  Liopropomini,  and 
Grammistini).  Some  early  life  history  stages  are  known  for  all 
of  the  tribes  except  Niphonini  (Fig.  268).  The  larvae  share  the 
elongation  of  one  or  two  anterior  dorsal  spines,  and  the  larvae 
and  adults  share  predorsal  bone  and  pterygiophore  arrange- 
ments which  presumably  function  to  support  the  larval  dorsal 
spines  (Johnson,  1983).  In  the  Epinephelini,  the  dorsal  spines 
are  stout  and  serrate,  whereas  in  the  other  three  tribes  they  are 
extremely  elongate,  flexible,  and  some  have  siphonophore-mim- 
icking  pigment  and  shape. 

The  following  is  a  summary  of  what  is  known  of  the  mor- 
phology of  early  life  history  stages  of  fishes  in  the  epinepheline 
tnbes  of  Johnson  (1983). 

J^iphonini.  —  Niphon  spinosus.  the  sole  member  of  this  tribe,  has 
unknown  larvae  but  Johnson  (1983)  speculated  that  on  the  basis 
of  first  dorsal  pterygiophore  morphology  and  presumed  rela- 
tionships, their  third  dorsal  spine  should  be  elongate. 

Epinephelini.— Larvae  are  known  only  for  those  genera  occur- 
ring in  Atlanto-American  waters.  Several  species  have  been 
reared  and  their  egg  and  larval  development  described  (see  Table 

123). 

Epinephelus.  — Larvae  of  species  from  every  ocean  belonging  to 
this  circumtropical  genus  are  known.  Smith  (1971)  placed  the 
American  members  of  the  genus  in  five  subgenera:  Epinephelus, 
Promicrops,  Cephalopholis,  Dermatolepis.  and  .-ilphestes.  These 
had  formerly  been  considered  genera,  and  members  of  these 
occur  in  other  parts  of  the  world.  Johnson  and  Ashe  ( 1 984)  were 
able  to  identify  larvae  of  most  species  of  American  Epinephelus 
primarily  on  the  basis  of  spinelets  on  the  elongate  dorsal  and 
pelvic  spines.  They  compared  spinelet  patterns  among  members 
of  the  subgenera  and  species  groups  of  Smith  (1971)  and  found 


Fig.  268.     Examples  of  epinepheline  larvae:  (A)  Epinephelini:  Paranthias  furcifer.  8.6  mm.  from  Kendall  ( 1 979);  (B)  Liopropomini:  Liopropoma 
sp.,  1 1.0  mm.  Collected  by  G.  R.  Harbison,  16  May  1981,  b^Sl.S'S,  150°21.8'E;  and  (C)  Grammistmi:  Ryplicus  sp.,  6.6  mm,  from  Kendall  (1979). 


508 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


that  most  share  common  patterns  (e.g.,  species  groups  E.  sthatus 
and  E.  adscensionis),  although  there  are  some  notable  problems 
(subgenera  Cephalopholis  and  Alphestes).  Thus,  in  Epinephehis 
there  is  general  concordance  between  the  only  distinguishing 
characters  of  the  larvae  (spinelet  patterns)  and  the  relationships 
hypothesized  based  on  a  variety  of  adult  characters;  but  thor- 
ough analysis  must  be  done  to  resolve  apparent  discrepancies. 

Paranthias.  —  One  species  (P.furcifer)  occurs  in  American  waters 
of  the  Atlantic  and  Pacific  Oceans.  The  larvae  have  a  unique 
spinelet  pattern  on  the  dorsal  fin  spines,  and  have  internal  no- 
tochord  pigment  not  found  in  other  epinephelines  (Johnson  and 
Ashe,  1984).  This  genus  as  an  adult  is  quite  distinct  ecologically 
and  morphologically. 

Mycteroperca.— This  American  genus  with  13  species  is  distin- 
guished from  the  other  epinepheline  genera  by  several  charac- 
ters, including  usually  having  more  anal  rays  (11-13).  The  species 
of  Mycteroperca  cannot  be  distinguished  as  larvae,  and  their 
spinelet  patterns  resemble  those  of  several  members  of  Epineph- 
elus  (e.g.,  E.  niveatus,  E.  flavolimbatus,  and  E.  acanthistius). 
However,  Mycteroperca  larvae  have  a  melanophore  at  the 
cleithral  symphysis,  which  is  not  found  in  any  of  these  species 
oi  Epinephelus  (Johnson  and  Ashe,  1984). 

Gon/op/ec/rus.  —  Postflexion  larvae  of  the  only  species,  Gonio- 
plectrus  hispamis,  are  known  (Kendall  and  Fahay,  1979).  The 
larvae  are  more  robust  and  have  shorter  elongate  dorsal  and 
pelvic  spines  than  other  American  epinephelines.  Also,  these 
elongate  spines  are  different  in  cross  section  and  spinelet  ap- 
pearance than  those  of  other  epinephelines  (Johnson  and  Keen- 
er, 1984). 

Diploprionini.  — A  photograph  of  a  transforming  larva,  a  draw- 
ing of  a  juvenile,  and  a  brief  description  of  the  juvenile  showed 
fish  with  long  flexible  dorsal  spines  and  rather  deep  bodies  (Hubbs 
and  Chu,  1934).  The  second  and  third  dorsal  spines  are  ex- 
tremely produced  in  the  larva,  but  only  the  third  is  in  the  ju- 
venile. The  photograph  of  the  larva  does  not  allow  more  detailed 
observation. 

Liopropomini.  — Larvae  of  Liopropoma/ Pikea  are  known  and 
cannot  presently  be  distinguished  on  the  basis  of  larval  char- 
acters (Kendall,  1977,  1979).  They  were  first  described  as  a  new 
genus,  Flagelloserranus.  by  Kotthaus  ( 1 970).  Jeboehlkia  is  known 
from  a  single,  small  specimen  which  shows  traits  of  being  a 
transforming  larva  (Robins,  1967). 

Lioproma/Pikea.  — The  general  body  shape  is  similar  to  that  of 
the  serranines,  although  the  gut  is  shorter  and  there  is  a  space 
between  the  anus  and  the  origin  of  the  anal  fin.  The  caudal 
peduncle  is  both  longer  and  deeper  than  it  is  in  serranines.  The 
most  outstanding  developmental  feature  is  the  presence,  even 
in  small  larvae,  of  two  elongate,  thin  dorsal  spines.  These  de- 
velop before  other  fin  rays,  reach  a  length  of  up  to  three  times 
the  fish  length,  and  become  the  second  and  third  dorsal  spines. 
These  spines  are  delicate  and  are  broken  in  many  specimens. 
Kotthaus  ( 1 970)  described  the  presence  of  thick  tissue  surround- 
ing these  spines;  the  tissue  around  the  second  spine  has  two 
vane-like  swellings  on  its  distal  third  and  the  tissue  around  the 
third  spine  is  tubular  for  its  entire  length.  The  distal  portion  of 
both  spines  is  pigmented  with  several  large  melanophores.  The 


remaining  fin  rays  develop  their  adult  proportions  without  any 
pronounced  elongations.  The  ventral  fins  develop  more  slowly 
than  those  of  most  other  serranids. 

Except  for  the  pigment  on  the  elongate  dorsal  fin  spines,  most 
larvae  are  unpigmented.  Some  spots  develop  on  the  hindbrain 
surface  in  larger  larvae,  probably  representing  the  onset  of  ju- 
venile pigment. 

Jeboehlkia.— The  single  species  {J.  gladifer)  is  known  only  from 
the  holotype,  a  40.8  mm  female.  Characters  that  indicate  that 
it  may  not  have  completed  transformation,  or  may  be  paedo- 
morphic,  include  the  virtual  lack  of  pigment,  the  enlarged  eye, 
and  the  elongate  first  dorsal  spine  (see  Robins,  1967). 

Grammistini.  — Fishes  in  this  tribe  have  been  variously  grouped 
as  members  of  families  separate  from  the  serranids  and  as 
subfamilies  of  the  serranids.  Larvae  of  four  of  the  seven  genera 
placed  in  this  tribe  by  Johnson  (1983)  are  known.  The  first  or 
second  dorsal  spine  is  elongate  and  flexible,  and  the  preopercular 
margin  is  armed  with  about  five  subequal  spines  in  larvae  of  all 
four  genera. 

Grammistes.—A  single,  1 1  mm  postflexion  larva  of  G.  se.xiline- 
atus  illustrated  by  Fourmanoir  (1976)  has  an  elongate  flexible 
first  dorsal  spine  and  five  spines  on  the  preopercular  margin.  It 
is  well  developed,  rather  deep-bodied,  and  appears  to  lack  pig- 
ment except  on  the  pectoral  fin  which  is  covered  with  fine  me- 
lanophores on  its  distal  third. 

Aporops.— The  anterior  portion  of  a  12  mm  postflexion  larva 
of  Aporops  bilinearis  illustrated  by  Fourmanoir  (1976)  has  the 
first  dorsal  spine  elongate  and  flexible  and  five  spines  on  the 
preopercular  margin.  It  is  well  developed  and  is  not  as  deep- 
bodied  as  the  aforementioned  Grammistes  larva.  No  pigment 
is  evident  in  the  illustration. 

Pseudogramma.  —  A  developmental  series  of  P.  gregoryi  was 
described  by  Kendall  (1977,  1979)  and  Leis  and  Rennis  (1983) 
illustrated  a  series  of  P.  polyacantha.  These  larvae  have  shallow 
tubular  bodies;  a  greatly  elongate,  flexible  dorsal  spine  (the  first 
or  second);  precocious  enlarged  pectoral  fins;  a  gap  between  the 
anus  and  the  anal  fin;  and  a  general  lack  of  pigment  except  on 
the  pectoral  fin  of  small  larvae  and  on  the  sheath  that  surrounds 
the  elongate  dorsal  spine. 

i?.V77;;cM.s.  —  Aboussouan  (1972b)  illustrated  and  briefly  de- 
scribed two  larvae,  and  Kendall  (1977,  1979)  compared  these 
with  specimens  he  described  from  the  western  Atlantic.  These 
larvae  have  the  first  dorsal  spine  produced,  flexible,  and  sur- 
rounded by  a  pigmented  sheath;  about  five  preopercular  spines; 
an  enlarged  pectoral  fin  that  may  be  pigmented;  rather  long  rays 
in  the  second  dorsal,  caudal,  and  anal  fins;  small,  late-developing 
pelvic  fins;  a  lack  of  body  pigment;  and  are  moderately  deep- 
bodied  at  the  nape. 

Relationships 

Although  known  larvae  of  serranids  show  a  diversity  of  char- 
acters that  will  probably  permit  them  to  be  used  in  definitive 
studies  of  relationships  within  the  group,  such  studies  are  pres- 
ently premature  (Fig.  269).  More  characters  need  to  be  traced 
ontogenetically,  and  larvae  of  more  species,  particularly  in  the 
Anthiinae  and  several  tribes  of  Epinephelinae,  need  to  be  de- 


KENDALL:  SERRANIDAE 


509 


preopercular 


interopercular 


preopercu 


interopercular 


B 


preopercu 


nteropercular 


D 


preopercular 


interopercular 


Fig.  269.  Representative  preopercular  and  interopercular  bones  from  larval  serranids  (from  Kendall,  1979);  (A)  Serraninae:  Serranus  sp.;  (B) 
Anthiinae:  Anthias  sp.  Type  1;  (C)  Epinephelinae:  Epinephelini,  Epmephelus  nivealus:  and  (D)  Epinephelinae:  Grammistini.  Pseudogramma 
gregoryi. 


scribed.  At  present,  however,  some  statements  can  be  made 
concerning  serranid  systematics  from  what  is  known  about  the 
larvae. 

The  serranid  subfamilies  are  clearly  distinct  as  larvae.  In  fact, 
it  is  not  possible  to  characterize  the  Serranidae  based  on  larval 
morphology,  because  no  characters  unite  the  subfamilies  while 
separating  them  from  larvae  of  all  other  families.  Serraninae 
larvae  seem  to  be  the  least  specialized  and  are  more  similar  to 
percoid  genera  thought  to  represent  the  basal  stock  from  which 
serranids  arose  (e.g.,  Morone.  Lateolahrax.  and  Dicentrarchus). 
The  serranine  genera  can  be  distinguished  from  each  other  and 
ordered  in  a  rough  progression  of  divergence  from  the  supposed 
ancestral  larval  form  (as  exemplified  by  Morone),  as  follows: 
Serraniculus.  Centroprislis-Paralabrax,  Diplectrum  Type  1, 
Serranus  (see  Kendall,  1979).  Characters  that  lead  to  this  as- 
sessment include  pigment,  body  shape,  sequence  of  dorsal  spine- 
soft  ray  development,  and  dorsal  fin  spine  elongation. 


Based  on  larval  and  other  evidence,  it  appears  that  two  major 
radiations  from  the  ancestral  serranines  arose  leading  to  the 
anthiine  and  the  epinepheline  lineages.  The  anthiines  form  a 
fairly  cohesive  group  of  fish  which  are  at  the  same  time  quite 
speciose.  The  generic  alignment  of  many  anthiines  is  unclear 
and  in  some  cases  larval  evidence  is  in  conflict  with  that  based 
on  adults.  Anthiine  larvae,  like  the  adults,  share  several  char- 
acters that  unite  them,  yet  they  are  quite  diverse  and  will  prob- 
ably prove  to  be  excellent  subjects  for  phylogenetic  investiga- 
tions. Larvae  of  only  about  half  of  the  presently  understood 
anthiine  genera  are  known  to  any  extent,  some  of  them  only 
from  one  transforming  larva.  Thus  the  lack  of  generic  revisions 
and  incomplete  knowledge  of  larval  development  makes  it  pres- 
ently unreasonable  to  attempt  a  thorough  systematic  assessment 
that  would  include  larvae.  Within  the  group,  a  progression  of 
increasing  spinyness  and  armature  is  apparent.  Among  the  lar- 
vae described  to  date,  armature  seems  to  be  added  as  follows: 


510 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES -AHLSTROM  SYMPOSIUM 


elongate  preopercular  and  interopercular  spines,  serrate  pre- 
opercular  and  interopercular  spines,  stout  pelvic  and  first  three 
dorsal  spines,  supraoccipital  spine,  serrate  dorsal  and  pelvic 
spines,  serrate  head  spines  on  several  bones,  and  spiny  scales 
developing  during  the  larval  stage. 

The  other  major  line  of  divergence  from  the  serranines  is  the 
five  tribes  of  the  epinephelines.  Johnson  (1983)  pointed  out  the 
adult  features  that  characterize  this  subfamily  and  the  tribes 
within  it,  although  he  did  not  provide  a  detailed  analysis  of  the 
relationships  among  the  tribes.  The  larvae  (representatives  of 
four  tribes  are  known)  all  have  one  or  two  quite  elongate  dorsal 
spines.  In  the  Epinephelini,  the  elongate  dorsal  spines  are  stout 
and  serrate;  in  the  other  tribes,  they  are  flexible,  thin,  and  in  an 
elaborately  pigmented  sheath.  Thus  it  appears  from  the  larvae 
that  the  Diplopionini,  Liopropomini,  and  Grammistini  may 
form  a  monophyletic  group  within  the  Epinephelinae. 

Epinephelini  larvae  are  all  quite  similar  but  some  genera  can 
be  separated  by  larval  characters  (Gonioplectrus,  and  Paran- 
ihias),  although  larvae  are  unknown  for  several  genera.  Gon- 
ioplectrus  larvae  are  most  similar  to  anthiine  larvae  and  may 
represent  the  most  primitive  extant  epinephelini  state.  Johnson 
(1983)  suggested  that  Niphon  represented  the  primitive  sister 


group  of  all  other  epinephelines  and  that  its  unknown  larvae 
may  have  an  elongate  third  rather  than  second  dorsal  spine. 
There  is  less  variation  in  size  of  the  second  and  third  dorsal 
spines  in  Gonioplectrus,  compared  to  other  Epinephelini,  which 
adds  credence  to  the  above  suggested  relationships. 

Few  larval  representatives  of  the  other  epinepheline  tribes 
[grammistine  lineage  of  Kendall  (1976)]  are  known  and  none 
of  them  have  been  studied  in  detail.  Their  elongate,  pigmented 
flexible  dorsal  spines,  lack  of  corresponding  elongate  pelvic 
spines,  five  subequal  preopercular  spines,  and  dearth  of  body 
pigment  unite  the  known  larvae.  Larvae  of  Diploprion  are  rather 
deep-bodied  compared  to  the  more  tubular  bodies  of  the  other 
known  larvae  grammistines.  The  second  and  third  dorsal  spines 
are  produced  in  Diploprion  and  Liopropoma.  but  only  one  spine 
is  produced  in  members  of  the  Grammistini.  In  this  group  of 
serranids  there  appear  to  be  larval  characters  that  will  be  helpful 
in  systematic  studies,  but  larvae  of  more  representatives  must 
be  known  in  more  detail  before  such  studies  will  be  meaningful. 

National  Marine  Fisheries  Service,  Northwest  and  Alaska 
Fisheries  Center,  2725  Montlake  Boulevard  East, 
Seattle,  Washington  981 12. 


Carangidae:  Development 
W.  A.  Laroche,  W.  F.  Smith-Vaniz  and  S.  L.  Richardson 


THE  family  Carangidae  (jacks,  trevallys,  and  pompano)  has 
traditionally  been  assigned  to  the  suborder  Percoidei,  an 
assemblage  of  generalized  perciform  fishes  (Lauder  and  Liem, 
1983).  The  family  is  notably  heterogenous,  including  species 
which  differ  widely  in  structure  and  appearance.  Phylogenetic 
relationships  within  the  suborder  and  even  the  familial  limits 
of  the  Carangidae  are  not  clearly  established  (see  Smith-Vaniz, 
this  volume).  The  family  is  composed  of  approximately  140 
species  and  30  genera  (Table  1 25)  many  of  which  remain  poorly 
defined. 

Carangids  are  found  world-wide  in  tropical  and  warm  tem- 
perate marine  and  estuarine  waters.  Carangids  are  actively 
swimming  fishes  which  range  from  small  schooling  planktivores 
to  large  solitary  piscivores  (Berry  and  Smith-Vaniz,  1978).  Some 
species  of  carangids  are  known  to  spawn  pelagically  offshore, 
i.e.,  Seriola  lalandi  =  S.  dorsalis  (Baxter,  1960)  and  Trachurus 
symmetricus  (Ahlstrom  and  Ball,  1954),  while  others  spawn 
close  to  shore  and  near  the  bottom,  i.e.,  Caranx  ignobilis  (von 
Westemhagen,  1974)  and  Oligoplites  saurus  (Aprieto,  1974). 
The  greatest  amount  of  information  concerning  early  life  stages 
exists  for  species  of  Decaplerus  and  Trachurus  on  which  research 
has  focused  due  to  their  commercial  importance. 

Development 

Eggs 

Carangids  have  spherical,  pelagic  eggs  which  have  a  narrow 

perivitelline  space  and  range  in  diameter  from  about  0.7  to  1.3 

mm.  One  to  several  oil  globules  are  usually  present,  and  egg 


envelopes  are  clear,  unsculptured,  and  lack  filaments  (Ahlstrom 
and  Ball,  1954;  Miller  and  Sumida,  1974;  James,  1976a).  The 
eggs  of  Naucrates  ductor  have  erroneously  been  reported  to  be 
demersal,  adhesive,  with  a  fine  entangling  filament  at  one  pole 
(Gilchrist,  1918)  and  attached  to  sharks  and  the  hulls  of  ships 
(Gilchrist,  1918;  Shuleikin,  1958).  They  are  actually  pelagic, 
non-adhesive,  and  without  filaments  (Barnard,  1926;  Sanzo, 
1931a;  Maksimov,  1969). 

Development  proceeds  in  the  typical  manner  of  pelagic  fish 
eggs  (Ahlstrom  and  Ball,  1954;  Miller  and  Sumida,  1974).  Eggs 
hatch  24  to  48  hours  after  spawning  at  water  temperatures  be- 
tween 18  and  30  C°  (temperature  range  within  which  eggs  and 
larvae  are  most  commonly  taken). 

Carangid  eggs  are  similar  in  size  and  appearance  to  those  of 
many  other  marine  fishes.  Thus,  identification  even  to  family 
level  may  be  difficult  or  frequently  impossible  using  presently 
known  characters. 

Larvae 

A/or/)/;o/(7gi'.  — Information  is  available  on  at  least  one  devel- 
opmental stage  for  58  of  the  140  valid  species  representing  24 
of  30  genera  (Table  125).  However,  even  among  those  taxa  for 
which  descriptive  information  is  available,  inconsistent  quality 
in  descriptive  text  and  coverage  of  the  developmental  period 
make  detailed  morphological  comparisons  and  identifications 
based  upon  these  descriptions  difficult  in  many  cases.  Laroche 
et  al.  (MS)  have  refined  developmental  terminology  for  caran- 
gids so  as  to  define  developmental  stages  more  precisely  and 
thus  improve  comparability  of  descriptions  between  taxa. 


LAROCHE  ET  AL.:  CARANGIDAE 


511 


Table  125.    Species  and  World  Distribution  List  for  the  Family  Carangidae.  Selected  literature  references  deal  with  descriptions  of  larvae 

and  juveniles. 


Species 


Ind.' 

Wesl 

Cent* 

East 

Wesl 

Easl 

Ocean 

Pac. 

Pac 

Pac, 

All. 

All. 

References* 


Aleclis  alexandnniis  (E.  Geoffrey  St.-Hilaire) 
Alectis  cilians  (Bloch) 

Aleclis  indicus  (Riippell) 

Alepes  djedaba  (ForsskSl) 

Alepes  melanoptera  Swainson 

Alepes  sp. 

Alepes  vari  (Cuvier) 

Atropus  alropos  (Bloch  and  Schneider) 

Atule  male  (Cuvier) 


Campogramma  glaycos  (Lacepede) 
Carangoides  armaliis  (Riippell) 
Carangoides  bajad  (ForsskSl) 
Carangoides  bartholomaei  (Cuvier) 

Carangoides  caeruleopinnalus  (Riippell) 

Carangoides  chrysophrys  (Cuvier) 

Carangoides  dinema  Sleeker 

Carangoides  equula  (Temminck  and  Schlegel) 

Carangoides  ferdau  (ForsskSl) 

Carangoides  fulvoguttatus  (ForsskSl) 

Carangoides  gymnosielhus  (Cuvier) 

Carangoides  hedlandensis  (Whitley) 

Carangoides  humerosus  (McCulloch) 

Carangoides  malabaricus  (Bloch  and  Schneider) 

Carangoides  oblongus  (Cuvier) 

Carangoides  orthogrammus  Jordan  and  Gilbert 

Carangoides  olrymler  (Jordan  and  Gilbert) 

Carangoides  plagiolaenia  Sleeker 

Carangoides  praeustus  (Sennett) 

Carangoides  ruber  (Bloch) 

Carangoides  talamparoides  Sleeker 

Carangoides  uii  (Wakiya) 

Carangoides  vinclus  (Jordan  and  Gilbert) 

Caranx  bucculentus  Alleyne  and  Macleay 

Caranx  caballus  Giinther 

Caranx  caniniis  Giinther 

Caranx  crysos  (Mitchill) 


Caranx  hippos  (Linnaeus) 

Caranx  ignobilis  (ForsskSI) 

Caranx  lalus  Agassiz 

Caranx  lugubris  Poey 

Caranx  melampygus  Cuvier 

Caranx  papuensis  Alleyne  and  Macleay 

Caranx  sem  Cuvier 

Caranx  senegallus  Cuvier 

Caranx  sexfasciatus  Quoy  and  Gaimard 

Caranx  tille  Cuvier 

"Caranx"  kohcru  Hector 

"Caranx"  para  Cuvier 

"Caranx"  rhonchus  E.  Geoffroy  St.-Hilaire 

Chloroscomhrus  chrysurus  (Linnaeus) 

Chloroscombrus  orqueta  Jordan  and  Gilbert 

Decaplcrus  kurroides  Sleeker 

Decapteriis  macarellus  (Cuvier) 

Decaplcrus  macrosoma  Blecker 

Decaplcrus  maruadsi  (Temminck  and  Schlegel) 

Decaplcrus  muroadsi  (Temminck  and  Schlegel) 

Decaplcrus  punctalus  (Cuvier) 


+ 

+ 

- 

+ 

+ 

- 

+ 

+ 

- 

- 

+ 

- 

+ 

+ 

- 

+ 

+ 

- 

+ 

+ 

+ 

+ 

+ 

- 

+ 

+ 

: 

+ 

+ 

_ 

+ 

+ 

- 

+ 

+ 

- 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

- 

+ 

+ 

- 

+ 

+ 

- 

- 

+ 

- 

+ 

+ 

- 

+ 

+ 

- 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

- 

+ 

+ 

— 

+ 

+ 

— 

+        *Aboussouan.  1975 
+       'Aboussouan,  1968a:  *Fowler,  1936; 
Ginsburg  1952;  *Johnson,  1978 

•Tsokur,  1977 


*Kuthalingam,  1959a;  *Miller  and  Sumida, 
1974;  •Miller  et  al..  1979;  Zvyagina  and 
Rass,  1977 


•Berry,  1959b;  •Johnson  1978;  Uroche  et 
al.  (in  prep.) 


+ 

+ 

+ 

+ 

+ 

+ 

+ 

- 

+ 

+ 

+ 

- 

+ 

- 

+ 

- 

+ 

- 

— 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

- 

- 

+ 

+ 

— 

•Berry,  1959b 


+  Aboussouan,  1975;  Berry,  1959b;  Johnson, 
1978;  Montolio.  1976;  McKenney  et  al., 
1958 

+        Berry,  1959b;  Johnson,  1978 


Berry,  1959b;  Johnson,  1978 


+       Aboussouan,  1975 

-       Ahlstrom  and  Sumida  (in  prep.) 


•Bapat  and  Prasad,  1952 

Aboussouan,  1967;  Aboussouan,  1975;  Co- 

nand  and  Franqueville   1973 
Aboussouan,  1968a;  Aboussouan,  1975; 

Laroche  et  al.  (in  prep.) 
Ahlstrom  and  Sumida  (in  prep.) 


?Delsman,  1926a 
Shojima,  1962 

■Aboussouan,  1975;  Aprieto.  1974;  Hilde- 
brand  and  Cable,  1930;  Johnson,  1978; 
Montolio,  1976 


512 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  125.    Continued. 


Genus' 


Species 


Ind.' 
Ocean 


Wcsl 
Pac. 


Cent.' 
Pac 


East 
Pac 


West 
All, 


East 
Atl. 


Decapterus  russelti  (Riippell) 

Decaptenis  scomhrinus  (Valenciennes) 
Decapterus  labl  Berry 
Decapterus  n.  sp.  "stonebrass  scad'" 
Elagatis  bipinnulata  (Quoy  and  Gaimard) 

Gnathanodon  speciosus  (ForsskSl) 

Hemicaranx  amhiyrhynchus  (Cuvier) 

Hemicaranx  bicolor  (Gunther) 
Hemicaranx  leucurus  (Gunther) 
Hemicaranx  zelotes  Gilbert 
Lichia  amia  (Linnaeus) 
Magalespis  cordyta  (Linnaeus) 
Naucrates  ductor  (Linnaeus) 


Otigoplites  altus  (Gunther) 
Oligoplhes  patometa  (Cuvier) 
Otigoplites  refidgens  Gilbert  and  Starks 
Otigoplites  saliens  (Bloch) 
Otigoplites  saurus  (Schneider) 

Panlotahus-  radiatus  (Macleay) 
Parastromateus  niger  (Bloch) 
Parana  signata  (Jenyns) 
Pseudocaranx  clulensis  (Guichenot) 
Pseudocaranx  dentex  (Bloch  and  Schneider) 

Pseudocaranx  nrigfiti  (Whitley) 

Scomberoides  commersonianus  Lacepede 

Scomberoides  lysan  (ForsskSl) 

Scomberoides  tala  (Cuvier) 

Scomberoides  tot  (Cuvier) 

Selar  boops  (Cuvier) 

Setar  crumenophlhatmus  (Bloch) 

Selaroides  leptolepis  (Cuvier) 
Selene  brevoortii  (Gill) 
Selene  brownii  (Agassiz) 
Selene  dorsalis  (Gill) 

Selene  oerstedii  Lutken 
Selene  peruviana  (Guichenot) 
Selene  setapinnis  (Mitchill) 


+ 
+ 
+ 


Selene  vomer  (Linnaeus) 


Seriola  carpenteri  Mather 
Seriota  dumerili  (Risso) 


Serwta  fasciata  (Bloch) 

Seriola  hippos  Gunther 
Seriola  latandi  Valenciennes 

Seriola  peruana  Steindachner 

Seriola  quinqueradiata  Temminck  and  Schlegel 

Seriola  rivoliana  Cuvier 


+ 

+ 

+ 

- 

— 

+ 

+ 

- 

+ 

+ 
+ 
+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

_ 

_ 

— 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 
+ 


+ 
+ 


+ 
+ 


+         + 

-         + 
+ 


?*Delsnian,  1926a;  ?*Tsokur,  1977;  Vijay- 
araghavan,  1958 


Aprieto,  1974;  Berry,  1969;  Johnson,  1978; 

Laroche  et  al.  (in  prep);  Okiyama,  1970 
Ahlstrom  and  Sumida  (in  prep.);  Miller  et 

al.,  1979 
Hoese  and  Moore,  1977;  Laroche  et  al.  (in 

prep.) 


Lo  Bianco,  1909;  Padoa,  1956c 

?Kuthalingam,  1959a 

Ahlstrom  and  Sumida  (in  prep.);  Lutken, 
1880;  Padoa,  1956c;  Pertseva-Ostroumova 
and  Rass,  1973;  Roule  and  Angel,  1930; 
Sanzo,  1930a,  1931 


Aprieto,  1974;  Johnson,  1978;  Laroche  et 
al.  (in  prep.) 


Phonlor,  1979 

♦James,  1976a;  *Padoa,  1956c;  ?*Schnaken- 
beck,  1931 


•Premalatha,  1977 


*Delsman,  1926a;  *Miller  et  al.,  1979;  Zvy- 

agina  and  Rass,  1977 
?Bapat,  1955 

Ahlstrom  and  Sumida  (in  prep.) 
Laroche  et  al.  (in  prep.) 
*Aboussouan,  1975;  *Conand  and  Franque- 

ville,  1973 

Ahlstrom  and  Sumida  (in  prep.) 
*Fowler,  1936;  *Ginsburg,  1952;  Johnson, 

1978;  Laroche  et  al.  (in  prep.);  Lutken, 

1880 
Aprieto,  1974;  Fowler,  1936;  Ginsburg, 

1952;  Johnson,  1978;  Laroche  et  al.  (in 

prep.);  Lutken,  1880 

?Hildebrand  and  Cable,  1930;  Johnson, 
1978;  Laroche  et  al.  (in  prep.);  Padoa, 
1956c;  Roule  and  Angel.  1930;  Sanzo. 
1930c,  1933b 

Ginsberg,  1952;  Johnson,  1978;  Laroche  et 
al.  (in  prep.) 


+        Ahlstrom  and  Sumida  (in  prep.);  ?Brownell, 
1979 

Lutken,  1880;  Mitani,  1960;  Uchida,  Dotsu 
et  al.,  1958 
+        *Ginsburg.  1952;  Laroche  et  al.  (in  prep.) 


LAROCHE  ET  AL.:  CARANGIDAE 


513 


Table  125.    Continued. 


Species 


Ind  ' 
Ocean 


West 
Pac. 


Cenl' 
Pac 


East 
Pac 


West 
Atl. 


East 
Atl. 


Seriola  zonata  (Mitchill) 

Seriolina  nigrofasciala  (Ruppell) 
Trachinotus  afncanus  Smith 
Trachinolus  anak  Ogilby 
Trachinotus  baiUonii  (Lacepede) 
Trachinolus  blochii  (Lacepede) 
Trachinotus  carolinus  (Linnaeus) 

Trachinotus  cayennensis  Cuvier 
Trachinotus  fatcatus  (Linnaeus) 


Trachinolus  goodei  Jordan  and  Eveimann 

Trachinotus  goreensis  Cuvier 

Trachinotus  kennedyi  Steindachner 

Trachinolus  marginatus  Cuvier 

Trachinolus  ma.xillosus  Cuvier 

Trachinotus  mookalee  Cuvier 

Trachinolus  ovatus  (Linnaeus) 

Trachinolus  paitensis  Cuvier 

Trachinotus  rhodopus  Gill 

Trachinolus  russelii  Cuvier 

Trachinotus  slilbe  (Jordan  and  MacGregor) 

Trachinotus  terala  Cuvier 

Trachinolus  velox  Ogilby 

Trachurus  declivis  (Jenyns) 

Trachurus  delagoa  Nekrassov 

Trachurus  japonicus  (Temminck  and  Schlegel) 

Trachurus  indicus  Nekrassov 

Trachurus  ialhainl  Nichols 

Trachurus  incditcrraneus  (Steindachner) 
Trachurus  murphyl  Nichols 
Trachurus  novaezelandiae  Richardson 
Trachurus  picluratus  (Bowdich) 
Trachurus  syinmelricus  (Ayres) 

Trachurus  trachurus  (Linnaeus) 


Trachurus  trecae  Cadenat 
Ulua  aurochs  (Ogilby) 
Ulua  nwnlalis  Cuvier 
Uraspis  hclvola  Forster 
Uraspis  secunda  Poey 
Uraspis  uraspis  Giinther 


+ 
+ 


+ 
+ 


+ 
+ 


+ 
+ 


+ 

+ 

- 

- 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

+ 

— 

— 

Aprieto,  1974;  Ginsburg,  1952;  Johnson, 
1978;  Lutken,  1880 


Fields,  1962;  Johnson,  1978;  Laroche  et  al. 
(in  prep.) 

Fields,  1 962;  Hildebrand  and  Schroeder, 
1928;  Johnson,  1978;  Laroche  et  al.  (in 
prep.) 

•Fields.  1962;  Johnson.  1978;  Laroche  et  al. 
(in  prep.) 

Aboussouan,  1975 


•De  Gaetani,  1940;  ♦Padoa,  1956c 


Shojima,  1962;  Uchida  et  al.,  1958 

♦Tsokur,  1977 

*de  Ciechomski  and  Weiss,  1973;  Johnson, 

1978;  Laroche  et  al.  (in  prep.) 
Demir,  1961;  Padoa,  1956c;  Sanzo,  1932a 
Santander  and  de  Castillo,  1971 

Aboussouan,  1975 

Ahlstrom  and  Ball,  1954;  Ahlstrom  and 
Sumida  (in  prep.) 

Aboussouan,  1975;  Arbault  and  Boutin, 
1968c;  Brownell,  1979;  Demir,  1961;  Eh- 
renbaum,  1905-1909;  Haigh,  1972b;  Kili- 
achenkova,  1970;  Kingetal.,  1977;  Leta- 
connoux,  1951;  Padoa,  1956c;  Russell, 
1976;  Schnakenbeck,  1931 

Aboussouan,  1967;  Aboussouan,  1975 


Johnson,  1978 


'  Carangid  getienc  limits  arc  not  well  established  and  some  taxa  here  rccogni/ed  ultimately  may  be  allocated  to  subgenenc  status.  Carnn^otdes  is  a  poorly  defined  group  that  may  include 
several  subunits  worthy  of  recognition-  The  three  species  assigned  to  "Carunx"  are  not  closely  related  and  their  generic  placement  is  uncertain. 

-  Panlntahus  Whitley,  1931  is  here  recognized  as  a  senior  synonym  oi  Absalom  Whitley.  1937.  Recent  examination  of  the  syntypes  oi  Caranx  parasilus  Garman  {type-species  of  Panlolabus) 
has  revealed  that  they  are  conspecific  with  C  radiants  Macleay.  type-species  of  Absalom 

'  Species  that  reach  their  western  distnbutional  limit  on  the  eastern  margin  of  the  Indian  Ocean  (including  western  Australia)  are  not  tabulated  as  occumng  in  the  Indian  Ocean 

*  Species  that  reach  their  eastern  distnbutional  limit  on  the  western  margin  of  the  Pacific  Plate  (see  Springer.  1 982)  are  not  tabulated  as  occumng  in  the  Central  Pacific;  Easier  Island  is  treated 
as  a  component  of  the  central  Pacific. 

'  Aslensk  indicates  scientific  name  used  in  cited  reference  differs  from  present  allocation;  question  mark  indicates  only  a  provisional  identification  given  in  cited  reference,  or  adult  taxonomy 
of  group  so  inadequate  at  time  of  publication  that  specific  identification  must  be  treated  as  suspect 


Development  in  carangids  proceeds  relatively  directly  to- 
wards the  adult  stage.  Adult  characters  are  gradually  acquired 
without  remarkable,  sudden  metamorphoses  (developmental  rale 
changes)  occumng  between  stages. 


Carangid  lai-vae  are  relatively  small  and  undeveloped  at 
hatching,  usually  1.0  to  2.0  mm  notochord  length  (NL),  with  a 
relatively  large  yolk  sac.  Head  size,  presence  of  24-27  myo- 
meres, and  possession  of  an  oil  globule  at  the  anterior  of  the 


514 


ONTOGE>fY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  270.     (A)  Flexion  larva  (5.4  mm)  of  Trachiirus  lalhaini:  postflexion  larvae  (5.5,  5.6  mm)  of  (B)  Decaplerm  punclatus  and  (C)  Selar 
crumenophthalmus;  and  (D)  early  flexion  larva  (4.6  mm)  of  Chloroscombrus  chrysurus. 


Fig.  271.     (A)  Early  flexion  lar\a  (3.1  mm)  of  Aleclis  ciliaris;  (B)  postflexion  lana  (4.9  mm)  of  A  tide  male:  and  (C)  flexion  larva  (4.0  mm)  of 
Gnathanodon  speciosus. 


516 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


B 


»T^; 


■f'-'  ™*'* 


,^-J^ 


^ 


Fig.  272.     (A)  Postflexion  larva  (5.9  mm)  of  Trachinolus  caroUnus;  (B)  late  flexion  larva  (4.7  mm)  of  Naucrales  ductor.  and  (C)  poslflexion 
larva  (5.3  mm)  of  Scnmberoides  lysan. 


LAROCHE  ET  AL.:  CARANGIDAE 


517 


Fig.  273.     Late  postflexion  larvae  oi (\)  Elagatis  bipinnulata  (1 1.4  mm);  (B)  Oligoplites  saurus  (8.6  mm)  and  (C)  Seriola  zonata  (9.5  mm) 


yolk  sac,  ventral  to  the  head,  are  the  most  outstanding  characters 
of  yolk  sac  larvae.  The  mouth  is  not  formed,  and  the  gut  is 
undeveloped.  Eyes  lack  melanistic  pigmentation;  fins  are  un- 
developed; the  notochord  is  straight;  and  head  spines  are  lacking 
(Ahlstrom  and  Ball,  1954;  Aprieto,  1974;  Miller  and  Sumida, 
1974).  The  present  state  of  knowledge  is  not  adequate  to  estab- 
lish a  set  of  characters  which  will  distinguish  pre-fin  formation 
carangid  larvae  from  larvae  of  all  other  marine  fish  families  in 
the  world.  Newly  hatched  carangid  larvae  are  difficult  to  identify 
even  to  family  due  to  the  paucity  of  diagnostic  morphological 
characters  and  multitude  of  perciform  taxa  which  co-occur  and 
have  similar-appearing  larvae.  Since  larvae  of  many  taxa  remain 
unknown,  the  problem  is  even  more  complicated.  However, 
within  restricted  and  well-defined  geographic  areas  it  may  be 
possible  to  define  such  a  character  set  if  the  fish  fauna  is  well 
known  (Laroche  et  al.,  MS). 


Following  yolk  absorption,  larval  carangids  range  from  rel- 
atively slender  forms,  i.e.,  body  depth  (BD)  20  to  27%  SL  in 
Oligoplites  saurus  (Fig.  273B),  to  relatively  deep  bodied  forms, 
i.e.,  BD  32  to  59%  SL  in  Selene  sp.  (Aprieto,  1974)  (Fig.  274A). 
The  gut  develops  as  a  narrow  straight  tube  on  the  first  day  after 
hatching.  A  single  gut  loop  is  present  in  larvae  3-4  mm  NL, 
which  is  about  5  days  after  hatching  in  Atule  mate  and  Oligo- 
plites saurus  (Aprieto,  1974;  Miller  and  Sumida,  1974).  This 
pattern  seems  to  be  common  among  other  species  although 
lengths  at  which  the  gut  loops  vary  slightly.  The  gut  extends  to 
midbody  with  snout  to  anus  length  in  preflexion  and  flexion 
larvae  usually  ranging  from  46  to  67%  SL  (Aprieto,  1974;  Lar- 
oche et  al.,  MS).  The  head  ranges  in  length  from  about  24  to 
41%  SL  and  is  typically  about  33%  SL. 

Head  spines  form  relatively  early  in  development.  The  first 
head  spine  to  develop  is  a  preopercular  spine  at  the  angle  of  the 


518 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  1 26.    Distinguishing  Characters  Useful  in  Identification  (to  Genus)  of  Flexion  and  Postflexion  larvae  of  Carangidae.  Presence 

of  character  indicated  by  "  +  ,"  absence  by  "-"  and  no  data  by  "'0."  Species  and  sources  on  which  this  table  is  based  are  listed  in  preceding  table. 

except  for  original  observations  on  Gnathanodon  speciosus,  Naucrates  diictor.  Parastromaleus  niger.  and  Scomheroides  lysan.  Character  definitions 

follow  Laroche  et  al.  (MS).  Information  in  this  table  should  be  considered  preliminary,  awaiting  more  thorough  descriptions. 


Supra- 

occipital 

ridge 

Angle 

preopercular 

spine 

Supraocular  ridge 

Posltemporal  and 
supracieilhral  spines 

Pterotic 
ndgc 

Vomer 
pigment 

Weak 

Prom 

nent 

Dors 

Genus 

Simple 

Ser- 
rated 

Spinule(s) 

Small 
spine 

Ser- 
rated 

Ser- 
rated 

2  or  3 
spines 

Weak 

Promi- 
nent 

lateral 
pigment 

Aleclts 

+ 

+ 

_ 

- 

+ 

_ 

— 

_ 

+ 

_ 

_ 

+ 

_ 

Alepes 

+ 

+ 

- 

- 

+ 

- 

- 

- 

-)- 

- 

- 

-1- 

+ 

Atropus 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

Atule 

- 

+ 

- 

- 

+ 

- 

- 

- 

+ 

- 

- 

_ 

+ 

Campogramma 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

Carangoides 

+ 

+ 

- 

- 

+ 

- 

- 

- 

+ 

- 

- 

0 

+ 

Caranx 

+ 

+ 

- 

- 

+ 

- 

- 

- 

-1- 

- 

— 

± 

+ 

"Caranx" 

+ 

+ 

- 

- 

+ 

- 

- 

— 

+ 

— 

_ 

_ 

-(- 

Chloroscombrus 

+ 

+ 

- 

- 

+ 

- 

- 

- 

+ 

- 

— 

+ 

_ 

Decaplerus 

+ 

+ 

- 

- 

+ 

- 

- 

- 

-1- 

- 

- 

— 

— 

Elagatis 

+ 

0 

+ 

- 

+ 

- 

- 

- 

+ 

- 

- 

+ 

+ 

Gnathanodon 

+ 

+ 

- 

- 

+ 

- 

- 

- 

4- 

_ 

- 

-1- 

-1- 

Hemicaranx 

0 

0 

0 

0 

+ 

- 

- 

- 

+ 

- 

_ 

0 

0 

Lichia 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

Megalaspis 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

Naucrates 

- 

+ 

- 

- 

- 

- 

- 

-1- 

- 

+ 

_ 

+ 

+ 

OUgoplites 

- 

- 

- 

+ 

+ 

- 

- 

- 

+ 

- 

- 

-1- 

+ 

Pamotabus 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

Parastromaleus 

+ 

+ 

- 

- 

- 

+ 

- 

- 

+ 

- 

-t- 

Parana 

+ 

+ 

- 

- 

- 

— 

+ 

— 

■¥ 

_ 

_ 

0 

-f 

Pseudocaranx 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

Scomheroides 

+ 

+ 

- 

- 

+ 

- 

- 

- 

-V 

_ 

— 

-1- 

■f 

Selar 

+ 

+ 

- 

- 

+ 

- 

- 

- 

+ 

- 

_ 

_ 

+ 

Selaroides 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

Selene 

+ 

+ 

- 

- 

+ 

— 

- 

— 

-h 

_ 

— 

+ 

_ 

Seriola 

- 

+ 

- 

+ 

+ 

- 

- 

— 

+ 

- 

— 

+ 

-1- 

Seriolina 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

Trachinotus 

- 

+ 

- 

- 

- 

-1- 

- 

- 

- 

+ 

-V 

^- 

+ 

Trachurus 

+ 

+ 

- 

- 

+ 

- 

- 

- 

-F 

- 

- 

+ 

-1- 

Ulua 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

Uraspis 

0 

0 

0 

0 

+ 

- 

- 

- 

+ 

- 

- 

0 

0 

posterior  margin  of  the  preopercle,  usually  first  appearing  in 
larvae  2.0  to  4.0  mm  NL,  which  coincides  with  yolk  sac  ab- 
sorption. 

Carangids  develop  two  series  of  preopercular  spines,  one  se- 
ries along  the  posterior  margin  of  the  preopercle  and  another 
along  the  anterior  margin,  called  the  "preopercular  crest"  by 
Ahlstrom  and  Ball  (1954).  Both  series  have  an  upper  and  lower 
segment  (Fig.  270).  The  number  of  preopercular  spines  does  not 
seem  to  reach  a  constant  number  as  in  larvae  of  many  other 
fish  families.  Instead,  the  number  of  spines  in  both  series  in- 
creases through  preflexion,  flexion,  and  postflexion  stages  to  a 
maximum  of  usually  about  9  in  the  anterior  and  1 1  in  the 
posterior  series,  then  decreases  in  number  during  transformation 
and  early  juvenile  stages.  Usually  just  prior  to  or  during  the 
early  juvenile  stage,  preopercular  spines  become  completely 
overgrown  by  tissue  and  bone.  Development  of  preopercular 
spines  in  both  the  anterior  and  posterior  series  proceeds  along 
the  margins  away  from  the  angle  of  the  preopercle.  Conversely, 
reduction  in  preopercular  spination  proceeds  toward  the  angle. 
When  spines  are  present  on  either  the  anterior  or  posterior 
margin,  a  spine  is  always  present  at  the  angle  of  the  margin,  and 
it  is  always  the  largest.  The  size  and  shape  of  this  spine  are 
particularly  useful  in  distinguishing  carangid  taxa  (Table  126). 


For  example,  Seriola  zonata  and  OUgoplites  saurus  (Fig.  273C, 
273B)  have  a  preopercular  spine  with  a  spinule(s),  and  Elagatis 
hipinmilata  has  serrated  preopercular  spines  (Fig.  273A). 

A  median  supraoccipital  crest  develops  on  the  head  during 
the  preflexion  stage  in  many  species  (Table  1 26)  and  persists 
until  late  in  the  transformation  stage  when  it  becomes  overgrown 
by  tissue  and  bone.  The  supraoccipital  crest  is  very  useful  in 
distinguishing  carangids  since  there  are  relatively  few  marine 
fish  families  which  have  larvae  with  a  crest.  The  shape  of  the 
supraoccipital  crest  has  been  used  to  distinguish  carangid  larvae 
of  various  taxa  (Aboussouan,  1975),  however,  the  difliculty  in 
defining  shape  characters  makes  them  somewhat  subjective  and 
of  questionable  reliability.  However,  some  taxa,  i.e.,  Elagatis 
hipinmilata  (Fig.  273A)  and  Chloroscombrus  chrysurus  (Fig. 
270D),  have  crests  which  do  appear  quite  distinct  from  those  of 
other  known  carangid  larvae. 

Among  other  head  spines,  supraocular  spines  and  serrations 
develop  in  many  taxa  (Table  126).  The  larger  multiple  supra- 
ocular spines  present  in  Naucrates  diictor  (Fig.  272B)  and  ser- 
rated pterotic  ridge  in  Trachinotus  spp.  (Fig.  272A)  are  notable. 
All  species  develop  posttemporal  and/or  supracleithral  spines 
which  vary  in  number,  usually  1-5,  and  relative  size  among 
taxa  (Fig.  270-274). 


LAROCHE  ET  AL.:  CARANGIDAE 


519 


Table  126.    Extended. 


Vemrolalt 

;rai  pigment 

Aligned 

along 

myosepla 

Internal 
melano- 
phoi^s 
over 
dorsal 
aona 

Lateral 
midline 
pigment 

Melano- 

phores 

on 

branchi- 
ostegal 
mem- 
bi^ne 

In- 
ternal 
melano- 
phores 
over 
noto- 
chord 

Dorsal 
and 
anal 
(inlet 

Melanophi 
dorsal  body 

ares  on 
margin 

Body  pigmentation 

Body  depth 

Dorsal 

fin 

spmes 

elongate 

{fonn 

early) 

Pelvic 

fin 

rays 

first  to 

develop 

Dorsal 
and  anal 
fin  rays 
elongate 
(form 
early) 

Number 

Scat- 
tered 

Anti- 
medial 
rows 

Median 
row 

Shallow 

(A  <  35% 

SL) 

Deep 

(i  >  35% 
SLl 

of 
myomeres 

Dense 

Light 

(typical) 

_ 

_ 

_ 

_ 

_ 

_ 

_ 

+ 

_ 

_ 

+ 

- 

+ 

- 

- 

+ 

24 

+ 

- 

0 

+ 

+ 

0 

- 

0 

0 

+ 

- 

- 

+ 

- 

- 

- 

24 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

24 

_ 

+ 

- 

+ 

+ 

- 

- 

-1- 

- 

- 

-1- 

- 

-1- 

- 

- 

- 

24 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

24 

+ 

- 

0 

+ 

+ 

0 

- 

0 

0 

- 

+ 

- 

-1- 

- 

- 

- 

24 

+ 

+ 

+ 

+ 

+ 

+ 

- 

+ 

+ 

- 

-t- 

- 

+ 

- 

- 

- 

24 

— 

+ 

0 

+ 

- 

0 

- 

0 

0 

- 

+ 

- 

+ 

- 

- 

- 

24 

- 

+ 

+ 

+ 

+ 

+ 

- 

- 

+ 

- 

+ 

- 

+ 

- 

- 

- 

24 

_ 

— 

- 

+ 

- 

- 

+ 

+ 

- 

- 

+ 

+ 

- 

- 

- 

- 

25 

+ 

- 

- 

+ 

+ 

- 

- 

+ 

- 

+ 

- 

+ 

- 

- 

- 

- 

24 

+ 

- 

- 

+ 

+ 

- 

- 

- 

-1- 

+ 

- 

- 

+ 

- 

- 

- 

24 

0 

0 

0 

0 

0 

0 

- 

0 

0 

0 

0 

- 

+ 

- 

0 

- 

26 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

24 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

24 

+ 

- 

- 

+ 

+ 

- 

- 

+ 

- 

+ 

- 

-1- 

- 

- 

- 

- 

25 

+ 

- 

— 

+ 

+ 

- 

- 

+ 

- 

+ 

- 

-1- 

- 

- 

- 

- 

26 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

24 

_ 

- 

- 

+ 

+ 

- 

- 

0 

0 

- 

+ 

- 

-f 

- 

0 

- 

24 

+ 

- 

0 

0 

0 

0 

0 

4- 

- 

- 

+ 

0 

0 

- 

- 

- 

27 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

25 

+ 

- 

- 

+ 

+ 

- 

- 

+ 

- 

-1- 

- 

+ 

- 

- 

- 

- 

26 

_ 

+ 

+ 

+ 

+ 

+ 

-1- 

+ 

— 

— 

+ 

+ 

— 

— 

— 

- 

24 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

24 

- 

- 

+ 

+ 

- 

+ 

- 

-1- 

- 

- 

-1- 

- 

+ 

+ 

+ 

- 

24 

_ 

_ 

+ 

+ 

0 

0 

- 

+ 

- 

-f 

- 

+ 

- 

- 

- 

- 

24 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

24 

+ 

- 

+ 

+ 

+ 

- 

- 

+ 

- 

-1- 

- 

-1- 

- 

- 

- 

- 

24 

+ 

- 

- 

+ 

- 

- 

- 

- 

-1- 

- 

+ 

+ 

- 

- 

- 

- 

24 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

24 

0 

0 

0 

0 

0 

0 

- 

0 

0 

0 

0 

- 

-1- 

- 

0 

- 

24 

Dorsal,  anal,  preanal.  and  caudal  finfolds  are  present  at  hatch- 
ing. Yolk-sac  larvae  rapidly  develop  the  pectoral  fin  base  and 
finfold.  The  sequence  of  fin  formation  in  most  species  is:  caudal, 
pectoral,  anal  and  soft  dorsal,  spinous  dorsal,  and  pelvic.  Species 
oi  Alectis  (Fig.  271  A)  and  Selene  (Fig.  274A)  are  exceptions, 
developing  pelvic  and/or  dorsal  fin  elements  precociously  before 
the  notochord  begins  to  flex.  The  sequence  of  fin  formation  in 
these  taxa  is:  either  pelvic,  spinous  or  soft  dorsal  followed  by 
caudal,  anal,  and  pectoral  (Aprieto,  1974;  Laroche  el  al.,  MS). 

Spinous  dorsal,  soft  dorsal,  and  anal  fins  generally  develop 
from  anterior  to  posterior,  although  the  first  element  in  each  fin 
may  lag,  and  the  most  posterior  element  in  the  soft  dorsal  and 
anal  fins  may  develop  precociously  in  some  species,  i.e.,  De- 
capterus  spp.  and  Selar  crumenophthabnus  (Laroche  et  al.,  MS) 
(Fig.  270B,  C).  In  many  species  at  least  some  dorsal  and  anal 
fin  spines  ossify  from  the  distal  tip  proximally  (Fig.  27 IC,  272C) 
which  may  be  an  unusual  condition  among  marine  fish  larvae 
and  may  help  characterize  the  Carangidae.  This  condition  has 
been  noted  in  a  number  of  species  and  may  occur  in  most  or 
all  species,  however,  lack  of  specimens  in  the  critical  stage  when 
this  condition  is  recognizable  does  not  yet  permit  documenta- 
tion of  its  occurrence.  Pectoral  fin  elements  develop  from  dorsal 


to  ventral.  The  pelvic  spine  develops  before  the  rays,  and  ray 
formation  proceeds  away  from  the  spine. 

The  separation  of  the  two  anteriormost  anal  fin  spines  from 
the  third  spine  by  a  distinct  gap  is  an  important  characteristic 
of  most  young  carangids  once  fins  have  formed.  This  gap  is 
caused  by  anterior  and  posterior  extensions  of  the  distal  part  of 
the  pterygiophores  supporting  the  second  (ultimate)  and  third 
(penultimate)  anal  fin  spines.  This  gap,  although  present,  is  rel- 
atively narrow  in  Elagatis  bipinrndala  and  Seriolina  nigrofas- 
ciata,  which  differ  from  other  carangids  in  having  only  two  anal 
fin  spines.  The  only  other  family  known  to  have  young  with  a 
similar  gap  is  the  Pomatomidae  (Laroche  et  al.,  MS). 

Development  of  an  "antrorse  spine"  on  the  anterodorsal  mar- 
gin of  the  first  dorsal  fin  pterygiophore  (Fig.  272A)  is  another 
character  that  is  found  in  most  young  carangids  following  fin 
development  and  is  shared  by  only  a  few  other  families,  i.e., 
Ephippidae.  This  "spine"  is  usually  covered  with  skin  but  is 
visible  in  larvae  and  juveniles. 

Scales  begin  to  develop  during  the  transformation  stage.  Many 
species  of  carangids  develop  modified  scales  in  the  form  of  scutes 
along  the  posterior  portion  of  the  lateral  line.  Ossifying  scales 
are  usually  first  visible  along  the  straight  part  of  the  lateral  line 
anterior  and  adjacent  to  the  caudal  peduncle,  where  scutes  form. 


520 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES -AHLSTROM  SYMPOSIUM 


y*J^-L' 


y.^*-j 


p^'-A 


n'>*^'%A 


'^^^ 


'.^ 


Fig.  274.     (A)  Early  postflexion  larva  (5.2  mm)  of  Selene  sp.  and  (B)  late  postflexion  larva  (9.2  mm)  of  Hemicaran.x  ainhlyrhynchus. 


Scale  development  proceeds  dorsally,  ventrally,  and  anteriorly 
from  this  location.  Berry  (1960)  presented  a  detailed  account  of 
scute  development  and  methodology  for  making  counts. 

Oi^eo/o^.  —  Developmental  osteology  has  been  described  for 
Trachurns  symmetricus  (Ahlstrom  and  Ball,  1954);  Decapterus 
punctatus.  Elagatis  bipinnulata.  Selene  vomer,  and  Seriola  :on- 
a/a  (Aprieto,  1974);  and  .4/M/r  ^wa/<' (Miller  and  Sumida,  1974). 
The  sequence  of  ossification  is  the  same  for  all  of  these  species. 
The  cleithrum,  premaxilla.  and  posterior  preopercular  angle  spine 
are  first  to  ossify  in  preflexion  larvae.  Although  the  cleithrum 
begins  to  ossify  early,  the  pectoral  and  pelvic  girdles  do  not 
completely  ossify  until  late  in  the  transformation  stage.  Near 
the  beginning  of  notochord  flexion,  the  maxilla,  dentary,  para- 
sphenoid,  supraoccipital,  articular,  frontal,  angular,  and  bran- 


chial arches  begin  to  ossify.  However,  much  of  the  cranium  does 
not  completely  ossify  until  late  transformation  stage.  Teeth  form 
along  the  anterior  margin  of  the  premaxilla  as  soon  as  it  ossifies. 
Aprieto  (1974)  noted  that  early  ossification  of  bones  related  to 
feeding  is  consistent  with  need  for  food  following  yolk  resorp- 
tion. The  first  branchial  arch  begins  to  ossify  first  with  ossifi- 
cation proceeding  from  the  angle  of  the  arch  outward.  The  other 
arches  ossify  similarly  in  sequence.  Gill  rakers  develop  following 
ossification  of  the  element  on  which  they  are  attached.  The  full 
complement  of  gill  rakers  is  not  attained  until  late  transfor- 
mation or  early  juvenile  stage.  Patches  of  small  teeth  form  on 
the  upper  pharyngeals  of  the  third  and  fourth  gill  arches,  and 
the  fifth  arch  has  tooth  patches  along  most  of  its  length.  Pha- 
ryngeal teeth  ossify  early  in  the  postflexion  stage. 

Vertebrae  begin  to  ossify  next,  in  the  middle  of  the  flexion 


LAROCHE  ET  AL.:  CARANGIDAE 


521 


Fig.  275.     (A)  Postflexion  larva  (5.5  mm)  of  Paraslromaleus  niger  and  (B)  small  juvenile  (25.6  mm)  of  I'raspis  secunda. 


522 


ONTOGENY  AND  SYSTEMATICS  OF  HSHES-AHLSTROM  SYMPOSIUM 


stage  (along  with  the  caudal  fin  rays)  in  most  species,  closely 
followed  by  neural  and  haemal  spines.  Vertebrae,  neural,  and 
haemal  spines  ossify  sequentially,  anteroposteriorly.  Centra  os- 
sify from  their  anterior  margin  posteriorly.  Neural  spines  of  the 
abdominal  vertebrae,  and  neural  and  haemal  spines  of  caudal 
vertebrae  begin  to  ossify  before  their  respective  centra.  Ribs 
ossify  at  about  the  same  time  and  also  develop  anteroposte- 
riorly. Pleural  ribs  ossify  before  the  epipleural  ribs.  The  urostyle 
begins  to  ossify  before  the  posteriormost  two  or  three  vertebrae 
during  the  flexion  stage.  Ossification  proceeds  from  its  anterior 
base  towards  its  distal  tip  as  it  also  does  in  the  hypurals. 

Pigmentation.  —  Details  concerning  the  development  and  vari- 
ety of  pigmentation  characters  are  discussed  by  Laroche  et  al. 
(MS)  and  are  summarized  for  genera  in  Table  126.  Although 
many  species  have  not  been  observed  and  this  table  is  tentative, 
it  reflects  the  potential  utility  of  pigmentation  characters. 

It  is  not  possible  to  describe  a  generalized  pigmentation  pat- 
tern that  is  unique  to  and  diagnostic  for  all  carangid  larvae.  By 
the  end  of  the  preflexion  stage,  most  species  have  rows  of  me- 
lanophores  along  the  dorsal  and  ventral  margins  of  the  tail. 
Melanophores  appear  on  the  head  over  the  brain  and  eventually 
form  a  cap  of  pigmentation.  Dorso-  and  ventrolateral  pigmen- 
tation may  be  present  or  absent  depending  on  the  species  (Fig. 
270A,  B).  A  row  of  small  melanophores  develops  along  the 
lateral  midline  at  midbody  during  the  preflexion  stage  and  per- 
sists into  the  juvenile  stage  (Figs.  270-275).  When  these  me- 
lanophores are  expanded,  they  appear  as  a  line  of  pigmentation. 
This  pigmentation  along  the  lateral  midline  has  been  referred 
to  as  the  "lateral  line  streak"  by  Ahlstrom  and  Ball  (1954)  and 
Miller  and  Sumida  (1974).  The  amount  and  pattern  of  mela- 
nistic  pigmentation  on  the  head,  body,  and  fins  of  carangid 
larvae  is  otherwise  quite  diverse,  grading  from  very  light  to  very 
dark  pigmentaton.  However,  larvae  can  usually  be  categorized 
as  either  lightly  or  darkly  pigmented  (Table  126,  Figs.  270-275). 
Darkly  pigmented  forms  usually  have  a  lightly  pigmented  caudal 
peduncle  (Figs.  272,  273). 


Systematic  considerations 

Although  considerable  taxonomic  confusion  still  exists  re- 
garding carangids,  and  developmental  stages  for  most  species 
remain  unknown,  similarities  among  larvae  of  species  assigned 
to  the  same  genus  suggest  a  congruence  between  adult  and  larval 
similarities  which  may  reflect  the  naturalness  of  some  generic 
groups.  For  example,  all  species  of  the  genus  Selene  for  which 
larvae  are  known  share  precocious  development  of  the  spinous 
dorsal,  pelvic,  and  caudal  fins,  while  all  species  of  Decapterus 
for  which  larvae  have  been  described  begin  development  of  a 
finlet  at  the  posterior  of  the  dorsal  and  anal  fins  before  more 
anterior  elements  begin  to  develop.  Interestingly,  Selar  cni- 
menophthalmus  (which  lack  finlets  as  adults)  larvae  also  begin 
development  of  a  fin  element  at  the  posterior  of  the  dorsal  and 
anal  fins  before  more  anterior  elements  begin  to  develop  (Fig. 
270C).  This  character  may  reflect  a  relationship  between  De- 
capterus and  Selar.  This  type  of  information  is  encouraging  and 
may  tend  to  raise  confidence  in  the  naturalness  of  taxonomic 
groups  and  in  the  potential  utility  of  developmental  characters 
for  use  in  systematic  studies  of  carangids. 

Developmental  information  is  available  for  too  few  species 
to  allow  interpretation  of  character  patterns  which  might  reflect 
phylogenetic  relationships  within  the  Carangidae.  Of  course, 
investigation  of  Carangidae's  relationship  to  other  groups  within 
Perciformes  is  a  much  larger  problem  and  will  require  that 
similar  information  be  gathered  for  other  taxa.  Careful,  com- 
parative developmental  studies  are  needed  to  supply  this  critical 
information  and  provide  the  most  direct  route  towards  a  better 
understanding  of  relationships. 

(W.A.L.)  School  of  Natural  Resources,  Department  of 
Fisheries,  Humboldt  State  University,  Arcata,  Cal- 
ifornia 95521;  (W.F.S.-V.)  Department  of  Ichthyology, 
The  Academy  of  Natural  Sciences,  19th  and  The 
Parkway,  Logan  Circle,  Philadelphia,  Pennsylvania, 
19103;  (S.L.R.)  Gulf  Coast  Research  Laboratory,  East 
Beach  Drive,  Ocean  Springs,  Mississippi  39564. 


Carangidae:  Relationships 
W.  F.  Smith- Vaniz 


DESPITE  the  great  economic  importance  and  broad  geo- 
graphic distribution  of  the  Carangidae,  knowledge  of  their 
systematics  is  very  inadequate.  The  few  attempts  to  determine 
their  phylogenetic  relationships  have  been  both  limited  in  scope 
and  methodologically  flawed.  These  classifications  largely  reflect 
the  distribution  of  characters  shared  between  taxa  rather  than 
being  based  on  evolutionarily  derived  characters.  Lack  of  knowl- 
edge of  an  appropriate  out-group  for  comparison  has  also  lim- 
ited progress  in  this  area. 

In  his  pioneering  study  of  carangid  osteology  and  relation- 
ships, Starks  (1911)  recognized  four  subfamilies  but  stressed  the 
difliiculty  of  establishing  intrafamilial  relationships.  Suzuki  ( 1 962) 
described  and  illustrated  the  osteology  of  1 8  genera  of  carangids. 


Unfortunately  only  Japanese  species  were  considered  and,  al- 
though much  useful  descriptive  information  was  presented,  little 
progress  was  made  towards  attaining  a  better  understanding  of 
carangid  phylogeny.  Vergara  (1972)  described  the  osteology  of 
the  Cuban  species  assigned  to  Caran.x  and  presented  a  phyletic 
analysis  of  their  relationships.  In  a  subsequent  paper  Vergara 
(1974)  expanded  his  analysis  to  include  all  Cuban  genera  of 
Carangidae  and  evaluated  the  phenetic  relationships  of  Cuban 
Caran.x.  Smith-Vaniz  and  Staiger  (1973)  concentrated  their  ef- 
forts on  the  Scomberoidini  and  presented  evidence  suggesting 
a  sister-group  relationship  between  Parana  and  Scomheroides  + 
Oligopliles.  The  detailed  comparison  and  osteological  descrip- 
tion of  Nematistius  by  Rosenblatt  and  Bell  (1976)  provided 


SMITH- VANIZ:  CARANGIDAE 


523 


Nematistiidae 

(8)  (1)(2) 

0) 

CO      '" 

(1) 

<D 

-O      CO 

CO 

—     T3 

CO 

T3 
CD 

centr 
haeni 

♦* 

c 

><    a 

CO 

0) 

n     '^ 

E 

^ 

u     L^ 

o 

o 

CO      o 

z 

LLI 

cr   o 

Trachinotini 

Carang 

ni 

Scor"*^'"'"i'^i"i               ' 

1 

1 

1 

(95) 

-D 

(5) 

(1) 

CD 

■o 

(4) 

en 

(20) 

(1) 

CO 

(1) 

(1) 

CO 

E 
E 

C 
CO 

(0 

(1) 

CO 
c 
o 

k_ 
CO 

01 

E 
o 
o 

ligoplite 
rachinot 

(1)      ^ 

.1    i 

(9) 

« 

o 

0) 

CO 

c 

o 

- 

CO 

(0 

ampogrj 
1  other  c 

Q. 

CO 

O 

H 

_l 

z 

w 

w 

LLI 

o 

to 

Carangidae 


"Carangoi(ds" 


Fig.  276.  Hypothesized  cladogram  of  "carangoid"  fishes  including  main  groups  of  Carangidae.  Numbers  opposite  rectangles  correspond  to 
characters  discussed  in  text.  Numbers  in  parentheses  are  estimated  total  number  of  species  m  taxon.  Open  rectangles  are  hypothesized  to  represent 
plesiomorphic  (phylogenetically  primitive)  character  states  and  solid  rectangles  derived  character  states;  characters  indicated  by  barred  rectangles 
are  hypothesized  to  have  evolved  more  than  once,  and  acquired  independently  in  each  lineage  so  marked;  half-barred  rectangles  indicate  that 
both  the  primitive  and  derived  character  states  occur  in  some  component  taxa  of  the  lineage  with  the  derived  condition  secondarily  evolved. 


524 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  127.    Selected  Characters  of  Carangid  Genera.  (Abbreviations:  Triseg.  =  trisegmental;  Br.  =  branchiostegal;  P  =  rayless  pterygio- 

phore.) 


1st  haemal  spine 

Caudal  peduncle 

AM  muscle 

Tnseg 

No.  species 

Scutes 

attachment 

grooves 

A.'div. 

radials 

Trachinotini 

Trachinotus 

(20) 

absent 

strong 

absent 

absent 

absent 

Lichta 

(1) 

absent 

strong 

absent 

absent 

absent 

Scomberoidini 

Parana 

(1) 

absent 

strong 

absent 

absent 

absent 

Scomberoides 

(4) 

absent 

strong 

absent 

absent 

absent 

Oligoplites 

(5) 

absent 

strong 

absent 

absent 

absent 

Naucratini 
Seriola 


(9) 


absent 


weak 


present 


present 


present 


Senolina 

(1) 

absent 

weak 

present 

present 

present 

Elagatis 

(1) 

absent 

weak 

present 

present 

present 

Naucrates 

(1) 

absent 

weak 

present 

present 

present 

Campogramma 

(1) 

absent 

weak 

present,  but 
rudimentary 

present 

absent 

Carangini 

Alectis 

(3) 

present 

strong 

absent 

present 

absent 

Alepes 

(4) 

present 

strong 

absent 

present 

absent 

Atropus 

(1) 

present 

strong 

absent 

present 

absent 

Atule 

(1) 

present 

strong 

absent 

present 

absent 

Carangoides 

(22) 

present 

strong 

absent 

present 

absent 

Caranx 

(14) 

present 

strong 

absent 

present 

absent 

Chloroscombrus 

(2) 

present 

strong 

absent 

present 

absent 

Decapterus 

(10) 

present 

strong 

absent 

present 

absent 

Gnathanodon 
Hemicaranx 
Megalaspis 
Pantolabus 


(1) 

present 

strong 

absent 

(4) 

present 

strong 

absent 

(1) 

present 

strong 

absent 

(1) 

present 

strong 

absent 

present 

absent 

present 

absent 

present 

absent 

present 

absent 

SMITH- VANIZ:  CARANGIDAE 


525 


Table  127.    Extended. 


Ir  rays 

Infenor  vertebral 
foramina 

Epural  bones 

Vertebrae 

'Predorsal  formulae 

7-8 
8 

absent 
present 

3 
3 

10  +   14 
10  +  14 

O/O/O  +  P/l  +  l/ 
0/0/0 +  P/P+1/ 
0/0/0+1/ 

V-VI  +  I,  17-29 

II  +  I,  16-18 

VII  + 1,  19-21 

II  +  I,  17-19 


9 

absent 

3 

10  +  17 

8 

absent 

2 

10  +  16 

7 

absent 

2 

10  +  16 

0/0/0+1/ 

0/0/0+1/ 
0/0/0 +P/1/ 
O/O/O  +  P/l  +  l/ 
0/0/0  + P/P+1/ 
0/0/0 +  P/P/1/ 
0/0/0 +  P/P+P/1/ 


VI  + 1.  32-38 

II  +  I,  34-38 

Vl-VIl  +  I,  19-21 

II  +  I,  16-20 

IV-VI  +  I.  18-21 

II  +  I,  19-21 


absent 

absent 
absent 
absent 
absent 


10  + 

14(6) 

11  + 

13(2) 

11  + 

14(1) 

11  + 

13 

10  + 

14 

10  + 

15 

10  + 

14 

0/0/0+1  +  1/ 
0/0/0  +  2+1/ 
0/0  +  0/1  +  1  +  1/ 
0/0/0/0+1/ 
0/0  +  0/1  +  1/ 
0/0  +  0/1  +  1  +  1/ 
0/0/0/1  +  1/ 

0/0/0+1/ 

0/0  +  0/1  +  1/ 


VII-VIII +  1,22-39 


7 

present 

2 

10  + 

14(2) 

0/0  +  0/1  +  1/ 

10  + 

15-16(1) 

0/0+0/P+l/ 

7 

present 

2 

10  + 

14 

0/0+0/2+1/ 

7 

present 

2 

10  + 

14 

0/0  +  0/2+1/ 

7 

present 

2 

10  + 

14 

0/0  +  0/2+1/ 

7 

present  or 

2 

10  + 

14(21) 

0/0  +  0/2+1/ 

absent  (1) 

10  + 

15(1) 

7 

present 

2 

10  + 
10  + 

14(11) 
15(3) 

0/0  +  0/2+1/ 

7 

present 

2 

10  + 

14 

0/0/0+1  +  1/ 
0/0+0/2+1/ 

7 

present 

2 

10  + 

14(9) 

0/0  +  0/2+1/ 

10  + 

15(1) 

0/0/0/2+1/ 
-/0/0/2+1/ 
-/0/0+0  +  2+1/ 

7 

present 

2 

10  + 

14 

0/0  +  0/1  +  1/ 

7 

present 

2 

10  + 

15(2) 

0/0  +  0/1  +  1/ 

10  + 

16(2) 

0/0  +  0/2+1/ 

7 

present 

2 

10  + 

14 

0/0  +  0/2+1/ 

7 

present 

2 

10  + 

14 

0/0  +  0/2+1/ 

II  +  I, 

15- 

■22 

VII  +  I, 

30- 

■37 

I+I, 

15- 

-18 

VI  +  I, 

25- 

-30 

I+I, 

18- 

-20 

IV-V  +  I, 

25- 

-29 

II+I, 

15- 

-17 

VI-VII  +  I. 

26- 

-28 

II+I, 

23- 

-25 

VI-VII  +  I, 

18- 

-22 

II+I, 

16- 

-20 

VIII  + 1, 

23- 

-27 

II+I, 

18- 

-23 

VIII  +  I, 

19- 

-22 

II+I, 

17- 

-18 

VIII  +  I, 

22- 

-25 

II+I, 

18- 

-21 

VIII  +  I. 

18- 

-35 

II+I, 

16- 

-29 

VIII  +  I, 

18- 

-25 

II+I, 

14- 

-21 

VII-VIII  +  I, 

25- 

-28 

II+I, 

25- 

-28 

VIII  +  I, 

,27- 

-37 

11  +  1,22-32 


VII  +  I.  18-21 

II  +  I,  15-17 

VI-VII  +  I.  20-25 

11  +  1,20-25 

VIII  +  I,  18-20 

II  +  I,  16-17 

VIII  +  I,  21-23 

II  +  I.  18-20 


526 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 

Table  127.    Continued. 


No,  species 

Scutes 

1st  haemal  spine 
attachment 

Caudal  peduncle 
grooves 

AM  muscle 
Ai'div. 

Tnseg. 
radials 

Parastromaieus 

(1) 

present 

strong 

absent 

present 

absent 

Pseudocaranx 

(3) 

present 

strong 

absent 

present 

absent 

Selar 

(2) 

present 

strong 

absent 

present 

absent 

Selaroides 

(1) 

present 

strong 

absent 

present 

absent 

Selene 

(7) 

present 

strong 

absent 

present 

absent 

Trachurus 

(12) 

present 

strong 

absent 

present 

absent 

Ulua 

(2) 

present 

strong 

absent 

present 

absent 

Uraspis 

(3) 

present 

strong 

absent 

present 

absent 

definitive  evidence  supporting  its  removal  from  the  Carangidae. 
Their  work  is  an  important  contribution  towards  elucidating 
carangoid  relationships. 

Much  more  effort  needs  to  be  focused  on  obtaining  basic  data 
on  the  biology,  ontogeny  and  systematics  of  carangids.  The  data 
presented  in  Table  127  and  the  following  discussion  are  a  first 
step  in  that  direction. 

Relationships 

Evidence  supporting  the  monophyly  of  carangoids  and  several 
major  groups  of  carangids  is  discussed  below.  Refer  to  G.  D. 
Johnson  (this  volume)  for  discussion  of  interfamilial  relation- 
ships of  the  three  echeneoid  families.  The  oldest  available  name 
for  each  of  the  four  carangid  tribes  herein  recognized  has  not 
been  determined  but  none  is  proposed  as  new.  No  synapomor- 
phies  were  found  to  support  the  inclusion  of  Lichia  in  the  tribe 
Trachinotini,  and  its  placement  is  one  of  convenience  in  accord 
with  the  practice  of  previous  authors  and  reflects  my  own  sub- 
jective bias.  Autapomorphies  that  define  Trachinolus.  Lichia 
and  the  naucratine  genera  are  not  included  in  the  cladogram 
(Fig.  276)  because  they  are  not  informative  about  relationships. 
These  taxa  are  recognized  individually  in  the  figure  to  make  it 
easier  for  the  reader  to  determine  the  character  state  distribu- 
tions and  the  number  of  species  comprising  each  genus.  The 
Carangini  includes  approximately  20  genera  (Table  127),  many 
not  well  established,  and  their  osteology  poorly  known.  Until 
this  presumably  monophyletic  assemblage  has  been  studied  in 
much  greater  detail  no  meaningful  discussion  of  relationships 
will  be  possible.  Several  recent  authors  have  considered  Para- 
stromateus  to  constitute  either  a  monotypic  family  or  carangid 
subfamily.  Available  evidence  suggests,  however,  that  it  should 
be  assigned  to  the  Carangini. 

The  following  character  states  are  the  basis  for  the  hypotheses 
of  carangoid  interrelationships  inferred  in  Fig.  276.  The  pre- 
sumed derived  character  state  is  listed  first,  followed  by  dis- 
cussion of  the  character  in  out-groups  when  necessary. 

(I)  Freihofer  (1978)  made  the  important  observation  that  in 
the  Nematistiidae,  Carangidae,  Coryphaenidae,  Rachycentridae 


and  Echeneididae  there  are  one  or  two  tubular  ossifications 
(prenasals)  around  the  anterior  extension  of  the  nasal  canal.  This 
presumed  specialization  of  the  lateralis  system  is  very  rare  in 
percoids  (also  present  in  the  unrelated  and  highly  specialized 
Toxotidae)  and  is  considered  to  be  a  synapomorphy  suggesting 
that  these  five  families  constitute  a  monophyletic  group. 

(2)  The  possession  of  small  adherent  cycloid  scales  is  a  derived 
character  shared  by  carangoids  in  contrast  to  the  typically  cte- 
noid scales  of  most  other  percoids.  Berry  (1969)  reported  that 
the  carangid  Elagatis  has  "ctenoid"  scales  and  Zheng  (1981) 
also  described  the  highly  modified  caudal  peduncle  scales  of 
Naucrates  as  ctenoid.  These  scales  are  not  typically  "ctenoid" 
and  appear  to  represent  modifications  of  the  carangoid  scale- 
type. 

(3)  Two  separate  prenasal  canal  units,  one  membranous  and 
one  bony  (Carangidae)  or  both  bony  (echeneoids).  In  contrast, 
Nematistiits  has  only  a  single  prenasal  canal  unit. 

(4)  Loss  of  the  bony  stay  (Fig.  277)  posterior  to  ultimate  dorsal 
and  anal  pterygiophores  that  is  present  in  most  other  percoids 
(see  Table  127,  G.  D.  Johnson,  this  volume). 

(5)  On  shoulder  girdle,  middle  part  of  coracoid  with  its  an- 
terior margin  consisting  of  a  lamella  of  bone  broadly  extending 
towards  the  median  cleithral  wing  (Suzuki,  1962:  figs.  36-44). 
In  Nematistius  the  middle  and  lower  parts  of  the  coracoid  are 
rodlike  with  lamellar  bone  restricted  to  its  posterior  margin 
(Rosenblatt  and  Bell,  1976;  fig.  8). 

(6)  Basioccipital  with  a  pair  of  foramina  (Rosenblatt  and  Bell, 
1976:  fig.  3)  into  which  anterior  processes  of  the  gas  bladder 
extend  forward  to  the  region  of  the  inner  ear. 

(7)  Anterior  shift  of  second  predorsal  bone  to  the  first  inter- 
neural  space  and  first  pterygiophore  greatly  expanded  and  plate- 
like. In  carangids,  as  in  most  other  percoids,  the  second  pre- 
dorsal bone  occupies  the  second  iniemeural  space  (predorsals 
absent  in  echeneoids),  and  in  both  echeneoids  and  carangids  the 
first  dorsal  pterygiophore  is  not  greatly  expanded. 

(8)  Spines  of  first  dorsal  fin  very  long  and  filamentous  and 
only  basally  connected  by  interradial  membrane. 

(9)  Tubular  ossifications  surrounding  both  prenasal  canal  units; 


SMITH- VANIZ:  CARANGIDAE 

Table  127.    Continued.    E.xtended. 


527 


Br.  rays 


Infenor  vertebral 
foramina 


Epural  bones 


'Predorsal  formulae 


Anal  tin 


7 

present 

2 

10  + 

14 

0/0+0/1  +  1  +  1/ 

7 

absent 

2 

10  + 

14(2) 

0/0+0/2+1/ 

10  + 

14- 

15(1) 

7 

present 

2 

10  + 

14 

0/0  +  0/2+1/ 

7 

present 

2 

10  + 

14 

0/0  +  0/2+1/ 

7 

present 

2 

10  + 

14 

0/0  +  0/2+1/ 

7 

present 

2 

10  + 

14 

0/0+0/2+1/ 

7 

present 

2 

10  + 

14 

0/0  +  0/2+1/ 

7 

present 

2 

10  + 

14 

0/0  +  0/2+1  +  1/ 

IV-V,41-44 


II  +  I, 
VIII  +  I, 


35-39 

23-28 


II  +  I, 
VIII  +  I, 


20-24 
23-27 


II  +  I, 
VIII  +  I, 


19-22 
24-26 


II  +  I, 
VIII  +  I, 


21-23 
20-24 


II  +  I, 
VIII  +  I, 


16-20 
28-36 


II  +  I, 
VIII  +  I, 


24-32 
21-22 


II  +  I, 
VIII  +  I. 


17-18 

24-32 


II  +  I,  17-28 


'  Methodology  of  predorsal  formulae  after  Ahlslrom  el  al.  (1976). 

posterior  canal  unit  unossified  in  carangids  and  absent  in  Ne- 
matistiidae. 

(10)  Absence  of  the  so-called  beryciform  foramen  in  the  an- 
terior ceratohyai. 

(11)  Absence  of  predorsal  bones. 

(12)  Several  anal  pterygiophores  anterior  to  the  first  haemal 
spine  (versus  one  in  Carangidae,  Nematistiidae  and  most  per- 
coids). 

( 1 3)  Larvae  very  elongate,  with  dorsal  fin  ray  development 
not  completed  until  two  or  three  times  size  at  flexion  (G.  D. 
Johnson,  this  volume).  In  contrast,  larvae  of  carangids  are  mod- 
erate to  deep-bodied  and  complete  dorsal  and  anal  fin  devel- 
opment in  conjunction  with  or  soon  after  flexion  (Laroche  et 
al..  this  volume).  Larvae  of  Nematistiidae  are  unknown. 

( 1 4)  Posteroventral  elongation  of  first  proximal  pterygiophore 
of  anal  fin  resulting  in  a  relatively  wide  gap  (Fig.  278b-e)  be- 
tween the  last  two  anal  spines.  The  carangid  genera  Elagatis 
and  Seriolina  (Fig.  278c)  are  exceptional  in  having  only  one 
spine  on  this  pterygiophore  so  the  gap  is  not  as  apparent. 

(15)  Presence  of  a  separate  dorsal  division  (A,')  of  the  ad- 
ductor mandibulae  muscle  originating  on  the  suborbital  shelf 
(Fig.  279).  The  relative  size  of  the  suborbital  shelf  in  carangids 
is  not  correlated  with  the  presence  or  absence  of  this  muscle, 
which  is  also  lacking  in  echeneoids,  Nematistius  and  most  per- 
coids. 

(16)  Some  lateral  line  scales  (at  least  those  on  caudal  peduncle) 
modified  as  thickened  scutes. 

( 1 7)  Caudal-peduncle  grooves  present  dorsally  and  ventrally; 
these  specialized  structures  undoubtedly  have  a  hydrodynamic 
function  related  to  swimming  mode.  Campogramma.  which 
appears  to  be  the  most  advanced  naucratine  (judging  from  the 
relatively  large  number  of  autapomorphic  characters  that  it  pos- 
sesses), is  exceptional  in  having  only  rudimentary  caudal-pe- 
duncle grooves  (absent  in  young). 

The  occurrence  of  caudal-peduncle  grooves  on  Nematistius 
which  shares  many  plesiomorphic  characters,  including  a  sim- 
ilar external  morphology,  with  naucratines  is  most  parsimoni- 
ously interpreted  as  parallelism.  These  structures  are  also  pres- 
ent on  carcharhinid  sharks. 


(18)  Premaxilla  non-protractile  and  in  adults  dorsal  margin 
of  upper  lip  attached  to  snout  by  a  broad  frenum. 

(19)  Epiotics  broadly  united  along  midline  of  cranium  beneath 
the  supraoccipital. 

(20)  Total  vertebrae  26  or  27  (versus  24  or  25). 

(21)  Cheeks  unsealed. 

(22)  Spines  of  dorsal  and  anal  fins  with  well  developed  venom 
glands  (Halstead  et  al,  1972;  Sazima  and  Uieda,  1979). 

(23)  First  proximal  pterygiophore  of  anal  fin  expanded  an- 
terolaterally  to  form  a  roof  over  anal  spines  (Smith-Vaniz  and 
Staiger,  1973:  fig.  15b). 

(24)  Juveniles  with  two  widely  spaced  rows  of  dentary  teeth 


Fig.  277.  Terminal  pair  of  dorsal  fin  rays  and  associated  pterygio- 
phores: (a)  Nemalislius  pecloralis  (Note  presence  of  large  bony  stay 
behind  last  medial  pterygiophore)  and  (b)  Naucrates  duclor. 


528 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES -AHLSTROM  SYMPOSIUM 


Nematistius 


Caranx 


Parastromateus 


Fig.  278.     Anterior  pterygiophores  and  associated  spines  and  rays  of  anal  fin  (Note  relative  spacing  between  last  two  spines);  (a)  Nemalislius 
pectoralis;  (b)  Senola  zonala;  (c)  Seriolina  nigrofasciata;  (d)  Caranx  sexfasciatus;  (e)  Parastromateus  niger. 


into  which  the  premaxiliary  teeth  fit  when  the  mouth  is  closed, 
and  the  outer  series  of  dentary  teeth  strongly  hooked  outward 
and  with  spatulate  tips.  Major  ( 1 973)  has  shown  that  this  dental 
arrangement  facilitates  lepidophagous  feeding  in  juvenile  Sconi- 
heroides  and,  on  the  basis  of  stomach  content  analyses  of  two 
species,  concluded  that  at  least  some  Oligoplites  have  similar 
feeding  habits.  Carr  and  Adams  (1972)  postulated  that  inten- 
tional removal  of  ectoparasites  is  also  an  important  activity  in 
juvenile  Oligoplites.  Presumably  such  unique  dentition  facili- 
tates both  types  of  specialized  trophic  ecology. 

(25)  Interosseous  space  between  coracoid  process  of  dentary 
and  posterodorsal  projection  of  anguloarticular  minute  or  ab- 
sent. 


(26)  Pleural  ribs  on  vertebrae  3  through  7  or  8  attached  high 
on  centrum  and  spatulate  in  cross-section. 

(27)  Posterior  dorsal-  and  anal-fin  rays  consisting  of  semi- 
detached finlets. 

(28)  Reduction  in  number  of  epurals  in  caudal  fin  from  3  to  2. 

(29)  Supramaxilla  minute  or  absent.  It  might  be  argued  that 
the  reductive-loss  supramaxilla  character  state  is  a  synapomor- 
phy  uniting  Trachtnotits  +  Lichia  with  the  Scomberoidini,  in 
which  case  the  well  developed  supramaxilla  of  Parana  would 
constitute  a  reversal.  Alternatively,  the  reductive  trend  of  the 
supramaxilla  in  the  two  taxonomic  pairs  under  consideration 
might  be  a  simple  case  of  parallelism.  In  the  absence  of  any 
other  obvious  synapomorphy  that  supports  the  first  hypothesis 


SMITH-VANIZ:  CARANGIDAE 


529 


Fig.  279.     Adductor  mandibulae:  (a)  Seriola  diiinerili;  (b)  Caran.x  sexfasciatus.  Note  the  presence  of  A,',  a  separate  dorsal  section  originating 
on  the  suborbital  shelf  (S). 


and  because  the  reversal  of  a  reductive  trend  is  involved,  I 
believe  it  is  more  conservative  (even  though  less  parsimonious) 
to  retain  the  unresolved  position  of  Trachinolus-Lichia  in  the 
cladogram. 

(30)  Pelvic  fins  absent  at  all  stages  of  development. 

(31)  Increase  in  number  of  caudal  vertebrae  from  16  to  17. 

(32)  Branchiostegal  rays  9  (versus  7  or  8). 

(33)  Basibranchial  dentition  consisting  of  large  median  tooth 
plates,  presumably  derived  from  fusion  of  the  large  paired  tooth 
plates  found  in  Scomhcroides  and  Oligoplitcs  (Smith-Vaniz  and 
Staiger.  1973:  figs.  24b-d). 

(34)  Lateral  line  with  5-9  dorsal  branches. 

(35)  Loss  of  dorsal-fin  spines  resulting  in  an  increase  in  the 
number  of  rayless  pterygiophores  (see  Table  127). 

(36)  Loss  of  mesopterygoid  teeth. 

(37)  Loss  of  supramaxilla  (minute  in  Scninberoides). 

(38)  Loss  of  suborbital  shelf  on  third  infraorbital  bone. 


(39)  Infraorbitals  2-4  enlarged  and  extending  posteriorly  across 
cheek  in  adults. 

(40)  Prominent  dark  spots  or  short  bars  on  sides  of  adults. 
Unlike  many  carangids,  the  juveniles  of  both  Scomberoides  and 
Oligoplites  are  unbarred. 

Recognition  of  the  family  Nematistiidae 
The  familial  placement  oi  Nematistius  has  long  been  contro- 
versial. Some  distinguished  ichthyologists  (Gill.  1863;  Jordan 
and  Evermann,  1896-1900;  Berg.  1947)  placed  it  in  a  separate 
family  while  others,  most  recently  Robins  et  al.  ( 1 980).  assigned 
it  to  the  Carangidae.  On  the  basis  of  a  detailed  osteological 
comparison.  Rosenblatt  and  Bell  (1976)  concluded  that  Ne- 
matistius should  not  be  classified  with  the  Carangidae.  They 
also  commented  on  the  striking  similarities  between  the  Ne- 
matistiidae and  certain  primitive  carangids,  especially  naucra- 
tine  genera.  Almost  all  of  the  many  features  shared  by  these  two 


530 


ONTOGENY  AND  SYSTEMATICS  OF  nSHES-AHLSTROM  SYMPOSIUM 


taxa  are  plesiomorphic  character  states,  the  one  notable  excep- 
tion being  caudal-peduncle  grooves. 

In  addition  to  possessing  different  character  states  3-8  and 
14  as  listed  above,  Freihofer  (1963)  observed  that  the  Caran- 
gidae  and  Nematistiidae  differ  in  the  course  of  the  nerves  of  the 
ramus  lateralis  accessorius  (RLA)  complex;  the  former  having 
pattern  9  and  the  latter  pattern  10  (reduced).  Nematistius  also 
differs  in  having  two  foramina  in  the  scapula;  a  typically  large 
one  and  a  smaller  more  posteriorly  positioned  foramen  (absent 
in  carangids)  that  also  occurs  in  the  Rachycentridae.  Like  the 
two  RLA  nerve  patterns,  the  derived  character  state  for  the  two 
scapular  foramina  conditions  has  not  been  determined.  Never- 
theless, the  inclusion  of  Nematistius  in  the  Carangidae  would 
make  the  family  paraphyletic  (unless  the  three  echeneoid  fam- 
ilies are  also  included)  and  impossible  to  define  based  on  shared 
derived  characters. 

Familial  position  of  Parastromateus 

Several  recent  authors  have  followed  Apsangikar  (1953)  or 
Suzuki  ( 1 962)  in  recognizing  Parastromateus  either  as  a  subfam- 
ily of  the  Carangidae  or  as  the  sole  representative  of  the  mono- 
typic  Formionidae  (=Apolectidae  or  Parastromatidae).  All  the 


characters  used  to  justify  the  latter  classification,  with  one  ex- 
ception discussed  below,  have  been  autapomorphic  characters 
which  can  provide  no  information  about  relationships.  That  the 
genus  should  be  assigned  to  the  Carangidae  is  clearly  indicated 
by  the  possession  of  derived  character  states  3-5  and  14-16 
discussed  previously. 

Haedrich  (1971)  noted  that  Parastromateus  (=Apolectus)  is 
the  only  fish  with  a  pattem-9  ramus  lateralis  accessorius  nerve 
system  that  has  a  "pons  moultoni."  In  an  addendum  to  his 
paper  it  was  suggested  that  retention  of  the  pons  is  a  primitive 
character  state.  It  should  be  emphasized  that  very  few  carangoids 
have  been  examined  for  the  presence  of  this  easily  overlooked 
structure.  Until  the  distribution  of  this  character  has  been  de- 
termined for  the  major  lineages  of  carangoids,  its  phylogenetic 
significance  can  not  be  evaluated.  Similarly,  no  data  have  been 
presented  to  substantiate  assigning  Parastromateus  to  its  own 
subfamily  within  the  Carangidae. 

Department  of  Ichthyology,  The  Academy  of  Natural  Sci- 
ences, 19th  and  The  Parkway,  Logan  Circle,  Phila- 
delphia, Pennsylvania  19103. 


Mugiloidei:  Development  and  Relationships 
D.  P.  DE  Sylva 


MUGILOIDEI  is  one  of  three  closely  related  suborders,  to- 
gether with  Sphyraenoidei  and  Polynemoidei,  in  the  Per- 
ciformes.  The  suborder  is  represented  by  a  single  family,  the 
Mugilidae.  Until  recently,  the  Atherinidae  had  been  considered 
close  relatives  of  the  Mugiloidei.  Within  the  family  Mugilidae, 
classical  morphological  taxonomic  analyses  have  been  applied 
to  regional  groupings  rather  than  to  the  family  as  a  whole  (Weber 
and  de  Beaufort,  1922;Roxas,  1934;  Smith,  1935,  1947;Schultz, 
1946;  Ishiyama,  1951;  Thomson,  1954;  Ebeling,  1957,  1961; 
Lindberg  and  Legeza,  1969;  Ben-Tuvia,  1975).  Hence,  the  sys- 
tematics  of  the  family  are  poorly  understood. 

Mullets  are  characterized  by  thick,  streamlined  bodies,  deeply 
forked  caudal  fin.  large  cycloid  or  weakly  ctenoid  scales,  and 
the  lack  of  a  lateral  line.  The  mouth  is  small,  the  jaws  have 
small  teeth  or  none,  and  the  gill  rakers  are  long  and  slender,  the 
latter  assisting  the  pharyngeal  jaw  apparatus  to  form  a  filtering 
apparatus  (Lauder  and  Liem,  1983).  They  share,  with  the  thread- 
fins  and  barracudas,  the  characteristic  of  having  two  widely 
separated  dorsal  fins.  Two  subfamilies  of  mullets  are  recognized, 
the  Mugilinae  and  the  Agonostominae  (Jordan  and  Evermann, 
1896-1900).  The  latter  have  sessile  teeth  which  attach  directly 
to  the  jaws,  a  flat  preorbital,  and  only  2  anal  spines  in  the  adult. 
The  Mugilinae  have  flat  labial  teeth,  if  any,  connected  to  the 
jaws  by  elongated  fibers,  a  ridged  and  grooved  preorbital,  and 
3  anal  spines  in  the  adult. 

The  Mugilinae  occur  worldwide  except  in  polar  regions,  while 
the  Agonostominae  are  confined  to  Central  America,  the  west- 


em  Indian  Ocean,  the  tropical  west  Pacific,  and  the  Australian 
coastline.  Mullets  occur  in  oceans,  bays,  estuaries,  and  fresh 
water.  They  are  uniformly  important  as  food  for  humans  and 
an  important  prey  in  the  food  web.  They  seldom  exceed  1  meter. 

Development 

Many  studies  exist  on  the  eggs,  larvae,  and  post-larval  stages 
of  mullets  in  comparison  to  other  families,  but  only  a  few  are 
comprehensive,  and  most  deal  with  a  single  species  (e.g.,  An- 
derson, 1957;Dekhnik,  1973;  Farrugio,  1977;  Kuo  et  al..  1973; 
Lai.  1979;  Martm  and  Drewry,  1978;  Sanzo,  1936;  Tung,  1973; 
Vialli,  1937;  Yang  and  Kim,  1962;  Yashouv  and  Bemer-Sam- 
sonov,  1970).  However,  a  general  overview  of  each  stage  can 
be  summarized. 

Eggs  are  pelagic,  spherical,  and  transparent,  with  the  surface 
of  the  egg  being  smooth  and  usually  without  sculpture  (Fig.  280). 
The  yolk  is  unsegmented.  the  perivitelline  space  is  narrow,  and 
there  is  one  or  more  oil  globules.  During  development,  several 
oil  globules  merge  with  each  other,  becoming  situated  on  the 
yolk  sac  upon  hatching.  Egg  sizes  for  various  species  of  European 
and  African  mugilids  range  from  0.6  to  1 .3  mm  and  vary  greatly 
in  diameter  from  one  geographic  area  to  another.  Although  most 
eggs  have  similar  pigmentation,  different  species  have  similar, 
though  sometimes  overlapping  spawning  seasons,  which  may 
offer  a  clue  in  the  analysis  of  phyletic  relationships  and  mugilid 
evolution. 

Larval  pigmentation  ranges  from  relatively  light  to  heavy  (Fig. 


DE  SYLVA:  MUGILOIDEI 


531 


Fig.  280.  Various  stages  of  development  of  eggs  of  silver  mullet.  Mugil  curema:  (a)  unfertilized  eggs;  (b)  2  hours  after  fertilization  (32  blastomeres); 
(c)  4  hours  after  fertilization  (blastodisc  well  formed,  cells  small);  (d)  8  hours  after  fertilization  (segmentation  cavity  forming);  (e)  12  hours  after 
fertilization  (early  embryo);  (0  16  hours  after  fertilization  (embryo);  (g)  24  hours  after  fertilization  (lateral  view  of  embryo);  (h)  24  hours  after 
fertilization  (top  view  of  embryo);  (i)  32  hours  after  fertilization  (lateral  view  of  embryo)  (from  Anderson,  1957). 


281).  All  larvae  have  stellate  melanophores  on  the  oil  globule, 
which  also  occur  on  the  forehead  of  some  species.  This  feature 
has  not  been  studied  for  mugilids  on  a  global  basis,  but  offers 
possibilities  for  phyletic  analysis. 

At  hatching,  stellate  melanophores  also  occur  on  the  yolk 
surface  and  body,  with  fine  spots  along  the  dorsal  and  ventral 
profile  of  the  caudal  trunk.  The  caudal  rays  form  first,  at  4  mm 
total  length.  The  second  dorsal  forms  at  between  4  and  5.7  mm. 


and  the  first  dorsal  forms  at  5.4  mm.  Scales  begin  to  develop 
at  between  8  and  10  mm  and  are  well  formed  at  1 1  mm.  Pig- 
mentation is  strong  at  from  2  to  5  mm,  and  the  dorsal  surface 
is  dark  by  5  mm  total  length.  By  a  length  of  8.2  to  10.9  mm 
they  are  silvery  white  to  silvery  green,  and  at  this  size  they 
resemble  the  adults  in  body  form,  there  being  no  distinctive 
metamorphosis  throughout  development  (Fig.  281). 

Identification  of  later  larvae  (Fig.  282)  is  based  upon  color. 


532 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  281.     Larvae  of  silver  mullet.  Mugil  curema.  (A)  Newly  hatched,  1.76  mm;  (B)  yolk-sac  stage,  2.15  mm;  (C)  yolk-sac  stage,  2.47  mm;  (D) 
yolk-sac  stage,  2.56  mm;  (E)  yolk-sac  stage,  2.56  mm;  {¥)  3.7  mm;  (G)  4.0  mm;  (H)  4.7  mm;  (I)  5.3  mm.  From  Anderson  (1957). 


pigmentation  pattern,  number  of  anal  elements,  longitudinal 
scales,  transverse  scales,  scale  morphology,  pyloric  caeca,  and 
gill  rakers.  The  general  profiles  of  the  head,  lips,  and  the  labial 
and  lingual  teeth  are  also  very  useful  characters  (Tung,  1973; 
Wallace  and  van  der  Elst,  1973;  Thomson,  1975;  Lai,  1979; 
Zisman,  1982). 

Relationships 

In  some  species  such  as  Mugil  cephalus,  as  presently  under- 
stood, which  has  a  worldwide  distribution,  there  is  considerable 
variability  in  meristic  characters  and  proportional  measure- 
ments. Additional  studies  are  warranted  to  determine  the  real 
extent  of  genetic  exchange  between  local  subunits  (Thomson, 
1982). 

At  the  generic  and  specific  levels,  mugilid  taxonomy  has  not 
been  resolved.  As  in  the  case  of  Mugil  cephalus,  those  species 
with  extensive  ranges  may  be  known  under  different  names  in 
various  parts  of  their  range. 

A  variety  of  external  morphology  features  have  been  used  to 
identify  genera  and  species  of  the  adult  stages,  ranging  from 


dentition  (Ebeling,  1957,  1961;  Farrugio.  1977)  and  scales 
(Thomson,  1982),  to  eye  coloration  (Alvarez-Lajonchere,  1975). 
Internal  anatomy  is  valuable  in  systematic  analysis,  including 
the  shape  and  number  of  pyloric  caeca  (Perlmutter  et  al.,  1957; 
Luther,  1975b),  the  alimentary  tract  (Thomson,  1966),  intestinal 
convolution  (Hotta,  1955),  osteology  (Luther,  1975a;  Mohsin, 
1978;  Kobelkowsky  and  Resendez,  1972;  Sunny,  1971;  Hotta 
and  Tung,  1972),  and  otoliths  (Morovic,  1953). 

Phyletic  studies  within  the  family  have  not  been  undertaken. 
Thomson's  manuscript  revision  (see  Thomson,  1982)  recog- 
nizes 14  genera  and  64  species  of  the  nominal  282  species.  Of 
these,  32  are  indeterminate  because  of  inadequate  descriptions 
or  missing  holotypes.  The  only  published  world  revision,  by 
Schultz  (1946),  recognizes  13  genera.  Relationships  are  based 
upon  the  adipose  eyelid,  type  of  scales,  labial  characteristics, 
preorbital  shape,  and  type  of  habitat.  Larval  mullets  have  been 
studied  extensively,  but  not  on  a  worldwide  basis,  and  no  phy- 
letic analysis  has  been  attempted.  It  is  known  that  in  certain 
species  the  young  stages  have  2  anal  spines,  but  larger  stages 
have  3  spines.  The  younger  stages  have  been  referred  to  as  the 


DE  SYLVA:  MUGILOIDEI 


533 


Fig.  282.  Postflexion  larvae  (A,  B)  and  juvenile  (C)  of  silver  mullet. 
Mugil curema.  (A)  7.0  mm;  (B)  14.5  mm;  (C)  25.5  mm.  From  Anderson 
(1957). 


"querimana  stage."  An  analysis  of  the  genera  and  species  pos- 
sessing this  trait  has  not  been  undertaken.  Biochemical  studies 
on  mugilid  systematics  have  been  undertaken  by  Callegarini 
and  Basaglia  (1978)  and  by  Autem  and  Bonhomme  (1980)  in 
the  Mediterranean,  but  no  studies  have  been  carried  out  on  a 
worldwide  basis. 

As  stated  in  the  discussions  on  Sphyraenoidei  and  Polyne- 
moidei  (this  volume),  they  have  been  closely  linked  with  the 
Mugiloidei  phyletically.  Previously,  the  athennids  had  been 
placed  within  this  assemblage,  but  Rosen  ( 1 964)  has  clearly 
shown  that  the  atherinids  belong  in  a  separate  superorder  con- 
taining the  flyingfishes  and  livebearers.  The  Mugiloidei  appear 
more  closely  related  osteologically  to  the  Sphyraenoidei  than 
they  are  to  the  Polynemoidei. 


A  brief  history  of  the  higher  classification  of  these  groups  is 
reviewed  here.  The  suborder  Percesoces  had  included  the  Ath- 
erinidae.  Mugilidae,  and  Sphyraenidae  (Jordan  and  Evermann, 
1868-1900),  but  Starks  (1900)  questioned  their  similanty,  though 
he  believed  them  to  be  quite  close  based  upon  the  decided 
branching  of  the  epiotic  crests.  Superficially,  the  mugiloid-sphy- 
raenoid  skeleton  resembles  that  of  atherinoids,  but  Hollister 
(1937)  pointed  out  an  important  developmental  difference  be- 
tween them.  In  Athehna.  the  lowermost  hypural  plate  develops 
as  a  single  entity.  In  Mugil  and  Sphyraena  this  plate  forms  from 
two  distinct  elements.  Berg  ( 1 940)  separated  the  Mugilidae,  with 
the  Sphyraenidae  and  the  Atherinidae,  from  the  Perciformes  as 
the  order  Mugiliformes  because  they  have  abdominal  pelvic  fins, 
a  relatively  primitive  character.  Rosen  (1964)  also  pointed  out 
the  similarities  among  mugiloids,  sphyraenoids,  and  polyne- 
moids  in  ossification  of  the  skull,  especially  the  common  pres- 
ence of  a  subocular  shelf,  the  jaw  suspension  and  feeding  mech- 
anism, jaw  musculature,  and  the  pharyngobranchial  and 
opercular  apparatuses.  Further,  Rosen  stated  that  "the  embryos 
of  mullet  (Anderson,  1957)  and  barracuda  (Orton,  1955b)  are 
small  and  contain  a  large  oil  globule  ....  A  forward-displaced 
heart  is  also  characteristic  of  Oryzias  .  .  .  but  not  of  Sphyraena 
(Orton,  1955b;Shojimaetal.,  1957),  and  probably  not  of  A/i/.g//." 

Removal  of  the  Mugilidae  from  the  suborder  Percoidei  is 
supported  by  studies  of  blood  plasma  and  plasma  proteins  (Sul- 
ya  et  al.,  1960).  Plasma  proteins  of  mugilids  are  less  complex 
than  those  of  any  other  family  considered  to  be  Perciformes, 
and  show  relationships  to  some  species  of  Cypriniformes  and 
Clupeiformes  (Gunter  et  al..  1961).  In  contrast,  plasma  proteins 
of  some  species  of  3  perciform  families,  the  Carangidae,  Sciaeni- 
dae,  and  Scombridae,  do  not  differ  greatly  from  those  of  the 
Mugilidae.  Based  on  this,  the  Mugilidae  could  be  regarded  either 
as  belonging  to  the  most  primitive  perciform  group  or  as  branch- 
ing from  some  early  perciform. 

The  early  life  history  stages  do  not  appear  to  offer  useful  hints 
as  to  phyletic  relations  with  other  taxa,  except  that  the  Mugi- 
loidei have  23  myotomes  dunng  larval  development,  a  feature 
shared  with  the  Polynemoidei  and  Sphyraenoidei. 


RosENSTiEL  School  of  Marine  and  Atmospheric  Science, 
University  of  Miami,  4600  Rickenbacker  Causeway, 
Miami,  Florida  33149. 


Sphyraenoidei:  Development  and  Relationships 


D.  P.  DE  Sylva 


SPHYRAENIDAE  is  a  closely  knit,  monogeneric  perciform 
family  of  the  suborder  Sphyraenoidei  (Gosline,  1971), 
Schultz  (1953)  revised  the  family,  which  has  since  been  partially 
modified  by  Smith  ( 1 956b),  Williams  ( 1 959),  and  de  Sylva  (1975) 
for  Indian  Ocean  species.  Six  genera,  including  three  new  names, 
were  proposed  by  J.  L.  B.  Smith  in  his  1 956  review  of  the  Indian 
Ocean  species.  These  have  been  synonymized  by  subsequent 
authors  to  include  the  single  genus  Sphyraena,  recognized  for 
all  living  species.  Fossil  genera  have  been  noted  in  the  Creta- 
ceous and  are  widespread  since  the  lower  Eocene.  These  are 
represented  by  the  genera  Sphyraenodus,  Protosphyraena,  Pro- 
sphyraena.  and  Sphyraena  (see  de  Sylva,  1963).  However,  be- 
cause most  fossil  generic  descriptions  are  based  only  upon  teeth 
or  dentary  fragments,  it  seems  presumptive  to  attach  very  great 
importance  to  the  validation  of  such  genera.  In  a  draft  revision 
of  the  family,  I  have  recognized  the  genera  named  by  Smith,  as 
well  as  other  genera  previously  proposed  for  other  sphyraenids, 
at  the  subgeneric  level  to  clarify  phyletic  relationships  on  a 
worldwide  basis  (Fig.  283;  Table  128). 

All  species  are  tropical  or  temperate,  and  are  schooling  or 
solitary  predators.  They  usually  live  in  the  littoral  zone  from 
the  surface  to  just  off  the  bottom  in  shelf  waters.  Several  are 


epipelagic  and  are  found  far  from  shoal  water.  They  are  im- 
portant food  fishes,  although  one  species,  Sphyraena  barracuda. 
is  frequently  responsible  for  ciguatera  poisoning  (de  Sylva,  1 963). 
Maximum  size  is  180  cm  and  48  kg. 

There  are  20  valid  species  of  the  69  nominal  species.  Sphy- 
raenids are  distinguishable  from  Polynemidae  and  Mugilidae 
by  their  well-developed  fang-like  teeth,  large  mouth,  and  point- 
ed snout,  with  the  upper  jaw  not  protrusible.  Gill  rakers  may 
be  absent,  bristle-like,  or  limited  to  one  or  two  at  the  angle  of 
the  gill  arch  (de  Sylva,  1975). 

Development 

Eggs  of  Sphyraenidae  have  been  described  for  only  3  species, 
and  they  are  similar  in  size  and  pigmentation.  Larval  stages 
have  been  described  for  5  (Raffaele,  1 888;  Bamhart,  1927;  Vial- 
h,  1956;  Orton,  1955b;  Shojima  et  al.,  1957;  Mannaro,  1971; 
Uchida  et  al.,  1958;  de  Sylva,  1963;  Houde,  1972b).  Larval 
stages  have  been  described  for  4  of  the  20  species,  from  rea- 
sonably complete  developmental  series  (e.g..  Figs.  284-287). 
Osteological  development  of  the  neurocranium  is  described  for 
only  1  species  (Gregory,  1933),  while  the  caudal  skeleton  and 
urophore  complex  have  been  studied  for  only  3  species  (Hoi- 


CO 

•H 

u   u 

u 

■H     QJ 

o 

C    i-^ 

O    -H 

(0 

Q.    QJ 

<U 

3 

M 

CO   ^ 

CO 

^-^ 

'i 

J^ 

nj 

•1—) 

■H 

■H 

w 

CD 

nJ 

-a 

J-i 

3 

nj 

DC 

•H 

w 

c 

QJ 

C 

c 

TO 

c 

CO 

■H 

CO 

CO 

to 

O 

^ 

0 

OO 

(d 

T) 

QJ 

4J 

c 

c 

rH 

CO 

•H 

CO 

x: 

OJ 

a> 

4J 

CO 

>,. 

-hI 

3 

CO 

CO 

c 

CO 

rH 

CO 

w 

nJ 

m  .H 

m 

3 

(J 

to 

(U 

to 

-3 

tn 

x: 

cd 

4-1 

W 

■H 

u 

o 

c 

OJ 

QJ 

•H     rH 

3 

(U 

3 

•H 

T-\ 

CJ 

3 

•H 

u 

u 

o 

3 

p. 

■l-\ 

x: 

to 

4-1 

u 

rH    -H 

CU 

to 

to 

e 

X3 

CO 

u 

3 

nJ 

•H 

tfi 

4J 

••-i 

s^ 

(U 

en 

cn 

C 

CO  -a 

T3 

V* 

to 

x: 

2 

O 

CU 

•H 

CO 

ftO 

1-1 

> 

>, 

J^ 

4-1 

14-1 

CO 

to 

CO 

QJ 

<U      3 

•H 

>1 

•H 

o 

3 

i-H 

•H 

CO     Q- 

U 

p 

Xi 

CO 

u 

o 

3 

CO 

> 

u 

■H 

an 

U     U 

U 

x: 

CO 

to 

■u 

f-A 

3 

J-i  ^— ' 

u 

■H 

r-H 

^ 

u 

0 

3 

■o 

u 

O   -H 

•H   dJ 

3 

3 

3 

QJ 

S^ 

<4-l 

CO 

a. 

1 

^ 

o 

I 

(0 
1 

1           1 

a 

XI     Q. 

>    tn 

at 

m|            a.[ 

■r-l 

CO 

X> 

\ 

\ 

(subgen. 

nov) 

/ 

\ 

(Australuzza) 

/ 

(Sphyraenella) 
3D,    4D,    5B.    \ 
7E,    8A,                           \ 

2a\3C,    4A, 

5A, 

(Sphyra 
7G,         \ 

2na) 

3A,    AA,    5B, 
7D,    8C 

\ 

/ 

(Agrioposphyraena) 

\ 

2B,    3A,    5B, 

!,    2B 

5A, 

7E?\ 

y 

^ 

8B 

6, 

\ 

\ 

3A, 

\ 

\ 

V 

\ 

\ 

8A, 

9A, 

lOA 

\ 

\ 

\ 

\ 

y 

Sph 

yraena    forster 

i 

\ 

X 

y 

7A,    8C, 

lOA 

( 

\ 

Callosphyraena) 

y 

S 

ph 

yraeni 

dae 

Fig.  283.     Diagram  of  relationships  among  sphyraenids  based  on  adult  and  larval  characters.  Numbers  refer  to  characters  listed  in  Table  128. 
Labelled  horizontal  lines  cross  branches  and  demonstrate  presumed  advanced  character  states. 


534 


DE  SYLVA:  SPHYRAENOIDEI 


535 


i     ^-"4 


Fig.  284.  Developmental  stages  of  Sphyraena  borealis  reared  in  the 
laboratory.  (A)  3.8  mm;  (B)  4.3  mm;  (C)  5.3  mm;  (D)  7.4  mm  (from 
Houde,  1972b). 


VJ.V- 


Fig.  285.  Developmental  stages  of  Sphyraena  borealis.  Specimens 
A,  B,  and  C  were  laboratory  reared;  specimen  D  was  collected  in  a 
plankton  net.  (A)  9.4  mm  SL;  (B)  12.3  mm  SL;  (C)  14.5  mm  SL;  (D) 
21.0  mm  SL(from  Houde,  1972b). 


Fig.  286.  Drawings  showing  changes  in  pigmentation  and  body  form 
with  larval  development  in  Sphyraena  barracuda.  (A)  5.5  mm  SL;  col- 
lected by  R/S  DANA,  Station  1293-V,  17°43'N,  64°56'W,  April  17, 
1922.  (B)  6.6  mm  SL;  collected  by  R/S  DANA,  Station  952,  17°55'N, 
64°48'W,  May  12,  1921.  (C)  8.6  mm  SL;  collected  by  R/S  DANA, 
Station  1352-V,  35°42'N,  73°43W,  May  21,  1922.  (D)  11.9  mm  SL; 
collected  by  Donald  P.  deSylva,  1  mile  southwest  of  the  harbor  entrance 
of  North  Bimini,  Bahamas,  June  6,  1956  (from  de  Sylva,  1963). 


lister,  1937;  Monod,  1968).  Development  of  Sphyraena  is  di- 
rect, with  no  metamorphosis  (Vialli,  1956;  de  Sylva,  1963; 
Houde,  1972b). 

Meristic  characters  are  not  especially  valuable  in  differen- 
tiating most  adult  species  of  this  family.  Although  little  work 
has  been  done  on  larval  meristic  characters,  it  would  be  expected 
similarly  that  they  would  not  prove  valuable.  Anal  rays  vary 
from  8  to  9,  and  the  dorsal  secondary  rays  of  the  caudal  fin  vary 
from  9  to  10  in  two  different  subgenera. 

Similarly,  morphological  characteristics  do  not  differ  widely 
in  the  early  life  history  of  the  species  except  that  two  groups 
can  be  broadly  identified— those  with  blunt  heads  and  more 
fusiform  bodies,  such  as  S.  barracuda  (de  Sylva,  1963)  (Figs. 
286,  287)  and  those  with  more  slender  heads  and  having  fleshy 
tips  on  the  lower  jaw  and  a  more  slender,  tapering  body,  as  in 
S.  sphyraena  and  S.  borealis  (ViaWi,  1956;  Houde,  1972b)  (Figs. 
284,  285). 

Adult  species  are  distinguished  by  the  shape  and  angle  of  the 
teeth,  number  of  lateral  line  scales,  opercular  and  preopercular 
bone  configuration,  lateral  pigment  pattern,  dorsal  fin  place- 
ment, and  kinds  of  gill  rakers. 

In  5.  barracuda,  adult  characters  are  acquired  over  a  size  range 
of  from  5.5  to  2 1 3  mm.  Pigmentation  is  acquired  gradually  from 
about  5.5  mm  to  24  mm,  then  rapidly  above  that  size. 

In  S.  barracuda,  the  caudal  fin  forms  first  followed  by  the 


536 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  128.    Characteristics  of  Sphyraenidae.  (+  =  occurs  in  this  species; 


■  does  not  occur  in  this  species;  0  =  no  information.) 


Sphyracna 
forslen 


S 
pingiiis 


S. 
Ilavicauda 


S  chryso- 
taenia 


S 
heilen 


S  aculi- 
pinnis 


S  novae-  S.  S.  S. 

hollandiae  tucasana         idtasles    argentea 


1.  Meristic 

Lat.  Hne  scales 

2.  Maximum  length  (mm),  SL 

3.  Gillrakers 

a.  Absent 

b.  Occur  as  spinules 

c.  One 

d.  Two 

4.  Lower  jaw 

a.  With  fleshy  knob 

b.  Without  fleshy  knob 

5.  Dorsal  fin 

a.  Above  pelvics 

b.  Behind  pelvics 

6.  Scales 

a.  Cycloid 

b.  Ctenoid 

7.  Adult  pigment  (lateral) 

a.  Axial  spot  beneath  pectoral  fin 

b.  Vertical  bars 

c.  Vertical  bars  festooned 

d.  Chevrons  angled  forward 

e.  Stripes  (one  or  two) 

f    Inky  blotches  on  lower  sides 
g.    No  lateral  markings 

8.  Teeth 

a.  Conical,  widely  spaced 

b.  Flattened,  erect,  contiguous 

c.  Flattened,  angled  backward, 

contiguous 

Larval  characters 

9.  a.    L.  jaw  with  fleshy  knob 

b.   L.  jaw  without  fleshy  knob 

10.  a.   Well-marked  pigmentation 
b.   Poorly  developed  pigmentation 


112-123 
640 


88-92 
350 


84-88 
320 


85-96 
231 


120-128 
800 


122-128 
434 


130-155       126-137        145        166 
500  467  530       907 


+ 


+ 


+  + 


+  +  + 


0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

+ 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

Fig.  287.  Drawings  showing  changes  in  pigmentation  and  body  form  transformations  in  Sphyraena  barracuda.  (A)  17.2  mm  SL;  collected  by 
Donald  P.  de  Sylva,  1  mile  southwest  of  the  harbor  entrance  of  North  Bimini,  Bahamas,  May  12,  1956.  (B)  23.7  mm  SL;  collected  by  Donald 
P.  and  Doris  D.  de  Sylva,  at  beach  east  of  Lemer  Marine  Laboratory,  Bimini,  Bahamas,  July  7,  1956.  (C)  213  mm  SL;  collected  by  David  K. 
Caldwell,  Spanish  Harbor  Key,  Monroe  County,  Florida,  June  7,  1956;  University  of  Florida  No.  7072.  (D)  790  mm  SL;  collected  by  Doris  D. 
de  Sylva,  north  of  North  Bimim,  Bahamas,  25°48'N,  79°17'W,  July  18,  1956  (after  de  Sylva,  1963). 


DE  SYLVA:  SPHYRAENOIDEI 
Table  128.    Extended. 


537 


s  s 

horealis  puuditla 


S. 
sphyraena 


S 

vthdensts 


S  S 

guachancho  putnamtae 


S.  lello 


S.  genie 


S-  afra 


S 

barracuda 


115-130        110-120         120-135         137-140         108-116         108-110         129-131         130-140         120-130        122-140        80-90  1 

450  400  1,370  540  470  600  873  1,250  1.150  1.720  1,650  2 


+ 


+ 


+ 


+ 


+ 


:  1 


6 


}       7 


0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

10 


second  dorsal  fin,  and  then  the  pectoral  and  anal  fins  (de  Sylva, 
1963).  By  6.6  mm,  the  second  dorsal,  artal,  and  pectoral  fins  are 
fully  ossified  (Fig.  286).  By  1 1.9  mm  the  first  dorsal  and  pelvic 
fins  have  developed.  Middorsal  and  midventral  pigmentation 
is  well  developed  at  9  mm,  and  is  useful  in  differentiating  among 
larval  stages. 

Juveniles  of  most  species  are  unknown,  and  characters  used 
to  separate  adult  species  would  be  expected  to  be  the  most  useful, 
especially  pigment  patterns. 

Relationships 

Most  sphyraenids  in  museums  have  been  misidentified.  The 
revision  of  the  family  in  the  Indian  Ocean  by  Williams  (1959) 
has  greatly  clarified  the  identification  of  several  important  Indo- 
Pacific  species  whose  identification  rests  largely  upon  the  pattern 
of  vertical  bars  or  chevrons,  festoons  (Figs.  288-289),  gill  raker 
characteristics,  relative  eye  size,  or  upon  the  relative  position 
of  the  first  dorsal  fin  (de  Sylva,  1975).  The  lack  of  any  analysis 
of  the  family  based  upon  osteology,  scale  morphology  (see  Bleek- 
er,  1854-1857),  or  internal  anatomy  precludes  an  exhaustive 
analysis  of  this  family.  Electrophoresis  and  functional  enzymic 


evolution  has  related  the  phylogeny  of  four  eastern  Pacific  sphy- 
raenids to  evolutionary  temperatures  (Graves  and  Somero,  1982) 
and  offers  much  promise  for  analysis  of  other  fishes.  As  pre- 
viously mentioned,  fossil  sphyraenids  are  so  incompletely  de- 
scribed that  they  shed  little  light  on  phyletic  affinities. 

Because  the  larvae  of  only  5  of  the  20  species  have  been 
described,  almost  nothing  can  be  deduced  about  the  phylogeny 
of  the  family  based  on  larval  characters. 

Sphyraenids  were  placed  by  Starks  ( 1 900)  in  the  subqrder 
Percesoces,  together  with  the  Mugilidae  and  the  Atherinidae. 
This  is  based  essentially  upon  their  widely  separated  dorsal  fins, 
elevated  pectoral  fins,  and  the  decided  branching  of  the  epiotic 
crests.  Hollister  (1937)  pointed  out  that  while  the  mugiloid- 
sphyraenoid  skeleton  superficially  resembles  that  of  atherinoids, 
there  was  an  important  difference  in  the  development  of  the 
hypural  plates.  Rosen  (1964)  placed  the  Atherinoidae  in  a  sep- 
arate order,  the  Atheriniformes.  Greenwood  et  al.  (1966)  rec- 
ognized the  Atheriniformes  as  a  superorder,  the  Atherinomor- 
pha,  based  upon  distinctive  habits  or  morphological  peculiarities, 
and  placed  in  the  superorder  Acanthopterygii  the  suborders  Mu- 
giloidei,  Sphyraenoidei,  and  Polynemoidei. 


538 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  288.  Variation  in  lateral  pigmentation  in  various  species  of  Sphyraena.  (A)  Sphyraena  idiasles.  21  cm,  Galapagos  Islands;  (B)  Sphyraena 
acutipinnis.  27  cm.  Hong  Kong;  (C)  Sphyraena  novaehollandiae.  43  cm,  Kapingamarangi,  Caroline  Islands;  (D)  Sphyraena  chn'sotaenia,  18  cm, 
South  Africa;  and  (E)  Sphyraena  flavicauda.  33  cm.  Strait  of  Jubal,  Red  Sea.  (All  drawn  by  J.  I.  Godfrey.) 


DE  SYLVA.  SPHYRAENOIDEI 


539 


"^^ 


■-^^^ 


^'^^::^^~ 


Fig.  289.  Variation  in  pigmentation  in  various  species  of  Sphyraena.  (A)  Sphyraena  genie.  29  cm,  Makassar,  Indonesia;  (B)  Sphyraena 
barracuda,  63  cm,  Biscayne  Bay,  Miami,  Rorida;  (C)  Sphyraena  forsleri,  5.9  cm,  Indonesia,  lateral  view;  (D)  Sphyraena  jorsteri,  5.9  cm.  Indonesia, 
dorsal  view;  and  (E)  Sphyraena  putnamiae.  45.7  cm,  Mahe,  Seychelles  Islands.  (All  dravm  by  J.  I.  Godfrey.) 


540 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES -AHLSTROM  SYMPOSIUM 


No  other  close  relatives  of  the  Sphyraenidae  have  been  dis- 
closed, although  it  has  been  postulated  that  the  Australian  sea 
pike,  family  Dinolesthidae,  is  an  early  offshoot.  However,  Eraser 
(1971)  critically  compared  the  internal  anatomy  of  the  two  fam- 
ilies and  concluded  that  their  apparent  similarity  is  a  result  of 
convergent  evolution. 

Larval  characters  of  the  Sphyraenidae  do  not  show  any  ob- 
vious similarity  to  either  the  Mugilidae  or  the  Polynemidae. 
There  are  only  two  illustrated  accounts  of  larval  Polynemidae 
(Aboussouan,  1966d;  Kowtal,  1972),  neither  of  which  discusses 
familial  relationships.  Superficially,  polynemid  larvae  resemble 


the  phyletically  distant  Sciaenidae.  Nor  do  the  Mugilidae  re- 
semble the  Sphyraenidae  in  the  larval  stages.  Undoubtedly  there 
are  similarities  in  the  larval  development  of  the  hypural  complex 
in  the  Mugilidae  and  Sphyraenidae,  but  I  am  unaware  of  any 
published  material  on  this.  The  question  of  whether  the  poly- 
nemids  should  be  grouped  within  the  Mugiloidei  and  Sphyrae- 
noidei  is  still  unresolved. 

RosENSTiEL  School  of  Marine  and  Atmospheric  Science, 
University  of  Miami,  4600  Rickenbacker  Causeway, 
Miami,  Florida  33149. 


Polynemoidei:  Development  and  Relationships 
D.  P.  DE  Sylva 


POLYNEMIDAE  is  the  only  family  of  the  suborder  Poly- 
nemoidei, containing  37  species,  most  of  which  are  Indo- 
Pacific.  Seven  genera  have  been  recognized:  Galeoides  Giinther, 
Filimanus  Myers,  Pentanemus  Giinther,  Polynemus  Linnaeus, 
Polydactylus'LaQtptdt,  Polistonemus G'\\\,  and  Eleutheronemus 
Bleeker  (see  Norman,  1930).  This  is  a  shallow-water  group 
dwelling  on  sand  or  mud  bottoms,  frequently  in  turbid  water. 
Most  are  common  in  tropical  brackish  environments,  and  some 
species  enter  rivers.  They  are  important  commercial  fishes,  es- 
pecially in  the  Indo-Pacific,  where  some  species  reach  2  meters. 
The  threadfins  resemble  mullets  (Mugilidae),  but  the  snout  is 
pointed  and  overhangs  the  large  mouth,  and  the  eyes  are  rather 
large  (Fig.  290).  The  feature  distinguishing  them  from  their  close 
relatives,  the  barracudas  and  the  mullets,  is  seen  in  their  4  to  7 
pectoral  rays  which  are  detached  from  the  rest  of  the  pectoral 
fin.  Polynemids  also  differ  from  mugilids  by  having  a  lateral 
line,  absent  in  mugilids,  which  extends  onto  the  caudal  fin. 
Polynemids  are  distinguished  from  sphyraenids  by  the  absence 
of  fang-like  chopping  teeth  and  the  rather  blunt,  terminal  mouth 
characteristic  of  the  Mugilidae.  With  the  mullets  and  the  bar- 
racudas they  share  the  characteristic  of  2  widely  separated  dorsal 
fins.  The  maxillary  attachment,  shape  of  the  preopercle,  length 
and  number  of  pectoral  filaments,  tooth  development,  and  de- 
velopment of  the  lower  lip  are  important  taxonomic  characters. 

Development 

Little  is  known  about  the  eggs  and  larvae  of  the  Polynemidae 
in  comparison  to  the  Mugilidae.  Eggs  have  been  obtained  through 
artificial  fertilization  of  Polydactylus  se.xfilis  in  Hawaiian  aqua- 
culture  ponds  (Morris  and  Kanayama,  1964-1969;  Lowell,  1971; 
Rao,  1977),  but  illustrations  of  the  egg  and  larval  stages  have 
not  been  published.  Larval  stages  of  the  Indian  species  Eleu- 
theronema  tetradactylum  from  India  show  developmental  stages 
from  egg  to  5.5  mm  (Sarojini  and  Malhotra,  1952;  Kowtal, 
1972).  The  small  egg,  which  averages  0.76  mm,  has  a  large  oil 
globule.  In  the  smallest  larva  descnbed  (3.8  mm),  caudal  fin 
development  has  started.  Some  rays  appear  in  the  caudal  fin  at 
4.7  mm,  and  melanophores  occur  on  the  maxillary  symphysis 
and  upper  side  of  the  pectoral  fin  bud.  A  related  African  species, 
Galeoides  polydactylus  from  Senegal,  shows  little  development 
of  the  dorsal  fin  at  2.7  mm  (Aboussouan,  1966d).  The  head  is 


relatively  large,  with  a  very  large  eye,  and  23  myotomes  can  be 
seen  (Fig.  291);  they  resemble  sciaenids.  Pigmentation  is  weak, 
in  contrast  to  the  Mugilidae,  except  for  some  melanophores  on 
the  opercle,  anal  fin  base,  and  gut.  By  4.3  to  4.4  mm,  the  two 
dorsal  fins  and  their  rays  have  formed.  At  the  largest  size  de- 
scribed, 7.6  mm,  pigmentation  occurs  around  the  opercular  se- 
ries and  posterior  trunk,  the  pectoral  filaments  are  forming,  and 
the  mouth  is  distinctly  inferior.  No  special  larval  characters 
occur  in  this  group,  and  development  is  direct  and  without  any 
peculiar  metamorphosis. 

Relationships 

No  modem  phyletic  analysis  has  been  undertaken  to  delineate 
the  relationships  among  the  7  genera.  The  only  revision  of  the 
family  is  by  Gill  (1862).  The  characters  which  separate  them 
from  one  another  are  the  extent  of  maxillary  attachment,  shape 
of  the  preopercle,  length  and  number  of  pectoral  filaments,  and 
development  of  the  teeth  and  lower  lip.  Except  for  the  number 
of  pectoral  filaments,  those  characters  at  best  offer  weakly  qual- 
itative differences  useful  in  identifying  species  rather  than  gen- 
era. 

Early  life  history  stages  shed  little  light  on  relationships  among 
members  of  the  Polynemidae.  Of  the  37  species,  larval  stages 
have  been  illustrated  for  only  2  species.  Osteological  studies  on 
the  axial  skeleton  have  been  carried  out  on  6  species,  based 


lower  pectoral 
fin  rays 


Fig.  290.     Major  features  of  the  family  Polynemidae  (from  Allen,  1981). 


DE  SYLVA:  POLYNEMOIDEI 


541 


Table  129.    Comparison  of  Meristic  Characters  of  Mwg;/ and. 4go- 

nostomus  (Mugilidae),  Polydaclylus  (Polynemidae),  and  Sphyraena 

(Sphyraenidae)  from  the  Western  Atlantic  Ocean  (Data  from 

Miller  AND  JoRGENsoN,  1973). 


Fig.  29 1 .  Larvae  of  the  polynemoid.  Galeoides  polydaclylus.  (A)  2.75 
mm;  (B)  3.13  mm;  (C)  4.3  mm;  (D)  4.4  mm;  (F)  7.6  mm.  From  Abous- 
souan(1966cl). 


upon  adult  specimens  (Marathe  and  Bal.  1958).  No  studies  of 
the  external  or  internal  anatomy  have  been  undertaken  on  any 
polynemid. 

The  suborder  Percesoces  was  established  by  Starks  (1900)  to 
show  the  close  relationships  among  the  families  Atherinidae. 
Mugilidae,  and  Sphyraenidae.  To  this  group  Tate  Regan  (1929) 
added  the  Polynemidae,  based  upon  the  well-developed  cranial 
crests,  the  position  of  the  exoccipitals  and  basioccipitals,  the 
alisphenoid  juncture,  the  poor  development  of  the  parapoph- 
yses,  and  the  24  vertebrae  shared  with  the  Sphyraenidae.  Based 
upon  extensive  osteological  evidence,  Gosline  ( 1 962)  concluded 
that  the  Polynemidae,  Sphyraenidae,  Atherinidae,  and  Phallo- 
stethoidei  are  more  closely  related  to  one  another  than  to  other 
fish  groups,  and  placed  them  in  a  separate  order,  Mugiliformes. 
He  did.  however,  show  that  the  Polynemidae.  Sphyraenidae, 
and  Mugilidae  were  more  closely  related  to  each  other  based  on 
the  similar  number  of  vertebrae,  the  postcleithral  strut,  and  the 
possession  of  nonadhesive  eggs,  than  to  the  Alhennidae  and 
Phallostethoidei.  The  pelvic  morphology  of  the  polynemids  and 
sphyraenids  is  so  primitive  as  to  suggest  that  these  groups  could 
not  have  arisen  from  any  advanced  percoid  groups,  and  that 
they  must  be  derived  from  a  very  low  level  of  percoid. 


No.  of  elemenls 

Mugd 

Agonos- 
tomus 

dactylus 

Sphyraena 

Vertebrae 

24 

25 

24 

24 

Precaudal 
Caudal 

12 
12 

12 
13 

10 
14 

12 

12 

First  count  dorsal  fin 
Second  dorsal  fin 

5 
7-8 

5 
8 

8-9 
10-12 

6 
9 

Anal  fin 

3,8-9 

2,  10 

3,  12-13 

2,9 

Total  caudal  elements 

28-29 

32-34 

41-43 

35 

Dorsal  secondary 
Dorsal  pnmary 
Ventral  pnmary 
Ventral  secondary 

7 

7 

7 

7-8 

9-10 

7 

7 
9-10 

12-13 

9 

8 
12-13 

9-10 
9 
8 
9 

Gosline  (1971)  removed  the  Phallostethoidei  from  the  sub- 
order Mugiloidei  (the  old  Mugiliformes),  but  concluded  that  the 
superfamily  Atherinoidae  belonged  in  this  suborder  together 
with  the  superfamilies  Polynemoidae,  Sphyraenoidae,  and  Mu- 
giloidae.  However,  Rosen  (1964)  had  removed  the  atherinoids 
from  the  percesocine  group  and  had  established  them  as  part 
of  a  new,  separate  order,  the  Atheriniformes,  a  practice  followed 
widely  today.  Thus,  the  Polynemoidei,  Sphyraenoidei,  and  Mu- 
giloidei have  no  relatives  closer  to  them  than  they  are  to  each 
other.  These  are  presently  recognized  as  separate  suborders  within 
the  order  Perciformes. 

There  is  no  salient  feature  in  the  early  life  history  which  relates 
the  Polynemoidei  to  other  taxa.  The  development  of  the  eggs 
and  larvae  of  Polynemoidei,  Mugiloidei.  and  Sphyraenoidei  seem 
to  follow  approximately  the  same  pattern,  and  all  have  pelagic 
eggs.  However,  a  major  departure  of  the  Polynemoidei  from  the 
other  two  is  that  one  species,  Polydaclylus  sexfilis.  is  a  protan- 
drous  hermaphrodite.  It  matures  first  as  a  male  at  a  fork  length 
of  about  20  to  29  cm,  and  then  transforms  into  the  female  at 
between  30  and  40  cm  following  a  hermaphroditic  stage  (San- 
terreandMay,  1977).  As  far  as  is  known,  none  of  the  Mugiloidei 
or  the  Sphyraenoidei  is  ever  hemaphroditic. 

Comparisons  of  meristic  characters  offer  some  noteworthy 
data  (Table  129).  The  vertebral  count  of  the  Polynemoidei  is 
10  +  14  =  24;  the  other  two  suborders  have  a  count  of  12  -I- 
12  =  24.  The  number  of  dorsal  and  anal  elements  of  the  Mu- 
giloidei and  Sphyraenoidei  resemble  each  other  more  closely 
than  they  do  the  Polynemoidei.  The  vertebral  formula,  as  well 
as  the  number  of  dorsal  and  anal  elements,  are  more  closely 
related  to  the  Gerreidae.  In  fact,  the  habits  of  the  Polynemoidei 
closely  resemble  those  of  the  Gerreidae.  To  my  knowledge,  there 
is  nothing  published  on  the  eariy  life  history  of  the  Gerreidae 
which  might  disclose  any  similarities  to  the  Polynemidae. 

The  Polynemoidei  (i.e..  Polydaclylus)  have  a  higher  number 
of  first  dorsal,  second  dorsal,  anal,  and  caudal  elements  than 
the  other  groups  (Table  129).  However,  a  companson  on  a 
worldwide  basis  is  required  before  such  an  analysis  can  reveal 
phyletic  relationships. 

RosENSTiEL  School  of  Marine  and  Atmospheric  Science, 
University  of  Miami,  4600  Rickenbacker  Causeway, 
Miami,  Florida  33149. 


Labroidei:  Development  and  Relationships 
W.  J.  Richards  and  J.  M.  Leis 


THE  most  recent  concept  of  this  group  concludes  that  the 
Pomacentridae,  Cichhdae,  Embiotocidae,  and  Labridae 
comprise  a  monophyletic  assemblage  (Kaufman  and  Liem.  1982). 
Kaufman  and  Liem  (1982)  include  the  Odacidae  and  Scaridae 
in  the  expanded  family  Labridae.  For  present  purposes,  we  em- 
ploy the  traditional  view  of  three  separate  families.  Pomacen- 
tridae is  a  large  primarily  marine  family  of  about  23  genera  and 
230  species  found  in  the  tropics  and  warm  temperate  waters  of 
the  world's  oceans  (Allen,  1 975a).  Cichhdae  is  a  fresh  and  brack- 
ish water  family  found  in  the  Americas,  Africa  including  Mad- 
agascar, coastal  western  Mediterranean,  and  the  coastal  areas 
of  India.  It  is  a  large  family  comprised  of  about  85  genera  and 
perhaps  700  species  making  it  the  second  largest  perciform  fam- 
ily (Stiassny,  1981).  Embiotocidae  is  found  only  in  the  North 
Pacific  with  2  species  around  Japan  and  Korea.  18  off  the  west 
coast  of  the  LI.S.,  and  1  confined  to  freshwater  of  California 
(Tarp,  1 952).  Labridae  is  a  tropical  and  warm  temperate  marine 
family  of  about  58  genera  and  about  400  species  (Russell,  1 980). 
Odacidae  is  a  temperate  marine  group  of  4  genera  and  1 2  species 
confined  to  New  Zealand  and  southern  Australia  (M.  F.  Gomon 
and  J.  R.  Paxton,  pers.  comm.).  Scaridae  is  a  tropical  marine 
family  of  about  12  genera  and  70  species  (Schultz,  1958).  Table 
1 30  summarizes  meristic  characters  of  labroid  fishes. 

Development 

The  family  Embiotocidae  is  a  small  family  of  viviparous  species 
that  has  several  unusual  morphological  specializations  during 
development  as  reviewed  by  Wourms  (1981).  During  gestation, 
the  vertical  fins  hypertrophy  and  develop  spatulate  extensions, 
and  the  alimentary  canal  hypertrophies,  especially  the  hind  gut. 
All  these  specializations  appear  due  to  viviparity  and  are  not 
treated  further  here. 

Cichhdae,  so  far  as  known,  all  undertake  elaborate  parental 
care  (Breder  and  Rosen,  1966).  The  eggs  are  slightly  elliptical 
or  irregularly  shaped.  The  eggs  are  also  adhesive  except  for  those 
which  are  orally  incubated.  There  is  a  vast  literature  on  repro- 
ductive behavior  most  of  which  describes  spawning  behavior 
and  parental  care,  but  there  is  little  descriptive  information  on 
larvae  since  many  species  transform  directly  from  yolk  sac  to 
juvenile  (Balon,  1981b;  Noakes  and  Balon,  1982).  Balon  (1959) 
described  the  young  of  Cichlasoma  cyanoguttalum.  The  larvae 


of  laboratory  reared  Cichlasoma  octofascialum  are  depicted  in 
Fig.  292.  The  larvae  of  Pterophy/him  have  an  adhesive  disk  on 
the  head  for  attachment  to  substrate  and  several  stages  are  de- 
picted in  photographs  in  Innes  (1956).  The  Cichlasoma  larvae 
(Fig.  292)  have  unusual  structures  on  the  head  though  they  were 
not  observed  to  be  used  as  holdfast  organs  (A.  W.  Kendall, 
pers.  comm.).  Larvae  of  Symphysodon  cling  to  the  mucus  of  the 
parent  and  actually  derive  nourishment  from  it  (Breder  and 
Rosen,  1966).  Balon  (1977)  thoroughly  describes  the  develop- 
ment of  Labeotropheus,  a  mouth  brooder  which  has  direct  de- 
velopment. 

Pomacentridae  have  demersal  eggs  with  an  adhesive  pedestal; 
the  male  guards  and  incubates  them.  Few  species  have  been 
studied  from  an  early  life  history  perspective  (Table  131).  Most 
have  pelagic  larvae,  but  at  least  one  species  (Acanthochromis 
polyacantha)  broods  and  protects  the  young  in  a  manner  similar 
to  cichlids  (Robertson,  1973).  Larval  development  is  direct  with 
few  larval  specializations  and  no  specialized  stages  between  lar- 
vae and  juveniles.  The  sequence  of  fin  formation  is  variable. 
All  fins  may  be  formed  as  early  as  3  mm,  but  depending  on 
species,  settlement  may  not  occur  until  1 8  mm.  The  gut  is  coiled 
at  hatching.  The  larvae  arc  very  similar  to  percoids  and  may 
be  easily  confused  with  numerous  families  (Leis  and  Rennis, 
1983).  In  general  they  have  a  short,  coiled,  triangular  gut,  an 
inconspicuous  gas  bladder  which  is  covered  by  melanophores, 
and  weak  preopercular  spination  (Fig.  293). 

Some  early  life  history  information  is  available  on  about  one- 
half  of  the  labrid  genera  (Table  1 30).  The  vast  majority  of  labrids 
spawn  small  (0.5-1.1  mm)  pelagic  eggs,  but  three  northeast 
Atlantic  genera  have  adhesive,  demersal  eggs  with  parental  care 
(Table  131).  Demersal  labrid  eggs  are  small  (<  1  mm)  and  ad- 
hesive, but  do  not  have  an  adhesive  pedestal.  Labrid  eggs  usually 
have  a  smooth  chorion  and  a  single  oil  globule.  Newly  hatched 
larvae  have  the  yolk  sac  protruding  anteriorly  in  front  of  the 
head  with  the  oil  globule  (if  present)  at  the  anteriormost  postion. 
The  larvae  are  generally  elongate  and  laterally  compressed  with 
a  deep  caudal  peduncle,  but  some  species  are  deep-bodied  (Fig. 
294).  The  gut  is  rugose  and  is  initially  straight;  coiling  may  be 
delayed  until  after  flexion  in  some  species.  The  head  is  com- 
pressed and  almost  always  lacks  spines.  Scales  do  not  form  prior 
to  settlement.  In  tropical  forms  the  eye  may  be  round,  ovoid. 


Table  130.    Some  Meristic  CHAEtACTERS  of  Labroid  Fishes.  N  is  the  approximate  number  of  recent  species  largely  after  Nelson  (1976).  Other 
data  from  Gunther,  1862;  Boulenger,  1915;  Tarp.  1952;  Miller  and  Jorgensen,  1973;  Russell,  1980;  Leis  and  Rennis,  1983;  Sanchez,  1981;  and 

J.  R.  Paxton,  pers.  comm. 


N 

D 

A 

F 

' 

Pj 

Venebrae 

Cichlidae 

700 

IX-XXV,  3 

-31 

III 

-XIII,  6-28 

_ 

_ 

1,5 

24-39 

Embiotocidae 

23 

VII- 

XVIII, 

9-28 

III 

-IV,  13-35 

17- 

-29 

1,5 

31-42 

Labridae 

400 

VIII 

-XX,  5 

-15 

III 

-V, 6-14 

11- 

-21 

1,5 

23-40 

Odacidae 

11 

XIV 

-XXVI 

9-23 

II- 

111,8-14 

11- 

-18 

0,  0  or  I,  4 

31-54 

Pomacentridae 

230 

VIII 

-XVII, 

10-18 

II, 

10-18 

14- 

-22 

1,5 

26 

Scaridae 

70 

IX, 

10 

III 

8-9 

13- 

-17 

1,5 

26 

542 


RICHARDS  AND  LEIS:  LABROIDEI 


543 


Fig.  292.     Larva  of  (upper)  Cichlasoma  oclofascialum.  5.4  mm  SL,  laboratory  .reared,  4  days  after  hatching,  drawn  by  B.  Vinter  and  (lower) 
5.9  mm  SL,  laboratory  reared,  10  days  after  hatching,  drawn  by  B.  Vinter. 


squarish,  or  nairow  and  have  choroid  tissue  associated  with  it. 
Larvae  of  temperate  species  tend  to  have  heavy  melanistic  pig- 
ment while  tropical  forms  have  few  melanophores  although 
erythrophores  may  be  abundant.  Meristic  characters  are  very 


useful  for  identifying  these  larvae.  Development  is  direct,  with 
only  the  non-round  eyes  (some  with  choroid  tissue)  and  elongate 
fin  rays  of  some  species,  and  perhaps  the  reduced  melanistic 
pigment  of  tropical  taxa  representing  lai-val  specializations.  Most 


Fig.  293.     Lar\'a  of  Microspathodon  chrysurus  3.7  mm  SL  from  specimen  reared  in  the  laboratory.  From  MS  by  Potthoffet  al.,  drawn  by  J. 
C.  Javech. 


Table  131.     Labroid  Taxa  for  Which  Information  is  Available  on  Ego  and  Larval  Stages.  References  dealing  with  spawning  which  do 
not  describe  eggs  or  larva  are  omitted.  YS— yolk-sac  stage;  pre^preflexion  stage;  flex  — flexion  stage;  post  — postflexion  stage;  D— demersal,  P— 

pelagic. 


Family/genus 


Number  of 
species 


Larvae— developmental  stage 


Egg  type 


Pre 


Flex 


References 


Pomacentridae 

Abudefduf 

Acanthochromis 
Amphipnon 

Chromis 


Scaridae 
Calotomus 

NicholsinaC!) 
Scarus 

Spansoma 


Unidentified 


Odacidae 

Neoodax 
Odax 


D 


D 
D 


D 


Several 


P-round 

P-spindle 
P-round 

P-spindle 


X 


Parental  care  of  larvae 
X  X  X  X 


X 


Microspathodon 

1 

D 

X 

X 

X 

X 

Pomacenlrus 

1 

D 

X 

X 

X 

Stegastes 

1 

D 

X 

Unidentified 

Several 

X 

X 

X 

X 

Labridae 

Bodiamts 

1 

X 

Cenlrolabrus 

1 

D 

X 

Cheilinus 

1 

X 

Choerodon 

1 

X 

Cirrhilabrns 

1 

P 

X 

Coris 

2 

P 

X 

X 

X 

X 

Clenolabrus 

1 

P 

X 

X 

X 

X 

Halichoeres 

4 

P 

X 

X 

X 

Imistius 

1 

X 

Labroides 

1 

P 

X 

Labrus 

2 

D 

X 

X 

X 

X 

Lachnolaimus 

1 

P 

X 

X 

X 

X 

Novaculichthys 

1 

X 

Oxyjulis 

1 

P 

Pseudocheilinus 

1 

X 

Pseudolabrus 

3 

P 

X 

X 

X 

Pleragogiis 

1 

P 

X 

Semicossyphus 

1 

P 

X 

StethojuUs 

2 

P 

X 

X 

X 

Symphodus 

6 

D 

X 

X 

Tautoga 

1 

P 

X 

X 

X 

Taulogolabrus 

1 

P 

X 

X 

X 

Thalassoma 

4 

P 

X 

X 

X 

Xyrichthys 

3 

P 

X 

X 

Unidentified 

Several 

P 

X 

X 

X 

X 

X 

X 


X 
X 


X 
X 


Shaw.  1955;  Mito,  1966;  Miller 

et  al..  1979;  Re,  1980;  Leis 

and  Rennis,  1983 
Robertson,  1973 
Delsman,  1930c;  Allen,  1972; 

Vatanachi,  1972;  Leis  and 

Rennis,  1983 
Page,  1918;  Padoa.  1956d;  Fujita, 

1957a;  Turner  and  Ebert,  1962; 

Ahlstrom,  1965;  Leis  and 

Rennis,  1983 
Potthofl'et  al.,  MS 
Leis  and  Rennis,  1983 
Miller  etal.,  1979 
Nellen,  1973b;  Leis  and  Rennis, 

1983 


Richards,  1984 

Sparta,  1956a;  Russell,  1976 

Leis  and  Rennis,  1983 

Leis  and  Rennis,  1983 

Suzuki  et  al.,  1981 

Page,  1918;  Sparta,  1956a; 

Fourmanoir,  1976 
Russell,  1976 
Mito,  1962b;  Pourmanoir,  1976; 

Richards,  1984 
Masuda  and  Tanaka,  1962 
Suzuki  etal.,  1981 
Sparta,  1956a;  Russell,  1976 
Colin,  1982;  Kelley,  pers.  comm. 
Vatanachi,  1972 
Bolin,  1930;  Orton,  1953a 
Fourmanoir,  1976 
Mito,  1962b;  Robertson,  1975a; 

Crossland,  1981 
Mito,  1962b 
Orton,  1953a 

Mito,  1962b;  Nellen,  1973b 
Sparta,  1956a;  Russell,  1976 
Kuntzand  Radclifle,  1917 
Kuntzand  Radclifle,  1917 
Kubo,  1939;  Sparta,  1956a;  Leis, 

1983;  Richards,  1984 
Sparta,  1956a;  Leis  and  Rennis, 

1983;  Richards,  1984 
Kamiya,  1925;  Mito,  1962b;  Dekhnik 

et  al.,  1966;  Pourmanoir, 

1976;  Miller  etal.,  1979; 

Crossland,  1982;  Leis  and 

Rennis,  1983 


Kamiya,  1925;  Leis  and  Rennis, 

1983 
Regan,  1916;  Aboussouan,  1969 
Winn  and  Bardach,  1960;  Mito, 

1962b 
Sparta,  1956a;  Winn  and  Bardach, 

1 960;  Randall  and  Randall, 

1963 
Watson  and  Leis,  1974;  Leis 

and  Rennis,  1983;  Richards, 

1984 


Regan,  1916 

Robertson,  1975a;  Crossland, 
1982  (as  unidentified  larva  1) 


RICHARDS  AND  LEIS:  LABROIDEI 


545 


->•> 


Fig.  294.  Labrid  larvae  from  top  to  bottom:  Lachnolaimus  maximus.  5.0  mm  SL,  from  specimen  reared  in  the  laboratory  (from  MS  by  Kelley) 
(drawn  by  J.  C.  Javech);  Thalassoma  bifasciatum  8.2  mm  SL  collected  on  R/V  OREGON  II  cruise  7239.  station  10,  18°00'N  latitude,  059°59'W 
longitude,  July  14,  1972  (drawn  by  B.  Washington);  Xynchlhys  sp.  (deep  body  form),  5.0  mm  SL,  collected  on  Ry'V  OREGON  II  cruise  7239, 
station  149,  23°29'N  latitude,  079°  1 3' W  longitude.  August  7.  1972.  [Note  narrow  eyes.  Freshly  caught  specimens  havered  pigment  (erythrophores) 
on  the  head,  trunk,  and  tail]  (drawn  by  B.  Washington);  and  Xynchthys  sp.  (narrow  body  form),  10.5  mm  SL,  collected  on  R/V  OREGON  II  cruise 
7239,  station  149,  23°29'N  latitude,  079°13'W  longitude,  August  7.  1972.  Note  narrow  eyes.  Freshly  caught  specimens  have  red  pigment 
(erythrophores)  on  the  head,  trunk,  and  tail  (drawn  by  B.  Washington). 


546 


ONTOGENY  AND  SYSTEM ATICS  OF  FISHES -AHLSTROM  SYMPOSIUM 


Fig.  295.  Larva  (upper)  of  an  unidentified  scarid,  9.3  mm  SL,  collected  on  RA'  OREGON  II  cruise  7239,  station  54,  18°58'N  latitude,  080°09'W 
longitude,  July  30,  1972;  [meristic  characters  for  Atlantic  scarids  and  the  labrid  Doralonotus  megalepis  are  identical]  (drawn  by  B.  Washington): 
and  larva  (lower)  of  Oda.x  pul/us,  12.2  mm  SL,  from  New  Zealand  (drawn  by  J.  C.  Javech). 


labrid  lai^ae  settle  out  at  less  than  1 5  mm,  but  some  may  remain 
pelagic  until  25  mm. 

Scarids  spawn  pelagic  eggs:  the  subfamily  Scarinae  appears 
to  spawn  spindle-shaped  eggs,  and  the  subfamily  Sparisomatin- 
ae  to  spawn  spherical  eggs  (Table  131).  Morphologically,  scarid 
larvae  are  similar  to  many  labrids:  they  are  elongate  and  com- 
pressed; have  an  initially  straight,  rugose  gut  that  later  coils; 
lack  head  spines;  have  squarish  to  narrow  eyes;  and  usually 
develop  choroid  tissue  (Fig.  295).  Scarid  larvae  differ  most  strik- 
ingly from  labrid  larvae  in  melanistic  pigment.  Scarid  larvae 
consistently  have  melanophores  over  the  posterior  gut  and  have 
a  ventral  series  of  melanophores  on  the  tail.  Melanophores  in 
the  cardiac  region  and  dorsally  on  the  caudal  peduncle  are  com- 
monly found  in  scarids.  Melanophores  in  these  regions  are  either 
absent  or  limited  to  one  or  two  melanophores  in  tropical  labrids. 
The  ventral  series  of  melanophores  on  the  tail  of  some  scarid 
larvae  resembles  a  set  of  developing  photophores  (a  histological 
study  is  warranted).  This  ventral  pigment  plus  the  narrow  eyes 
and  choroid  tissue  (particularly  of  sparisomatines)  give  some 
scarid  larvae  a  gonostomatid  or  myctophid  appearance,  result- 
ing in  some  identification  problems.  Scarines  seem  to  settle  out 
at  10  mm  or  less,  while  Calatomus  (a  sparisomatine)  may  re- 
main pelagic  until  15  mm. 

Little  is  known  of  the  early  life  history  of  odacids,  but  they 
spawn  pelagic  eggs  (Table  131),  and  their  larvae  are  generally 
similar  to  elongate  labrids  with  high  numbers  of  myomeres 
(Table  131).  Only  three  larvae  of  two  species  have  been  de- 
scribed, so  it  is  difficult  to  generalize,  but  these  are  elongate. 


compressed,  have  unlooped  guts,  no  head  spines,  and  round 
eyes.  One  species  has  very  elongate,  early-forming,  anterior  spines 
in  the  dorsal  fin,  and  a  pigment  pattern  of  blotches  along  the 
body  margins  (Fig.  295).  The  other  species  is  unpigmented  and 
lacks  elongate  fin  elements. 

Relationships 

Kaufman  and  Liem  (1982)  include  in  the  Labroidei  the  Po- 
macentridae,  Cichlidae,  Embiotocidae,  Labridae,  Odacidae,  and 
Scaridae  and  further  include  the  Odacidae  and  Scaridae  in  the 
Labridae.  They  consider  the  Pomacentridae  to  be  the  primitive 
sister  group  of  all  the  other  labroids,  the  cichlids  the  primitive 
sister  group  of  embiotocids  and  labrids,  and  embiotocids  the 
primitive  sister  group  of  the  labrids. 

Labroids  are  characterized  by  (1)  united  or  fused  fifth  cera- 
tobranchials  resulting  in  the  formation  of  one  functional  unit, 
(2)  a  true  diarthrosis  between  upper  pharyngeal  jaws  and  the 
basicranium  without  an  intervening  part  of  the  transversus  dor- 
salis  anterior  muscle,  and  (3)  the  presence  of  an  undivided 
sphincter  oesophagi  muscle  forming  a  continuous  sheet  (Kauf- 
man and  Liem,  1982). 

Kaufman  and  Liem's  (1982)  arrangement  and  composition 
of  the  Labroidei  receives  only  limited  support  from  ELH  char- 
acters. The  monophyly  of  the  Labroidei  cannot  be  established 
from  early  life  history  characters.  Pomacentrid  and  cichlid  lar- 
vae are  morphologically  and  developmentally  nearly  indistin- 
guishable from  many  percoid  larvae  (e.g.,  mullids,  gerreids, 
sparids),  while  the  labrids,  scarids,  and  odacids  are  quite  dif- 


RICHARDS  AND  LEIS:  LABROIDEI 


547 


ferent.  A  cursory  study  indicates  larvae  of  these  latter  families 
share  at  least  four  derived  characters:  almost  total  lack  of  head 
spination;  a  long,  rugose,  straight  gut  which  loops  relatively  late 
in  development;  compressed,  elongate  body;  and  a  reduction  in 
principal  caudal  ray  number  from  the  typical  percoid  comple- 
ment of  9  +  8.  The  "percoid"  larval  type  of  the  pomacentrids 
and  cichlids  might  be  a  primitive  character  state,  but  there  are 
no  derived  characters  which  unite  their  larvae  with  the  labrid 
type  of  larvae.  At  least  gut  development  and  head  spination  of 
the  labrids  are  shared  with  the  pseudochromids,  which  are  gen- 
erally very  similar  to  some  labrid  larvae  which  settle  at  small 
sizes  (Leis  and  Rennis,  1983).  This  may  be  the  result  of  con- 
vergence, but  a  labrid/pseudochromid  relationship  should  be 
investigated  as  an  alternative  to  Kaufman  and  Liem's  (1982) 
proposed  phylogeny. 

If  Kaufman  and  Liem  (1982)  are  correct  in  proposing  the 
pomacentrids  as  the  primitive  sister  group  of  the  other  labroids, 
then  either  parental  care  of  hatched  young  evolved  indepen- 
dently in  the  pomacentrid  Acanihochromis  and  the  cichlids  (vi- 
viparity in  embiotocids  might  be  a  derivation  of  parental  care 
of  eggs  and  hatched  larvae,  but  this  remains  to  be  shown),  or 
was  present  in  a  pre-pomacentrid  common  ancestor  and  was 
secondarily  lost  in  all  labroids  but  the  cichlids  and  Acaniho- 
chromis. Similarly,  either  demersal  eggs  and  parental  care  of 
them  evolved  independently  in  some  labrids,  pomacentrids,  and 
cichlids.  or  were  present  in  a  pre-pomacentrid  common  ancestor 
and  secondanly  lost  in  most  labrids  and  all  scarids  and  odacids. 
Therefore,  neither  demersal  eggs  nor  parental  care  of  hatched 
young  offer  much  support  to  Kaufman  and  Liem's  (1982)  phy- 
logeny. 


ELH  characters  may  be  useful  in  studying  the  intrafamilial 
relationships  of  labroid  fishes.  Larval  labrids  are  very  diverse 
in  development  and  morphology,  and  this  may  prove  useful  in 
elucidating  labrid  interrelationships.  Within  the  labrids,  de- 
mersal eggs  and  parental  care  of  eggs  are  unique  to  some  mem- 
bers of  the  tribe  Labrini.  Egg  shape  and  larval  morphology  sup- 
port the  subfamilial  divisions  within  the  Scaridae.  Too  little  is 
known  of  pomacentrid  and  cichlid  development  to  say  if  ELH 
characters  might  be  useful  in  elucidating  intrafamilial  relation- 
ships. 

In  conclusion,  ELH  characters  support  Kaufman  and  Liem's 
(1982)  labroid  phylogeny  only  in  the  close  relationship  of  the 
labrids,  scarids,  and  odacids.  In  spite  of  the  similarities  uniting 
the  three  families,  there  are  enough  differences  between  their 
known  larvae  to  lead  us  to  suggest  the  labrids,  scarids,  and 
odacids  should  not  be  combined  into  one  family  at  this  time. 
M.  F.  Gomon  (pers.  comm.)  argues  that  the  alternative  to  com- 
bining the  three  families  into  one  is  splitting  the  group  into  as 
many  as  five  smaller  families.  While  we  do  not  advocate  this 
course,  the  great  larval  diversity  found  within  the  group  could 
provide  evidence  supporting  this  alternative.  However.  ELH 
information  for  more  genera  of  labrids,  scarids,  and  odacids 
must  be  gathered  before  firm  statements  can  be  made. 

(W.J.R.)  National  Marine  Fisheries  Service,  Southeast 
Fisheries  Center,  75  Virginia  Beach  Drive,  Miami, 
Florida  33149;  (J.M.L.)  The  Australian  Museum,  6-8 
College  Street,  Sydney  2000,  Australia. 


Acanthuroidei:  Development  and  Relationships 
J.  M.  Leis  and  W.  J.  Richards 


THE  acanthuroid  fishes  are  marine,  tropical  and,  except  for 
the  pelagic  Luvaridae,  are  associated  with  coral  reefs.  The 
suborder  consists  of  about  1 10  species  distributed  among  four 
families:  Acanthuridae  (Randall,  1955a),  Luvaridae  (Roule, 
1924;  Tyler,  Nakamura  and  Collette,  MS  in  prep.).  Siganidae 
(Woodland,  1983),  and  Zanclidae  (we  follow  Randall,  1981,  and 
consider  the  Zanclidae  distinct  from  the  Acanthuridae).  Ap- 
parently, all  species  have  a  specialized  pelagic  stage  between 
larvae  and  juveniles,  often  referred  to  as  an  acronurus  larval 
stage  (we  prefer  to  restrict  this  term  to  its  original  usage  in  the 


Acanthuridae).  This  specialized  pelagic  stage  has  provided  the 
basis  for  the  description  for  many  supposedly  new  species  and 
genera,  and  has  been  used  as  evidence  for  uniting  the  group 
(e.g.,  Lauder  and  Liem,  1 983).  The  siganids  are  usually  consid- 
ered the  most  generalized  (=primitive)  family  of  the  suborder, 
and  the  zanclids  are  considered  closely  related  to  if  not  included 
in  the  acanthurids  (Tyler,  1 970).  Luvanis  has  recently  been  shown 
to  be  closely  related  to  the  acanthurids  (Tyler,  Nakamura  and 
Collette,  MS  in  prep.).  The  chaetodontids  have  been  suggested 
as  the  percoid  group  from  which  acanthuroids  were  derived 


Table  132.  Meristic  Characters  of  Acanthuroid  Fishes.  N  is  the  number  of  recent  species,  principally  after  Nelson,  1976.  Note  that  in  the 
Luvandae  there  is  a  progressive  loss  offin  rays  from  the  larval  stage  (adult  counts  in  parentheses).  Maximum  larval  counts  are  followed  by  adult 
counts  in  parentheses.  (Data  from  Randall.  1955b.  c:  Smith.  1966a;  Weber  and  de  Beaufort.  1936;  Gregory  and  Conrad.  1943;  and  Leis  and 

Rennis.  1983). 


N 

D 

A 

P. 

P; 

c 

Venebrae 

Acanthundae 

85 

IV-IX.  19-33 

II-IV.  18-32 

14-19 

I.  3-1,  5 

16 

22-23 

Siganidae 

23 

XII-XIV.  9-11 

VII.  9-10 

14-21 

1,3,1 

17 

23 

Zanclidae 

1 

VII.  38-42 

111.31-35 

18-19 

1.5 

16 

22 

Luvaridae 

1 

11.24(12-13) 

18(13-14) 

17-20 

1.4(0) 

16 

23 

548 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  133.     Acanthuroid  Taxa  for  Which  Information  is  Available  on  Ego  and  Larval  Stages.  YS— yolk-sac;  pre— preflexion;  flex- 
flexion  stage;  post  — postflexion  stage;  D— demersal;  P— pelagic. 


Number  of 
species 


Larvae— developmental  stage 


Egg  type 


Pre 


Flex 


Acanthuridae 

Acanihurus 


Clenochaetus 
Naso 

2 
2 

Zebrasoma 
Unidentified 

2 
Several 

Luvaridae 

Luvarus 

1 

Siganidae 

Siganus 

6 

Zanclidae 

Zanchis 


X 
X 


X 
X 

X 
X 


Lutken,  1880;  Breder.  1927;  Whitley 
andColefax,  1938;  Randall,  1956, 
1961;  Aboussouan,  1965;  Burgess, 
1965 

Randall,  1955c 

Fourmanoir,  1976;  Leis  and  Rennis, 
1983 

Randall,  1955b;  Aboussouan,  1966a 

Dekhnik  et  al.,  1966  (Fig.  37-1,  misiden- 
tified  as  Balistidae);  Randall,  1955c; 
Nellen,  1973b;  Watson  and  Uis,  1974 


Roule,  1924;  Roule  and  Angel,  1930; 
Blache,  1964 


Fujita  and  Ueno,  1954;  Uchida  et  al., 
1958;  Mito,  1966;  Popper  etal.,  1973; 
May  et  al.,  1974;  von  Westemhagen 
and  Rosenthal,  1975,  1976;  Bryan  and 
Madrisau,  1977;  Leis  and  Rennis, 
1983 


Strasburg,  1962 


(Tyler,  1970).  Table  132  summarizes  meristic  characters  of  the 
suborder,  and  Table  133  reviews  current  state  of  knowledge  of 
its  early  life  history. 

Development 

Siganids  have  small  (<  1  mm)  demersal  eggs  with  oil  droplets 
(Table  133).  No  parental  care  has  been  recorded.  Larvae  hatch 
in  a  poorly  developed  condition.  Moderately  long  preopercular 
spines  and  serrate  ridges  form  on  the  head,  and  the  first  fin 
elements  to  form— the  pelvic  spine  and  second  dorsal  spine— 
form  very  early  (Fig.  296).  The  body  does  not  become  very 
deep,  and  although  the  pelvic  and  dorsal  spines  are  elongate 
and  serrate,  they  do  not  exceed  three  times  the  eye  diameter. 
No  scales  form  prior  to  settlement,  but  the  pelagic  stage  may 
grow  to  30  mm  and  is  very  silvery  in  life,  particularly  over  the 
gut.  Early  larvae,  in  particular,  are  very  percoid  in  appearance. 
Figment  in  preflexion  larvae  is  limited  to  dorsal  and  pelvic  fin 
membranes,  gut,  and  a  ventral  series  on  the  tail.  Older  larvae 
are  more  heavily  pigmented. 

Acanthurids  have  small  (<  I  mm)  pelagic  eggs  with  a  single 
oil  droplet  (Table  133).  Larvae  hatch  in  a  poorly  developed 
condition,  but  very  soon  develop  serrate  ridges  on  the  head  (but 
no  elongate  preopercular  spines  form).  The  first  fin  elements  to 
form  (the  pelvic  spine  and  second  dorsal  spine)  do  so  very  early, 
and  these  are  quickly  followed  by  the  second  anal  spine  (Fig. 


296).  These  fin  spines  are  serrate,  and  at  least  one  exceeds  three 
times  the  diameter  of  the  eye.  The  head  and  trunk  become 
remarkably  deepened.  This  is  accentuated  by  the  elongate  pel- 
vic, dorsal,  and  anal  spines  at  the  opposite  edges  of  the  deepest 
point  of  the  body:  the  body  becomes  distinctly  kite-shaped. 
Small,  triangular  scales  arrayed  in  vertical  rows  begin  to  form 
shortly  after  flexion.  The  pelagic  stage  may  reach  60  mm  and 
is  very  silvery  in  life  around  the  gut.  Preflexion  larvae  are  lightly 
pigmented  in  specific  patterns.  Late  larval  stages  may  acquire 
aspects  of  the  juvenile  pigment  pattern.  The  caudal  peduncle 
armature  forms  late  in  the  larval  period.  In  Naso  the  spines  form 
from  existing  scales  (i.e.,  they  pass  through  an  unspecialized 
scale  stage).  In  Acanihurus  it  forms  directly  without  the  unspe- 
cialized scale  stage. 

Nothing  is  known  of  luvarid  or  zanclid  eggs  or  preflexion 
larvae  (Table  133).  Luvarus  larvae  apparently  have  early-form- 
ing pelvic  and  anterior  dorsal  fin  spines.  They  also  have  early- 
forming  scales,  serrations  on  the  head,  but  lack  elongate  pre- 
opercular spines  (Fig.  297).  The  dorsal  and  pelvic  spines  of 
Luvarus  are  more  than  three  times  the  diameter  of  the  eye. 
Luvarus  larvae  are  deep-bodied,  but  not  as  kite-shaped  as  acan- 
thurids, and  have  a  more  square-shaped  head.  With  growth,  the 
spines  of  the  fins,  and  many  of  the  soft  rays  are  lost,  and  the 
body  becomes  more  fusiform.  Late  Zanclus  larvae  are  very 
similar  to  acanthurid  larvae  (Fig.  297)  and  are  scaled  similarly. 


Fig.  296.  Urvae  of  (upper)  Acanihurus  sp.,  6.0  mm  SL,  OREGON  II  cr.  7343,  sta.  87,  Caribbean  Sea,  16''54'N.  062°03'W,  February  16. 
1973,  drawn  by  J.  C.  Javech;  (middle)  Naso  unicornis.  5.9  mm  SL,  PROVIDENCE  II,  st.  T-429,  Indian  Ocean,  09°27'S,  050°2rE,  December 
18,  1974,  drawn  by  J,  C.  lavech;  and  (lower)  Siganus  fuscescens,  5.4  mm  TL,  modified  after  Uchida  et  al.  (1958). 


LEIS  AND  RICHARDS:  ACANTHUROIDEI 


549 


550 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  297.     Larvae  of  (upper)  Luvarus  imperiatis,  6.8  mm  TL,  modified  after  Fahay  ( 1 983);  and  (lower)  Zanclus  canescem.  16  mm  SL.  modified 
after  Strasburg  (1962). 


but  only  the  third  dorsal  spine  is  elongate  (the  unknown  smaller 
larvae  may  have  elongate  spines  in  other  fins). 

Relationships 

Acanthuroids  share  the  following,  probably  derived  charac- 
ters (we  assume  acanthuroids  have  a  percoid  ancestry):  long 
pelagic  period;  early-forming,  elongate  dorsal  and  pelvic  spines; 
serrate  fin  spines;  moderately  to  very  deep,  compressed  body; 


serrate  ridges  on  the  head;  silvery  gut;  22-23  vertebrae;  and  16- 
1 7  principal  caudal  rays.  This  is  strong  evidence  for  the 
monophyly  of  the  group. 

Tyler  (1970)  notes  that  acanthuroids  have  been  considered 
as  chaetodontid  derivatives.  We  find  no  support  for  this  view 
among  ELH  characters.  Chaetodontids  and  pomacanthids  do 
have  large,  specialized  pelagic  stages,  but  these  differ  greatly 
from  acanthuroids  (Leis  and  Rennis,  1983)  and  resemble  ca- 


LEIS  AND  RICHARDS:  ACANTHUROIDEI 


551 


rangids  at  early  stages.  Leiognathid  larvae  are  similar  to  siganid 
larvae  in  many  respects  (head  spination,  fin  spine  development, 
silvery  gut— see  G.  D.  Johnson,  this  volume  and  Leis  and  Gold- 
man, 1 983),  and  we  suggest  the  leiognathids  should  be  evaluated 
as  a  potential  primitive  sister  group  of  the  acanthuroids.  There 
is  little  evidence  from  ELH  characters  to  support  the  notion 
that  the  acanthuroid  fishes  are  the  primitive  sister  group  of  the 
tetraodontiform  fishes  (Leis,  this  volume). 

Intra-ordinal  relationships  of  acanthuroid  fishes  as  suggested 
by  ELH  characters  fully  support  those  based  on  adult  characters. 
The  siganids  are  distinguished  from  the  other  acanthuroids  by 
the  following  derived  characters:  demersal  egg,  two  spines  in 
pelvic  fin,  and  seven  spines  in  the  anal  fin.  Larvae  of  acanthurids, 
luvarids,  and  zanclids  have  the  following  derived  characters:  no 
elongate  preopercular  spines;  kite-shaped  body;  elongate  snout; 
extremely  elongate  dorsal  and  pelvic  spines;  early-forming  spe- 


cialized scales;  and  reduced  number  of  dorsal  fin  spines.  Thus 
the  siganids  appear  to  be  the  primitive  sister  group  of  the  other 
acanthuroids.  Interrelationships  of  the  acanthurids,  zanclids, 
and  luvarids  cannot  be  clarified  given  the  current  knowledge  of 
zanclid  and  luvarid  ELH  characters.  Larval  zanclids  have  an 
extremely  elongate  dorsal  spine  and  a  retrose  preorbital  spine. 
Acanthurids  have  caudal  peduncle  armature,  and  luvarids  have 
ontogenetic  reduction  in  fin  elements,  no  anal  spines,  and  a  very 
squared  head.  None  of  these  specializations  are  shared  by  any 
two  of  the  families,  so  they  shed  no  light  on  interrelationships. 

(J.M.L.)  The  Australian  Museum,  6-8  College  Street, 
Sydney  2000,  Australia;  (W.J.R.)  National  Marine 
Fisheries  Service,  Southeast  Fisheries  Center,  75  Vir- 
ginia Beach  Drive,  Miami,  Florida  33149. 


Blennioidei:  Introduction 
R.  H.  Rosenblatt 


THE  modem  concept  of  the  perciform  suborder  Blennioidei 
dates  from  the  paper  of  Regan  (1912b),  who  defined  and 
delimited  the  group  as  "Percomorphous  Teleosts  with  the  pelvic 
fins  jugular  or  mental,  each  of  a  spine  and  four  soft  rays  or  still 
further  reduced,  with  the  dorsal  and  anal  rays  typically  corre- 
sponding in  number  to  the  vertebrae,  each  basal  bone  attached 
to  its  own  neural  or  haemal  spine  (rays  more  numerous  in  Ophi- 
diiformes)  with  well  developed  wings  of  the  parasphenoid  as- 
cending in  front  of  the  prootics,  and  with  all  or  most  of  the  ribs 
inserted  on  strong  parapophyses." 

As  Regan  himself  indicated  this  definition  encompasses  a 
heterogeneous  group,  and  his  series  "Ophidiiformes"  has  now 
been  removed  from  the  Perciformes.  Subsequent  to  Regan  sev- 
eral widely  differing  classifications  have  been  proposed,  with 
groups  often  being  added  or  removed  without  comment.  Jordan 
(1923)  proposed  the  most  radical  arrangement.  He  placed  in  the 
order  Jugulares  almost  all  spiny  rayed  fishes  with  advanced 
pelvic  fins.  Jordan's  Jugulares  was  divided  by  him  into  1 2  series, 
comparable  to  suborders,  and  no  less  than  62  families.  Jordan, 
in  his  magisterial  fashion,  provided  an  outline  classification, 
without  substantiation  by  characters. 

Berg  in  his  1940  classification  rationalized  the  classification 
of  the  Blennioidei.  He  restricted  the  suborder  mainly  to  Regan's 
series  "Blenniformes"  and  "Cliniformes,"  and  redistributed  the 
remainder  of  the  Jugulares,  either  to  the  Percoidei  or  to  the 
suborders  Ophidioidei  (equivalent  to  Regan's  series  Ophidi- 
iformes), Ammodytoidei,  or  Callionymoidei.  Some  indication 
of  relationships  is  perhaps  implicit  in  Berg's  placement  of  por- 
tions of  Jugulares  auctorum  immediately  preceding  the  Blen- 
nioidei. 

Although  a  number  of  works  on  various  blennioid  groups 
have  appeared  (see  particularly  Hubbs,  1952;  Makushok,  1958, 
and  the  papers  of  V.  Springer)  the  only  subsequent  attempt  to 
characterize  and  deal  with  the  group  as  a  whole  is  that  of  Gosline 
(1968).  The  classification  given  by  Nelson  (1976)  differs  from 


that  of  Gosline  as  well  as  the  outline  classification  of  Greenwood 
et  al.  (1966).  The  discussion  of  larval  forms  given  here  mostly 
accords  with  Nelson's  Blennioidei  as  a  convenience,  regardless 
of  the  eventual  disposition  of  the  taxa.  The  only  major  departure 
from  the  arrangement  of  Nelson  is  that  the  family  Zoarcidae  is 
treated  here,  although  Nelson  included  it  in  the  Gadiformes  (see 
Anderson,  this  volume). 

The  reasons  for  the  varying  treatment  of  these  fishes  are  not 
difficult  to  find.  The  unraveling  of  phyletic  lines  within  the 
Perciformes  is  made  difficult  by  the  sheer  number  of  species 
and  genera.  One  is  faced  with  the  choice  of  mining  a  narrow 
vein  for  nuggets  of  knowledge  which  lie  isolated,  or  engaging  in 
a  strip  mining  operation  which  reveals  broad  patterns  at  the 
expense  of  ignoring  contradictory  details.  In  other  terms,  in- 
sufficient knowledge  of  morphological  variation  within  the  Per- 
ciformes precludes  at  this  time  either  identification  of  unequiv- 
ocal synapomorphies  or  the  determination  of  polarity  of  a  number 
of  characters  within  almost  any  presumed  lineage. 

A  number  of  features  taken  to  characterize,  if  not  to  define, 
the  Blennioidei  may  be  the  product  of  convergent  or  parallel 
evolution,  correlated  with  the  assumption  of  benthic  life. 

As  pointed  out  by  Gosline  ( 1 968)  the  blennioids,  as  compared 
with  percoids,  have  less  deep  bodies,  with  a  short  trunk  and  a 
relatively  attenuated  caudal  region.  The  dorsal  and  anal  are  long 
and  low,  terminating  near  the  caudal,  and  the  pectoral  and  usu- 
ally the  caudal  fins  are  rounded.  There  is  an  exact  correspon- 
dence in  number  between  dorsal  and  posterior  anal  soft-rays 
and  vertebrae  supporting  them.  The  pelvic  fins  are  inserted  in 
advance  of  the  pectoral  fins,  and  the  number  of  rays  is  generally 
reduced;  the  spine  often  rudimentary  or  splint-like,  and  the  soft- 
rays  three  or  fewer. 

The  deep,  relatively  compact  body  of  a  generalized  perciform 
is  that  of  a  fish  which  hovers,  probably  near  the  substrate,  but 
which  makes  rapid  bursts  either  in  feeding  or  predator  avoid- 
ance, or  both.  The  body  shape  is  adapted  for  slow  swimming. 


552 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


alternating  with  bursts  of  acceleration.  There  is  a  general  but 
far  from  universal  trend  for  bottom  dwelling  fishes  to  become 
elongate;  the  eel-like  body  is  widely  distributed  taxonomically. 
Bottom-living  fishes  often  use  crevices  for  shelter  and  forage  in 
interstices,  and  may  burrow.  Elongation  of  the  body  accom- 
panied by  an  increase  in  the  number  of  vertebrae  produces  the 
flexibility  necessary  for  these  activities.  The  elongate  body  form 
requires  either  anguilliform  swimming  or  undulation  of  the  me- 
dian fins.  In  either  case  the  role  of  the  caudal  fin  is  reduced. 
The  pectorals  are  used  in  short  darts  or  lunges,  and  their  fan 
shape  is  associated  with  accelerating  a  large  amount  of  water 
per  thrust.  This  function  is  important  even  in  relatively  elongate 
forms  in  accelerating  the  head  in  feeding  strikes,  and  pectorals 
are  reduced  or  lost  in  only  a  few  lineages. 

The  pelvics  of  bottom  living  forms  no  longer  have  a  hydro- 
dynamic  function  as  brakes  or  rudders.  Instead  they  may  func- 
tion as  props  which  hold  the  head  ofl^  the  bottom  (as  in  the 
Cottidae  and  Gobiidae  as  well).  A  reduction  in  the  number  of 
rays  is  also  seen  in  the  Cottidae. 

That  morphological  features  are  functional  does  not  mean 
that  their  joint  possession  cannot  be  taken  to  demonstrate  com- 
mon ancestry.  However,  it  does  indicate  caution.  The  only  one 
of  Gosline's  characters  for  the  Blennioidei  that  is  not  clearly 
functional  is  the  1:1  relationship  of  median  fin  rays  and  ver- 
tebrae. However,  the  reduction  in  the  number  of  fin  rays  per 
segment  to  one  is  the  culmination  of  a  functional  trend  begun 
in  the  Paleozoic,  and  we  cannot  yet  be  sure  that  it  happened 
but  once. 

Although  Gosline  regarded  his  classification  as  owing  more 
to  that  of  Jordan  than  Regan,  his  mam  characters  of  pelvic 
position  and  median  fin  ray  arrangement  are  exactly  those  given 
by  Regan  in  his  diagnosis.  Gosline's  concept  of  the  Blennioidei 
and  its  superfamilies,  although  not  completely  accepted  (see 
Nelson,  1976),  has  not  been  superseded,  except  that  his  Con- 
grogadoidae  is  no  longer  included;  the  Congrogadidae  is  now 
placed  in  the  Percoidea  (Winterbottom,  1982)  and  the  Pero- 
nedysidae  has  been  synonymized  with  the  Clinidae  (George  and 
Springer,  1980). 

According  to  Gosline  the  Blennioidei  (without  the  Congro- 
gadoidae)  may  be  divided  into  four  superfamilies.  The  first  of 
these,  the  Notothenioidae,  is  clearly  the  most  heterogeneous.  In 
addition  to  the  Antarctic  and  sub-Antarctic  families  (Bovich- 
thyidae,  Nototheniidae,  Harpagiferidae,  Bathydraconidae  and 
Channichthyidae)  usually  placed  in  this  group  (Berg,  1940),  the 
tropical  Mugiloididae  (=Parapercidae)  Trichonotidae  and  Chei- 


marrhichthyidae  were  included  although  they  do  not  share  with 
them  the  specialized  features  of  a  smgle  nostril  and  a  loss  of 
one  pectoral  actinost.  There  appears  to  be  no  reason  to  regard 
the  two  groups  of  families  as  closely  related. 

The  Trachinoidae  was  said  to  be  comprised  of  the  Trachin- 
idae,  Leptoscopidae,  Uranoscopidae  and  Dactyloscopidae.  All 
are  adapted  for  lying  buried  in  the  substrate,  and  it  is  likely  that 
their  structural  similarities  are  related  to  this  habit.  The  Dac- 
tyloscopidae has  recently  been  placed  in  the  Blennioidae  (George 
and  Springer,  1982). 

The  superfamily  Blennioidae  was  regarded  as  composed  by 
the  families  Tripterygiidae,  Clinidae,  Chaenopsidae,  and  Blen- 
niidae.  Subsequently  the  subfamily  Labrisominae  of  the  family 
Clinidae  was  raised  to  family  status  and  the  Dactyloscopidae 
transferred  from  the  Trachinoidei  (George  and  Springer,  1980). 
Within  the  Blennioidei,  the  superfamily  may  be  characterized 
by  the  combination  of  two  nostrils  on  each  side,  pelvic  soft- 
rays  four  or  fewer,  prootic  excluded  from  orbital  rim  (that  is, 
ascending  wing  of  paraspheroid  meets  frontal),  and  basisphe- 
noid  present. 

The  remaining  superfamily,  the  Zoarceoidae,  was  regarded  as 
composed  of  1 1  families,  some  poorly  understood.  Anderson 
(1983,  this  volume)  recognized  8  families  in  the  group:  Bathy- 
masteridae,  Stichaeidae,  Pholididae,  Anarhichadidae,  Ptil- 
ichthyidae,  Zaproridae,  Scytalinidae  and  Zoarcidae.  Although 
composed  of  forms  differing  greatly  in  morphology,  the  super- 
family  may  be  diagnosed  as  blennioids  with  a  single  nostril  on 
either  side  of  the  head,  prootic  excluded  from  rim  of  orbit,  and 
basisphenoid  absent.  There  is  no  merit  in  the  removal  of  the 
Zoarcidae  to  the  Gadiformes  (see  also  Anderson,  this  volume). 

It  should  be  clear  from  the  foregoing  that  no  satisfactory  def- 
inition of  the  Blennioidei  has  as  yet  been  framed.  Perhaps  lines 
of  relationships  would  best  be  recognized  by  restricting  the  Blen- 
nioidei to  the  Blennioidae  and  Zoarceoidae  of  Gosline,  and 
returning  his  other  two  superfamilies  to  the  Percoidei.  It  appears 
that  ontogeny  and  larval  characters  have  as  yet  little  to  con- 
tribute to  questions  of  suprafamilial  and  subordinal  relation- 
ships between  and  among  these  fishes. 

Perhaps  it  is  fitting  to  end  with  a  quote  from  Jordan  (1923), 
addressing  issues  such  as  this:  "I  may  repeat  a  warning  as  old 
as  science  itself:  that  we  must  not  expect  a  degree  of  accuracy 
which  the  subject  in  question  does  not  permit." 

ScRipps  Institution  of  Oceanography,  University  of  Cal- 
ifornia, San  Diego,  La  Jolla,  California  92093. 


Schindlerioidei:  Development  and  Relationships 
W.  Watson,  E.  G.  Stevens  and  A.  C.  Matarese 


THIS  suborder  contains  a  single  paedomorphic  family  com- 
posed of  two  species  of  the  genus  Schindlena.  Both  species 
inhabit  neritic  surface  waters  of  the  subtropical  and  tropical 
Indian  and  Pacific  Oceans  (Bruun,  1940;  Schindler,  1932;  R.  J. 
Lavenberg,  pers.  comm.).  Their  early  life  histories  are  known 


from  the  work  of  Watson  and  Leis  (1974),  Miller  et  al.  (1979), 
and  Ozawa  and  Matsui  (1979).  Classification  of  Schindlerioidei 
is  speculative,  and  its  placement  here  by  Nelson  (1976)  follows 
Gosline  (1971),  who  tentatively  considered  this  taxon  a  percoid 
derivative,  possibly  related  to  Ammodytoidei. 


WATSON  ET  AL.:  SCHINDLERIOIDEI 


553 


B 


.:^:^Za:Z^^^^ 


Fig.  298,  Lateral  views  of:  (A)  Schmdlena  pietschmanni  larva,  2.7  mm;  (B)  5.  pietschmanni  larva.  3,5  mm  (redrawn  from  Miller  et  al..  1979); 
(C)  5.  pietschmanni  larva.  4.7  mm  (from  Miller  et  al.,  1979);  (D)  S  pielschmanm  adult  female.  15.1  mm  (redrawn  from  Jones  and  Kumaran, 
1964);  (E)  S.  praematura  larva,  3.6  mm  (from  Ozawa  and  Matsui.  1979);  and  (¥)  S  praematura  adult  female.  20.1  mm  (redrawn  from  Jones 
and  Kumaran,  1964). 


Development 

Eggs 

Although  ovarian  eggs  are  well-known  for  both  species  (Jones 
and  Kumaran.  1964;  Sardou.  1974),  the  mode  of  spawning  is 
unknown.  Watson  and  Leis  (1974)  reported  planktonic  Schind- 


leria  sp.  eggs  which  they  suggested  were  either  pelagic  or  perhaps 
dermersal  eggs  extruded  in  the  net.  The  largest  ovarian  eggs  lack 
oil  droplets  and  are  irregular  in  shape,  0.35-0.40  mm  in  di- 
ameter (S.  praematura).  or  oval.  0.30  x  0.65  mm  (S.  pietsch- 
manni'). Hydrated,  planktonic  eggs  of  Schmdlena  sp.  are  oval, 


554 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


0.50  X  1. 30  mm,  contain  no  droplets,  and  have  an  unsculptured 
chorion  with  a  cap-Uke  structure  at  one  end.  Incubation  time 
is  not  known. 

Larvae 

Morphology.  — Larval  size  and  degree  of  development  at  hatch- 
ing are  unknown.  However,  5.  pietschmanm  at  1 .9  mm  NL  has 
a  rather  large  yolk  sac  (containing  an  apparently  segmented  yolk) 
in  addition  to  pigmented  eyes  and  an  open,  presumably  func- 
tional, mouth.  Notochord  flexion  occurs  after  2.7  mm  but  before 
3.5  mm  NL  in  5.  pietschmanni,  and  before  4.3  mm  in  S.  prae- 
matura.  Development  to  the  essentially  larval  mature  form  is 
gradual.  The  juvenile  stage  may  be  taken  to  begin  with  com- 
pletion of  the  dorsal  and  anal  fins  and  the  acquisition  of  the 
principal  caudal  rays  (ca.  4-5  mm),  and  the  adult  stage  to  begin 
when  the  male  genital  papilla  or  the  ovaries  of  the  female  be- 
come discemable  (longer  than  ca.  9  mm  SL).  The  distinctive 
schindleriid  terminal  section  at  the  rear  of  the  vertebral  column 
does  not  become  apparent  until  the  late  larval  or  early  juvenile 
period. 

Aside  from  fin  development,  morphology  changes  little  during 
larval  development.  The  swim  bladder  moves  posteriorly  from 
myomeres  6-8  to  myomeres  14-1 5  in  5.  pietschmanni;  a  similar 
migration  presumably  occurs  in  S.  praematura  (e.g.,  Sardou, 
1974).  Preanal  length  is  greater  in  S.  praematura  than  in  S. 
pietschmanm. 

Pigmentation.  — Sdn'mdXmids  are  lightly  pigmented  throughout 
development  (e.g..  Miller  etal.,  1979;  Ozawa  and  Matsui,  1979). 
During  the  larval  and  early  juvenile  period,  S.  pietschmanm 
has  one  to  four  pairs  of  melanophores  along  the  sides  of  the  gut 
(usually  two  or  three  pairs),  one  to  four  melanophores  along  the 
ventral  midline  of  the  tail  (usually  two  or  three),  and  pigment 
on  the  posterior  dorsal  surface  of  the  swim  bladder.  The  pos- 
terior tail  melanophore  is  typically  more  elongate  than  the  others 
(Fig.  298).  All  but  the  swim  bladder  pigment  is  lost  during  the 


juvenile  stage.  Larval  pigmentation  of  5.  praematura.  as  shown 
by  Ozawa  and  Matsui  (1979),  and  juvenile  pigment,  shown  by 
Sardou  (1974),  are  very  similar  to  that  of  5.  pietschmanni.  Like 
S.  pietschmanni,  S.  praematura  retains  only  the  posterior  swim 
bladder  pigment  in  the  adult  stage  (Fig.  298). 

A/m5/(C5.  —  Meristics  for  Schindleria  are:  Vertebrae  15-25  -I- 
12-21  =  33-44;  D  15-22;  A  10-14;  P  15-17;  and  C  13prin.  A 
combination  of  caudal  vertebrae  and  anal  fin  ray  counts  usually 
will  distinguish  the  two  species. 

The  caudal  fin  rays  are  the  first  to  develop,  followed  by  the 
dorsal  and  anal  fin  rays  (forming  simultaneously).  Pectoral  fin 
rays  are  the  last  to  ossify.  Pelvic  fins  never  form. 

Relationships 

Early  life  history  characters,  to  the  extent  that  they  are  pres- 
ently known,  do  little  to  clarify  the  phylogenetic  position  of  the 
Schindlerioidei.  For  example,  Gosline  ( 1 963b,  1971)  speculated 
that  Schindlerioidei  might  be  derived  from  an  ammodytoid 
ancestor;  however,  while  both  suborders  share  some  characters 
(e.g.,  an  elongate  larval  form  with  preanal  length  just  over  50% 
body  length),  they  differ  in  other  important  ways  (e.g.,  late  de- 
velopment of  pectoral  fin  rays  in  schlindleriids  and  early  de- 
velopment in  ammodytoids).  Knowledge  of  spawning  and  early 
development  might  aid  in  ascertaining  schindleriid  relationships 
although  at  present  this  group  seems  destined  to  remain  an 
enigma. 

(W.W.)  Marine  Ecological  Consultants,  531  Encinitas 
Boulevard,  Suite  1 10,  Encinitas,  California  92024;  (E. 
G.S.)  National  Marine  Fisheries  Service,  Southwest 
Fisheries  Center,  PO  Box  271,  La  Jolla,  California 
92038;  (A. CM.)  National  Marine  Fisheries  Service, 
Northwest  and  Alaska  Fisheries  Center,  2725 
Montlake  Boulevard  East,  Seattle,  Washington  98 112. 


Trachinoidea:  Development  and  Relationships 
W.  Watson,  A.  C.  Matarese  and  E.  G.  Stevens 


THE  blennioid  infraorder  Trachinoidea,  as  used  here,  con- 
tains about  140  species  in  1 1  families  of  morphologically 
quite  diverse,  but  generally  small,  primarily  shallow-living  tem- 
perate and  tropical  marine  demersal  or  burrowing  fishes  (Chias- 
modontidae  is  bathypelagic;  Cheimarrhichthyidae  inhabits  fresh 
water).  These  families  have  not  always  been  considered  as  closely 
related  (e.g.,  Gosline,  1968,  1971),  but  we  follow  Nelson  (1976) 
in  considering  them  together  here.  Nelson  ( 1 976)  originally  placed 
16  families  in  the  Trachinoidea,  but  subsequently  synonymized 
the  Limnichthyidae  with  Creediidae  (Nelson.  1978).  Springer 
(1978)  removed  Oxudercidae  to  the  Gobiidae.  Three  other  fam- 
ilies are  treated  elsewhere  in  this  volume;  Bathymasteridae  and 
Dactyloscopidae  with  the  Blennioidea  (Matarese  et  al.,  this  vol- 
ume) and  Opistognathidae  with  the  Percoidei  (G.  D.  Johnson). 
In  this  brief  review,  we  summarize  the  present  state  of  knowl- 


edge of  the  early  life  histories  of  trachinoid  fishes  and  attempt 
to  determine  whether  such  information  contributes  to  our  un- 
derstanding of  their  phylogenetic  relationships.  Unfortunately, 
early  life  histories,  mostly  incomplete,  are  known  for  only  a 
small  number  of  species  (Table  134).  This  paucity  of  early  life 
history  data  makes  generalizations  about  development  tenuous 
at  best,  but  for  purposes  of  this  paper  the  known  taxa  are  con- 
sidered representative. 


Development 
Eggs 

Eggs  are  unknown  for  the  Percophididae,  Trichonotidae,  and 
Leptoscopidae.  Only  ovarian  eggs  have  been  described  for  the 


WATSON  ET  AL.:  TRACHINOIDEA 


555 


Table  134.    Summary  of  Early  Life  History  Information  Available  for  Trachinoid  Fishes. 


Approxi- 

Number 

mate 

Descriptions 

llluslralions 

of 

niimh«*r  nf 

Family 

genera 

IJ  Ul  J  1  t-n^  t    ^Jl 

species 

Dislnbution 

Genera 

Species 

Genera 

Species 

Early  life  history  source 

Trichodontidae 

2 

2 

North  Pacific 

1 

1 

1 

1 

Breder  and  Rosen,  1966;  Marliave, 
1981 

Cham  psodontidae 

1 

10 

Indo-Pacific 

I 

1 

1 

I 

Mito,  1962d,  1966 

Chiasmodontidae 

4 

23 

Worldwide  (temperate 
and  tropical,  marine) 

1 

1 

1 

1 

Ahlslrom,  pers.  comm.;  Lavenberg, 
pers.  comm. 

Percophididae 

6 

17 

Atlantic,  Indo-Pacific 

1 

1 

1 

1 

Crossland,  1982 

Mugiloididae 

3 

36 

Atlantic,  Indian,  Pacific 
(subtropical  and  tropi- 
cal) 

1 

5 

1 

4 

Leis  and  Rennis.  1983;  Mito,  1966; 
Robertson,  1973,  1975a;  Watson, 
unpubl. 

Trichonotidae 

2 

5 

Indo-Pacific 

1 

2 

1 

1 

Leisand  Rennis,  1983 

Cheimarrhichthyidae 

1 

1 

New  Zealand  (freshwa- 
ter) 

0 

0 

0 

0 

McDowall,  1973c 

Creediidae 

7 

14 

Indo-Pacific 

2 

2 

2 

3 

Leis,  1982;  Leis  and  Rennis,  1983; 
Regan,  1916;  Watson  and  Leis, 

1974 

Trachinidae 

1 

4-5 

Eastern  Atlantic,  Medi- 
terranean 

1 

4 

1 

-> 

Breder  and  Rosen,  1966;  Dekhnik. 
1973;  Ehrenbaum,  1905-1909;  Mari- 
naro,  1971;  Padoa,  1956g;  Russell, 
1976;  Schnakenbeck,  1928;  Vod- 
vanitskv  and  Kazanova,  1954 

Uranoscopidae 

8 

25 

Atlantic,  Indian,  Pacific 
(shallow  temperate 
and  tropical) 

3 

4 

2 

3 

Dekhnik,  1973;  Fritzsche,  1978;  Mito, 
1966;  Robertson,  1974 

Leptoscopidae 

2 

3 

Australia,  New  Zealand 
(marine) 

0 

0 

0 

0 

Cheimarrhichthyidae  (McDowall,  1973c).  Six  of  the  seven  re- 
maining families  spawn  small  to  moderate  (0.70-2.45  mm  di- 
ameter), spherical,  single  pelagic  eggs  (Table  135).  McDowall 
(1973c)  suggested  a  pelagic  spawning  mode  for  Cheimar- 
rhichthyidae as  well,  unusual  for  the  suggested  riparian  spawn- 
ing habitat  but  consistent  with  the  close  relationship,  or  identity, 
of  Cheimarrhichthyidae  with  Mugiloididae.  All  pelagic  eggs  have 
oil  droplets  (most  have  only  one,  0.16-0.26  mm  in  diameter) 
and  all  except  some  Uranoscopidae  have  smooth,  unsculptured 
chorions.  Incubation  periods  range  from  2  to  6  days  and  larvae 
are  not  well  developed  at  hatching  (Trachinidae  are  somewhat 
better  developed,  with  pigmented  eyes  and  pelvic  buds). 

Demersal  egg  masses  (750-1,000  eggs)  are  produced  only  by 
the  Trichodontidae  (Table  135).  These  eggs  are  large  (3.52  mm 
in  diameter),  slightly  flattened,  with  an  unsculptured  chorion 
and  no  oil  droplet.  Incubation  is  estimated  at  about  one  year 
(Mariiave.  1981)  and  larvae  are  well  developed  at  hatching. 

Larvae 

Larval  stages  are  unknown  for  the  Cheimarrhichthyidae  and 
Leptoscopidae.  The  described  trachinoid  larvae  display  only  a 
few  unifying  characteristics:  (1)  all  are  pelagic,  hatching  at  ca. 
2-15  mm  (Table  1 36);  (2)  they  pass  through  no  specialized  stages 
(except  the  gargaropteron  juvenile  stage  of  the  chiasmodontid 
genus  Kali):  and  (3)  they  metamorphose  gradually  to  the  de- 
mersal juvenile  stage  at  a  small  to  moderate  size  (ca.  10-60 
mm). 

Mo/7)/)o/og)'.  —  Morphology  is  quite  variable;  however,  larvae 
are  either  relatively  long  and  slender  (Fig.  299:  Trichodontidae, 
Chiasmodontidae,  Percophididae,  Trichonotidae,  Creediidae) 
or  rather  robust  (Fig.  300:  Champsodontidae,  Mugiloididae, 


Trachinidae,  Uranoscopidae).  All  the  robust  larvae  and  one  of 
the  slender  types  (Trichodontidae)  have  somewhat  rounded  heads 
with  relatively  short  snouts.  Preanal  length  in  both  types  usually 
is  not  more  than  50%  of  standard  length  (60%  or  more  in  Cree- 
diidae and  Trichonotidae)  and  changes  little  during  develop- 
ment. Head  and  body  spination  are  extremely  variable.  Pre- 
opercular  spination  is  known  for  six  families:  Trichodontidae, 
Chiasmodontidae,  Champsodontidae,  Mugiloididae,  Creedi- 
idae, and  Trachinidae.  Champsodontid  larvae  develop  a  serrate 
crest  on  the  snout  and  head  during  the  postflexion  period,  and 
chiasmodontid  larvae  (except  Kali:  R.  J.  Lavenberg,  pers.  comm.) 
develop  small  body  spicules  (Fig.  299)  just  before  or  during 
notochord  flexion. 

Pigmenlalion.  —  Pigmenlalion  of  trachinoid  larvae  is  quite  vari- 
able, from  nearly  absent  to  quite  intense  (Table  137).  Larval 
champsodontids,  mugiloidids,  trichonotids,  and  creediids  re- 
main lightly  pigmented  throughout  development,  while  larval 
trichodontids,  chiasmodontids.  trachinids,  and  uranoscopids 
may  become  rather  heavily  pigmented.  Pigmentation  usually 
increases  with  increasing  larval  size;  trichonotids  and  creediids 
change  little  in  pigmentation  with  growth. 

Head.  — Eyes  are  pigmented  at  hatching  for  the  demersally- 
spawned  Tnchodontidae,  and  for  two  of  the  six  families  with 
pelagic  eggs  (Table  137).  Pigmentation  is  present  at  hatching, 
or  subsequently  develops,  over  the  brain  in  five  families.  The 
degree  of  pigmentation  of  other  areas  of  the  head  is  variable. 

Gut.  — Pigmentation  typically  is  present  dorsally  over  the  gut 
and  swim  bladder  throughout  larval  development  (absent  only 
in  creediids  and  postflexion  trichonotids).  Other  gut  pigment  is 
variable. 


556 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  135.    Characteristics  of  Trachinoid  Eggs. 


Pelagic  (P) 

Egg 

Oil  droplets 

or 

Single 

diameter 

number:  size 

Attachments  or 

Incubation 

Family 

demersal  (D) 

or  mass 

(mm) 

range  (mm) 

ornamentation 

Pigmentation 

penod 

Source 

Trichodontidae 

D 

Mass  750- 
1,000 
eggs 

3.52 

0 

None 

Amber 

2  mo.-l 

yr- 

Breder  and  Rosen, 
1966;  Mariiave, 
1981 

Champsodontidae 

P 

Single 

1.09-1.19 

1:  0.17-0.22 

None 

Melano- 
phores  on 
embryo 
and  oil 
droplet 

McDowell,  1973c; 
Mito,  1966 

Chiasmodontidae 

P 

Single 

1.08-1.14 

1:  0.26 

None 

Chorion  rose 
to  amber 

Ahlstrom.  pers. 
comm. 

Percophididae 

Unknown 

Mugiloididae 

P 

Single 

0.77-1.25 

1:  0.16-0.25 

None 

Melano- 
phores  on 
embryo 
and  oil 
droplet 

5-6  days 

Mito,  1966;  Rob- 
ertson, 1973, 
1975a 

Trichonotidae 

Unknown 

Cheimarrhichthyidae 

P  (assumed) 

Single 
(ovarian) 

McDowell.  1973c 

Creediidae 

P 

Single 

0.70-1.10 

400-600  in 
8-12  clus- 
ters; co- 
alesce to 
3-8:  0.05- 
0.10 

None 

Melano- 
phores  on 
embryo 

2  days 

Lets,  1982;  Watson 
andLeis,  1974 

Trachinidae 

P 

Single 

0.94-1.37 

1-30,  co- 
alesce: 
0.19-O.25 

None 

4-5  days 

Breder  and  Rosen, 
1966;  Dekhnik, 
1973;  Marinaro, 
1971;  Padoa, 
1956g;  Russell, 
1976 

Uranoscopidae 

P 

Single 

1.52-2.45 

3-27:  0.02- 
0.15 

Polygonal 
network 
on  cho- 
rion in 
Uranosco- 
pus 

Melano- 
phores  on 
yolk  and 
embryo 

Dekhnik,  1973; 
Fritzsche,  1978; 
Mito,  1966;  Rob- 
ertson, 1975a 

Leptoscopidae 

Unknown 

Trunk  and  tail.  — Most  trachinoid  larvae  display  some  degree 
of  pigmentation  along  the  ventral  margin  of  the  tail  (absent  in 
some  mugiloididsand  preflexion  trachinids).  Pigmentation  (typ- 
ically rather  light)  occurs  along  the  dorsal  margin  of  the  trunk 
and  tail  at  some  time  during  larval  development  in  many  tra- 
chinoids.  Internal  pigment  may  develop  above  and  below  the 
vertebral  column  (e.g.,  Trichodontidae). 


Hypural  margin.  — Hypural  pigment  typically  is  light  or  absent 
although  its  presence  as  a  bar  is  diagnostic  for  the  Trichodon- 
tidae. 

Fins.  — Fins  typically  are  unpigmented  in  trachinoid  larvae,  al- 
though for  some  groups  fin  pigmentation  can  be  diagnostic  (e.g., 
the  caudal  and  posterior  dorsal  and  anal  fin  pigment  of  Trich- 


Table  136.    Size  (mm  SL)  of  Trachinoid  Larvae  at  Selected  Developmental  Stages. 


Prejuvenile  or 

Family 

Hatching 

Nolochord  flexion 

specialized  stages 

Juvenile 

Trichodontidae 

14.5 

Before  hatching 

None 

32-60 

Champsodontidae 

3.4-3.7 

4.6-5.0 

None 

9.6-10.7 

Chiasmodontidae 

ca.  4 

Before  ca.  9 

ca.  45 

ca.  12-45 

Percophididae 

<16.0 

Mugiloididae 

2.2-3.0 

3.7-4.8 

None 

10.0  to  >12.6 

Trichonotidae 

5.2-6.3 

None 

>18.8 

Cheimarrhichthyidae 

s25 

Creediidae 

2.6-3.5 

7.0-10.2 

None 

>11.0,  S29.2 

Trachinidae 

3.2 

5.0-10.0 

None 

13-15 

Uranoscopidae 

>2.5-4.38 

None 

£23 

Leptoscopidae 

No  information 

WATSON  ET  AL.:  TRACHINOIDEA 


557 


B 


^^S^ 


D 


Fig.  299.  (A)  Trichodontidae:  Trichodon  trichodon.  13.0  mm,  from  Marliave  (1981);  (B)  Chiasmodontidae:  Pseitdoscopelus  sp..  14.0  mm, 
CalCOFI  station  5710-5-130.80  (approximately  24°49'N,  1  16°49'W);  (C)  Percophididae:  Hemerocoeles  sp.,  16.0  mm,  redrawn  from  Crossland 
(1982);  (D)  Tnchonotidae:  Tnchonotus  sp.,  5.9  mm,  from  Leis  and  Rennis  (1983);  and  (E)  Creediidae:  Limmchlhys  donaldsom.  1 1.0  mm,  from 
Leis(1982). 


onotidae  or  the  early  developing  heavily  pigmented  pelvic  fins 
of  trachinids), 

Meristic  characters— WtrXthraX  and  fin  ray  counts  are  summa- 
rized in  Table  138.  The  sequence  of  fin  ray  formation,  incom- 
pletely described  for  most  families,  appears  to  be  quite  variable 
except  that  the  caudal  fin  is  first  to  begin  ossification  of  rays  in 
all  but  the  trachinids  (the  caudal  is  second  in  this  family,  fol- 


lowing the  pelvic  fins).  Dorsal  and  anal  fin  rays  are  second  to 
form  in  four  families  (Mugiloididae,  Tnchonotidae.  Creediidae, 
Uranoscopidae),  while  pectoral  fin  rays  are  second  in  two  (Trich- 
odontidae and  Chiasmodontidae)  and  pelvic  fin  rays  in  one 
(Champsodontidae). 

Special  structures. Sipec\a.\  structures  are  generally  lacking  in 
trachinoid  larvae.  Only  the  elongate  opercular  appendage  of 


558 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  137.     Summary  of  Pigmentatton  (Melanin  Only)  of  Larval  Trachinoid  Fishes.  Key:  +,  present;  0,  absent;  J,  increasing  with  devel- 
opment; 1,  decreasing  with  development;  0-+,  initially  unpigmented,  becommg  pigmented  with  development;  An.  anterior;  Po.  posterior. 


Eye  at 
hatching 

Head 

Gut 

Trunk 

and  tail 

Ventra 

margin 

Family 

Brain 

Jaws 

Snout 

Opercle 

Isthmus 

Nape 

An- 
terior 

Dor- 
sal 

Ven- 
tral 

Ut- 
eral 

Pre- 
flexion 

Flexion 

Trichodontidae 

+ 

+1 

+ 

+ 

0 

0 

0-  + 

0 

0 

0 

-1- 

+ 

Champsodontidae 

+ 

+ 

0 

+  1 

0 

0 

0-  + 

0 

-1- 

0-  + 

+ 

-1- 

-(- 

Chiasmodontidae 

0 

+ 

+ 

0 

0 

0-  + 

0 

0 

-1- 

0 

0 

4- 

+ 

Percophididae 
Mugiloididae 

0 

0 
0,  or 

+1,  or 

+T 

0 
0 

0 
0,  or 

+  1 

+ 
0 

0 

0 

0 
0 

0 

0,  or 

+ 

-1- 

+  T 

0 
0,  or 

+ 

0 

0-  + 

0,  or 

+ 

+ 

Trichonotidae 

Cheimarrhichthyidae' 

Creediidae 

0 
0 

0 
0 

0 
0 

0 

0 

0 

0 

0 
0 

0 
0 

0 
0 

+1 
0 

0 

+ 

0 
0 

+  Po 
+  Po 

+  Po 
+  Po 

Trachinidae 

+ 

0.  or 

+  1 

0,  or 

+T 

0,  or 

+  T 

0-+I 

0 

0,  or 

+ 

0 

+ 

0 

0-  +  I 

0,  or 

+ 

+ 

Uranoscopidae 
Leptoscopidae' 

0 

+  1 

0-  + 

+1 

0-  + 

0-  + 

0-  + 

+ 

+T 

0-+T 

or 

+  1 

0-+1 
or 

+  T 

+  T.  or 

+  1 

+  T.  or 
+  1 

'  Larvae  unknown 


Table  138.    Selected  Meristics  of  Trachinoid  Fishes. 


Family 


Dorsal  fin 


Pectoral 
fin 


Pelvic 
fin 


Pnmary 
caudal 
fin  rays 


Trichodontidae  X-XVI  +  O-I,  13-19 

Champsodontidae  V  +  I,  1 8-22 

Chiasmodontidae  VI-XIII  +  18-28         0-1,17-28       10-15        1,5 


,27-31 

21-23 

I,  5 

12-15 

12-15  +  34-40 
47-50 

,  17-20 

9-13 

1,5 

10  +  19-22  = 

29-32 

17 


33-44 


Marliave,  1981;  NWAFC, 
unpubl. 

de  Beaufort  and  Chap- 
man, 1951;  Matsubara 
et  al.,  1964;  Milo,  1962d 

Johnson  and  Cohen. 
1974;  Lavenberg,  pers. 
comm.;  Norman, 
1929 


Percophididae 

0-IX  +  14-31 

O-I, 

15-42 

20-28 

I.  5 

15 

8-9  -1-  19-21  = 
27-30 

Ginsburg,  1955;  Iwamoto 
andStaiger,  1976;  Mil- 
ler and  Jorgenson,  1973 

Mugiloididae 

IV-VII.  19-28 

1 

16-26 

15-22 

I,  5 

14-15 

10-16  +  18-22  = 
28-38 

Cantwell,  1964 

Trichonotidae 

III-VII,  40-46 

36-40 

12-14 

I,  5 

13 

15  +  40  =  55 

Leisand  Rennis,  1983; 
Masudaet  al.,  1975 

Cheimarrhichthyidae 

IV-VI,  18-21 

I-II 

14-16 

14-18 

I,  5 

12-15 

12  +  20-21  = 
32-33 

McDowall,  1973c 

Creediidae 

18-40 

25-40 

11-17 

None, 
or  I, 

3-5 

10 

37-59 

Leisand  Rennis,  1983; 
Smith,  1961 

Trachinidae 

V-VII  +  21-32 

25-36 

15 

1,5 

14 

11-12  +  23-31  = 
34-43 

Padoa,  1956g;  Russell, 
1976 

Uranoscopidae 

0-V  +  12-19 

O-I 

12-19 

13-24 

1,5 

11-14 

9-12  +  14-17  = 
25-29 

Berry  and  Anderson, 

1 96 1;  Fritzsche,  1978; 
Marshall,  1965;  Miller 
and  Jorgenson,  1973; 
Mito,  1966;  Scott  et  al., 
1974;  Smith,  1961; 
Wade,  1946 

LeptoscopicJae 

34-35 

37 

I,  5 

lO-H 

Gosline,  1968;  Scott  et 
al.,  1974 

WATSON  ET  AL.:  TRACHINOIDEA 
Table  137.    Extended. 


559 


Trunk  and  rail 

Venlral 

margin 

Dorsal  margin 

Latera! 

Fins 

Post- 

Pre- 

Post- 

Prc- 

Post- 

Hypural 

Pec- 

flexion 

flexion 

Flexion 

flexion 

flexion 

Rexion 

flexion 

margin 

Dorsal 

Anal 

toral 

Pelvic 

Caudal 

Source 

+ 

0-  + 
An 

+  1 

0 

+  T 

+ 

0-  + 

0 

0--h 

0 

0-4- 

Marliave,  1981 

+ 

+ 

+ 

+ 

0 

0,  or 

-1- 

0,  or 

+ 

0 

+  1 

+1 

0 

0 

+  1 

Mito,  1962d,  1966 

+ 

0,  or 

+ 

0,  or 

+ 

+ 

0 

0 

0 

0 

0 

0 

0 

0 

0 

Ahlstrom,  pers.  comm.; 
Lavenberg,  pers. 
comm. 

0 

0 

0 

0 

0 

0 

0 

0 

McDowall,  1973c 

+  ,  or 

+  1  Po 

0,  or 

0 

0,  or 

0,  or 

0.  or 

0 

0 

0 

0,  or 

0,  or 

0 

Leisand  Rennis,  1983; 

+  1 

+  lPo 

+  Po 

+  Po 

+  An 

+ 

-(- 

Mito,  1966;  Robert- 
son, 1973;  Watson, 
unpubl. 

+  Po 

+  Po 

+  Po 

-1-  Po 

-1-  Po 

+  Po 

+  Po 

0 

+  Po 

+  Po 

0 

0 

+ 

Leisand  Rennis,  1983 

+  Po 

+  Po 

+  Po 

0,  or 
+  Po 

0 

0 

-1- 

0--h 

+  lPo 

+  iPo 

0 

0 

+ 

Leis,  1982;  Leisand 
Rennis,  1983;  Regan, 
1916 

+ 

0,  or 
+  An 

0.  or 
+  An 

0,  or 
-1-  An 

0 

0,  or 

+  An 

-1-  An 

+ 

0-+T 

0 

0 

-1- 

0 

Ehrenbaum,  1905-1909; 
Padoa,  1956g;  Rus- 
sell, 1976 

0,  or 

+ 

0 

0 

4- 

+  An 

-1- 

-1- 

0,  or 

0,  or 

0 

0 

0 

Dekhnik,  1973;  Mito, 

+ 

+1 

+1 

1966;  Pearson,  1941 

Champsodon.  and  the  body  spicules  of  the  chiasmodontids  (ex- 
cept Kali)  are  distinctive.  Trachinus  vipera  has  precocious,  en- 
larged, and  heavily  pigmented  pelvic  fins. 

Relationships 

The  trachinoid  families  summarized  here  are  presumed  to  be 
deinved  from  the  Percoidei,  or  in  some  cases  to  belong  them- 
selves to  the  Percoidei  (e.g.,  Trichodontidae,  Champsodontidae, 
Chiasmodontidae:  Gosline,  1971).  Therefore,  in  the  following 


discussion  we  consider  early  life  history  characters  shared  with 
the  Percoidei  as  primitive.  Characters  shared  with  other  Blen- 
nioidei  are,  somewhat  arbitrarily,  considered  to  be  derived.  Our 
purpose  in  classifying  characters  into  these  categories  is  not  to 
develop  a  new  phytogeny  of  the  Trachinoidea  based  on  early 
life  history,  since  far  too  little  in  known  to  allow  such  an  un- 
dertaking, but  rather  to  determine  whether  such  characters  sup- 
port our  treatment  of  the  Trachinoidea  as  a  monophyletic  group. 
Six  of  the  1 1  trachinoid  families  retain  the  pelagic  spawning 


Table  1 39.  Summary  of  Early  Life  History  Characteristics  of  the  Trachinoidea.  The  percoid  condition  is  assumed  to  be  primitive,  while 
the  blennioid  condition  is  assumed  to  be  derived.  The  percoid  condition  includes  spawning  of  pelagic  eggs  which  soon  hatch  to  poorly-differentiated 
larvae,  a  moderately  deep  body,  myomeres  mid-to-upper  twenties,  development  of  dorsal  and  anal  fin  rays  before  pectoral  and/or  pelvic  fin  rays, 
and  five  pelvic  fin  rays.  The  blennioid  condition  includes  spawning  of  non-pelagic  eggs  with  an  extended  incubation  period  and  hatching  of  well 
developed  larvae  having  pigmented  eyes,  an  elongate  shape,  myomeres  thirty  or  more,  development  of  pectoral  and^'or  pelvic  fin  rays  before 
dorsal  and  anal  fin  rays,  and  fewer  than  five  pelvic  fin  rays.  It  should  be  understood  that  spawning  mode,  incubation  period  and  development  at 

hatching  tend  to  be  correlated,  as  are  larval  shape  and  number  of  myomeres. 


Eggs 

Larvae 

Urval 
shape 

Number  of 
myomeres 

Pectoral 
pelvic  fin 
ossification 

Reduced 
number  of 
pelvic  rays 

Pre-   • 
opercular 
spination 

Larval 

Family 

Spawning 
mode 

Incubation 
penod 

Development 
at  hatching 

pigmen- 
tation 

Trichodontidae 

Champsodontidae 

Chiasmodontidae 

Percophididae 

Mugiloididae 

Tnchonotidae 

Cheimarrhichthyidae 

Creediidae 

Trachinidae 

Uranoscopidae 

Leptoscopidae 

Derived 

Primitive 

Primitive 

Primitive 

Primitive? 
Primitive 
Primitive 
Primitive 

Derived 

Derived 

Primitive 
Derived 

Derived 
Derived 
Primitive 

Primitive 

Primitive 

Derived 

Primitive 

Derived 

Primitive 

Derived 

Derived 

Primitive 

Derived 

Derived 

Primitive 

Primitive 

Derived 

Primitive 

Derived 

Primitive 

Primitive 

Derived 

Derived 
Derived 
Primitive 

Derived 

Derived 

Derived 

Primitive 

Primitive 

Primitive 

Primitive 

Derived 

Primitive 

Primitive 
Primitive 
Primitive 
Primitive 
Primitive 
Primitive 

Derived 

Primitive 

Primitive 

Present 
Present 
Present 
Absent 
Present 
Absent 

Present 
Present 

Absent 

Heavy 

Light 

Heavy 

Light 
Light 
Light 

Light 

Heavy 

Heavy 

560 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


B 


Fig.  300.  (A)  Champsodontidae:  Champsodon  snyderi.  9.6  mm,  from  Mito  (1962a);  (B)  Mugiloididae:  Parapercis  schauinslandi.  5.3  mm,  Kahe 
Point,  Oahu,  Hawaii  (approximately  21°16'N,  ISS'S'W);  (C)  Trachinidae:  Trachinus  vipera.  7.5  mm,  redrawn  from  Schnakenbeck  (1928);  and 
(D)  Uranoscopidae:  Astroscopus  giMatus.  4.9  mm,  from  Pearson  (1941). 


WATSON  ET  AL.:  TRACHINOIDEA 


561 


mode  (Tables  135  and  139)  typical  of  the  mainne  percoids,  one 
shares  with  the  other  Blennioidei  the  condition  of  spawning  non- 
pelagic  egg  masses.  Among  the  pelagic  spawners,  four  retain  the 
percoid-like  condition  of  early  hatching  of  poorly-differentiated 
larvae;  two  share  with  the  demersal  spawners  the  condition  of 
a  relatively  long  incubation  and  hatching  of  well  developed  lar- 
vae with  pigmented  eyes. 

The  larvae  of  four  families  are  moderately  deep-bodied,  a 
character  shared  with  the  majority  of  percoids.  Each  of  these 
families  (except  Trachinidac)  contains  at  least  some  species  with 
myomeres  numbering  in  the  mid-to-upper  twenties:  typical  per- 
coid  counts.  Five  trachinoid  families  resemble  blennioids  in 
having  elongate  larvae,  usually  with  more  than  30  myomeres. 

All  trachinoid  larvae  (except  some  Trachinidae)  follow  the 
typical  perciform  pattern  of  beginning  caudal  fin  ossification 
first;  lai^ae  of  five  families  follow  the  percoid  pattern  of  begin- 
ning ossification  of  dorsal  and  anal  fin  rays  before  pectoral  and 
pelvic  fin  rays.  Four  families  share  with  the  other  blennioids 
the  early  acquisition  of  pectoral  and/or  pelvic  fin  rays.  All  tra- 
chinoid families  share  with  the  other  blennioids  the  jugular 
placement  of  pelvic  fins,  but  only  one  family  (not  all  species) 
also  shares  the  blennioid  condition  of  fewer  than  five  pelvic  fin 
rays. 


Larval  pigmentation  and  preopercular  spination  of  the  Tra- 
chinoidea  (Table  139)  are  difficult  to  assess,  since  both  range 
from  absent  to  highly  developed  in  both  the  Percoidei  and  Blen- 
nioidei. The  distribution  of  these  characters  is  listed  in  Table 
1 39  to  aid  in  determining  relationships  among  the  Trachinoidca. 

Based  solely  on  early  life  history  characters  (Table  1 39),  the 
Uranoscopidae  and  Mugiloididae  (including  Cheimarrhichthyi- 
dae?)  appear  to  be  the  most  percoid-like  members  of  the  Tra- 
chinoidca, while  Trichodontidae  are  most  like  the  other  Blen- 
nioidei. Two  points  become  clear  in  considering  the  contribution 
of  early  life  history  to  the  understanding  of  trachinoid  phylog- 
eny:  (1)  the  Trachinoidea  is  a  very  diverse,  probably  polyphy- 
letic,  group;  and  (2)  much  more  early  life  history  data  are  needed 
before  any  substantial  contribution  can  be  made  to  the  under- 
standing of  this  group. 

(W.W.)  Marine  Ecological  Consultant.s.  531  Encinitas 
BouLEVARtJ,  Suite  110,  Encinitas,  California  92024; 
(A.C.M.)  National  Marine  Fisheries  Service,  North- 
west AND  Alaska  Fisheries  Center,  2725  Montlake 
Boulevard  Ea.st,  Seattle,  Washington  98112;  (E.G.S.) 
National  Marine  Fisheries  Service,  Southwest  Fisheries 
Center,  PO  Box  271,  La  Jolla,  California  92038. 


Notothenioidea:  Development  and  Relationships 

E.  G.  Stevens,  W.  Watson  and  A.  C.  Matarese 


NOTOTHENIOIDEA  comprises  5  families  with  35  genera 
and  about  100  species  (Table  140).  These  familes  are 
endemic  to  the  Antarctic  and  Subantarctic  regions  (DeWitt, 
1971;  Norman,  1938a;  Wyanski  and  Targett.  1981).  Adults, 
ranging  from  100  to  900  mm  SL,  occupy  several  habitats  from 
the  surface  to  several  hundred  meters  depth  and  are  often  as- 
sociated with  continental  and  island  slopes  and  shelves.  Some 
species  are  adapted  for  living  close  to  the  undersurface  of  ice. 
Discussions  of  the  systematic  position  of  notothenioids  are 
found  in  Gosline  (1968)  and  Norman  (1938a),  who  consider 
them  Perciformes  or  perciform  relatives  on  the  basis  of  the 
adult  cranial  osteology;  the  jugular  position  of  the  pelvic  fins, 
which  have  one  spine  and  five  rays;  and  the  caudal  fin  ray 
number,  usually  14.  Both  note  the  reduced  number  of  pectoral 
radials  found  in  Notothenioidea.  Gosline  (1968)  unites  the  no- 
tothenioids with  trachinoids  and  blennioids  using  characters 
such  as  the  one  to  one  ratio  of  vertebrae  to  dorsal  and  anal  fin 


rays,  more  than  2  5  vertebrae,  and  fewer  than  1 5  branched  caudal 
rays.  Gosline  ( 1 968),  Norman  ( 1 938a),  and  other  recent  workers 
(i.e.,  Andersen  and  Hureau,  1979)  separate  Nototheniidae  and 
Harpagiferidae  making  a  total  of  five  families  (this  classification 
is  used  here),  whereas  Nelson  (1976)  follows  Berg  (1940)  and 


Table  141.    Notothenioidea:  Egg  Diameter  (mm)  and  Larval  Size 
at  Selected  Developmental  Stages  (mm  SL). 


Egg 

Nolochord 

Family 

diamcler 

Hatchmg 

flexion 

Juvenile 

Bovichthyidae 

Unknown 

Unknown 

Unknown 

ca.  25 

Nototheniidae 

1.2-4.0 

6-14 

9-20 

25-60 

Harpagifendae 

2.4-3.0 

7-13 

ca.  9-13 

35-38 

Bathydraconidae 

1.5-3.0 

Unknown 

18-24 

24-34  + 

Channichthyidae 

2.8-4.5 

ca.  14 

18-42 

50-60  + 

Table  140.    Notothenioidea:  General  Summary  and  Early  Life  History  Information. 


Number  of 
genera 

Approximate 

number  of 

species 

Distnbulion 

Ear 

y  life  history 

Descnplions 

Illustrations 

Famil\ 

Genera 

Species 

Species 

Bovichthyidae 

Nototheniidae 

Harpagiferidae 

Bathydraconidae 

Channichthyidae 

3 
8 
5 

10 
U 

12 
50 
15 

14 

17 

Antarctic,  Subantarctic 
Antarctic,  Subantarctic 
Antarctic,  Subantarctic 
Antarctic 
Antarctic 

0 
5 
2 
7 
8 

0 
16 
4 
8 
8 

0 

12 

3 

6 

7 

562 


ONTOGE>fY  AND  SYSTEMATICS  OF  HSHES-AHLSTROM  SYMPOSIUM 


Fig.  301.  Notothenioid  larvae  (from  top  to  bottom):  Nototheniidae:  Patagonololhcn  larsent.  35  mm.  from  North  and  While  (1982);  Harpa- 
giferidae:  Harpagifer  bispinis,  18.2  mm,  from  Everson  (1968);  Harpagiferidae:  Artedidraco  inints.  24.0  mm.  from  Efremenko  (1983);  Bathydra- 
conidae:  Psilodraco  breviceps.  16.9  mm,  from  Efremenko  (1983);  and  Channichthyidae:  Pagetopsis  macroplerus,  19  mm,  redrawn  by  H.  Orr  from 
Regan  (1916). 


STEVENS  ET  AL.:  NOTOTHENIOIDEA 


563 


Table  142.    Notothenioidea:  Selected  Meristics.  Sources  listed  here  do  not  include  the  following  cited  in  the  text:  Andriashev  (1959);  DeWitt 

(1970);  Norman  (1937,  1938);  Nybelin  (1947,  1951);  Regan  (1913d.  1916);  Yefremenko  (1979a,  b).  Omissions  indicate  no  data  were  found  in 

the  literature.  While  this  paper  was  in  press  a  revision  of  the  Nototheniidae  was  published  (Anderson,  1984). 


Fin  rays 

Vertebrae 

Number    _ 
of 

Family 

Pnncipal           Pre- 

Genus 

species 

Di 

D; 

A 

Pec. 

Pel. 

caudai           caudal 

Caudal 

Tolal 

Sources 

Bovichthyidae 

Bovichlhys 

10 

VII-VIII 

18-21 

13-19 

14-16 

1.4-5 

41 

(Isp.) 

Hureau  and 
Tomo,  1977 

Colloperca 

1 

VI-VIII 

21-24 

20-24 

17 

Pseudophrites 

1-2 

VII-VIII 

19-20 

23-25 

18 

1.5 

Scott,  1962 

Nototheniidae 

Dissostichus 

2 

VIII-X 

25-29 

25-30 

25-29 

1,5 

18-20 

54-55 

de  Ciechomski 
and  Weiss, 
1976 

Eleginops 

1 

VIII-IX 

24-26 

22-24 

22-24 

1,5 

12 

30-31 

42-43 

de  Ciechomski 
and  Weiss, 
1976 

Pleuragramma 

1 

Vl-VIII 

33-38 

36-39 

20-21 

1,5 

18-20 

53-55 

Aelhotaxis 

1 

VII 

34 

30 

27-28 

1.5 

16 

36 

52 

DeWitt,  1962b 

Cryolhenia 

1 

IV-VI 

34-36 

33-35 

24-26 

1,5 

14-17       13-18 

33-37 

50-53 

Daniels,  1981 

Pagothenia 

2 

IV-VI 

30-37 

26-33 

23-24 

1,5 

14 

DeWitt,  1964b 

Palagonololhen 

16 

IV-VIII 

27-38 

27-37 

24-28 

1.5 

54-55 
(Isp.) 

Notothenia 

ca.  27 

III-VIII 

25-41 

21-36 

16-30 

1,5 

14-16       14-22 

27-37 

44-58 

Harpagiferidae 

Artedidraco 

5 

II-IV 

24-30 

16-21 

14-17 

1.5 

Dollwdraco 

1 

II-III 

22-25 

14-15 

16-18 

1.5 

13 

21-22 

34-35 

DeWitt  and 

Tyler,  1960 

Histiodraco 

1 

11-111 

26 

17 

18-19 

1,5 

Pogonophryne 

7 

11-111 

25 

16-18 

19  (Isp.) 

1,5 

15 

23 

38  (Isp. 

) 

Harpagtfer 

1 

Ill-V 

21-26 

16-21 

15-18 

1,5 

10-11 

24-25 

35-36 

Bathydraconidae 

Balhydraco 

4 

0 

34-40 

29-31 

-)T 

I,  5 

52-53 

Gerlachea 

1 

0 

45-47 

34-35 

26-28 

1,5 

25 

38 

63 

Racoviizia 

1 

0 

30-37 

27-31 

22-25 

1.5 

19-20 

32-34 

52-54 

DeWitt.  1964a 

Phonodraco 

I 

0 

34-38 

29-33 

21-22 

16 

33-34 

49-50 

Cygnodraco 

1 

0 

61-66 

35-38 

22-23 

Parachaenichthys 

2 

0 

42-46 

30-33 

21-23 

Psilodraco 

1 

0 

27-30 

27-29 

25-27 

16 

48 

Gymnodraco 

1 

0 

28-30 

22-26 

21-22 

I,  5 

13         20-21 

28 

48-50 

DeWitt  and 

Tyler.  1960 

Akarota.xis 

1 

0 

31-33 

27-28 

22-23 

1.5 

12-13          17 

33 

50 

DeWitt  and 
Hureau,  1979 

Vomeridens 

1 

0 

34 

32 

31-33 

1,5 

12             18 

36 

54 

DeWitt  and 
Hureau,  1979; 
DeWitt,  1964a 

Channichthyidae 

Champsocephalus 

2 

VII-XI 

32^1 

31-tO 

22-29 

1,5 

16-18 

60-64 

Kock,  1981 

Pagetopsis 

2 

IX-XV 

26-31 

24-27 

22-23 

1,5 

18-20 

56-58 

Pseudochaenichthys 

VIII-XI 

28-32 

27-31 

22-25 

1.5 

16-18 

51-55 

Kock,  1981 

Neopagetopsis 

XIV 

34 

29 

26 

Dacodraco 

III 

32 

29 

24 

1,5 

Chaemchthys 

VI-IX 

30-35 

27-34 

18-22 

1,5 

54-56 

Chaenocephalus 

V-IX 

37-42 

36-40 

23-26 

1,5 

18-20 

60-64 

Kock.  1981 

Cryodraco 

III-V 

40-45 

39-46 

23-25 

1,5 

20-22 

63-65 

Chionodraco 

V-VII 

37-51 

33-38 

21-24 

1,5 

16-18 

60-64 

Chaenodraco 

VI-VIII 

38-42 

32-35 

22-25 

1,4 

18-20 

60-62 

Chionobathyscus 

564 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Greenwood  et  al.  ( 1 966)  in  placing  Harpagiferinae  as  a  subfam- 
ily of  Nototheniidae  (Table  140). 

Development 

The  first  work  on  the  early  life  history  of  these  fishes  was 
undertaken  with  material  collected  on  the  polar  expeditions  of 
the  early  20th  century.  Regan  (1916)  illustrated  larvae  of  seven 
notothenioid  species.  Additional  early  life  history  data  were 
sparse  until  the  Antarctic  expeditions  of  the  second  half  of  the 
century.  In  the  last  30  years  efforts  have  been  directed  toward 
understanding  the  biology,  ecology,  population  dynamics,  and 
physiological  adaptations  of  these  fish.  In  these  investigations 
some  early  life  history  data  have  been  acquired.  Larvae  of  36 
species  have  been  described  (Table  140).  The  most  comprehen- 
sive summaries  are  the  key  in  North  and  White  (1982),  the 
papers  of  Yefremenko  (1979b,  c)  and  the  atlas  of  Efremenko 
(1983a).'  No  early  life  history  data  are  available  for  the  family 
Bovichthyidae  except  for  a  brief  description  of  the  behavior  and 
an  illustration  of  a  25  mm  prejuvenile  of  Bovichthys  variegatus 
(Robertson  and  Mito,  1979). 

Eggs 

Eggs  of  four  notothenioid  families  have  been  described  in- 
cluding some  known  only  from  studies  of  ovaries  (Table  141). 
Eggs  are  moderate  to  large  (1.2-4.5  mm  diameter)  with  large 
yolks,  no  oil  globules,  and  small  perivitelline  spaces  (Marshall, 
1953;  Andriashev,  1965;  Dearborn,  1965).  In  one  species,  No- 
tothenia  ( Trematomus)  hernacchii.  eggs  are  bright  yellow  to  deep 
brown.  Most  species  are  demersal  spawners;  nesting  behavior 
has  been  observed  in  N.  hernacchii  (Moreno,  1980)  and  Har- 
pagifer  hispinis  (Daniels,  1978).  However,  Notothenia  micro- 
lepidola  eggs  have  been  reported  from  plankton  collections 
(Robertson,  1975a).  The  demersal  eggs  are  sticky,  clinging  to 
substrate  or  algae.  One  species,  N.  neglecta.  reared  in  the  lab- 
oratory from  artificially  fertilized  eggs,  has  an  incubation  time 
of  103-150  days  and  hatches  with  a  well-developed,  heavily 
pigmented  body,  black  eyes,  and  a  large  yolk  sac  (White  et  al., 
1982).  Daniels  (1978)  reports  an  incubation  time  of  14  to  18 
weeks  for  H.  hispinis. 

Larvae 

Morphology— The  described  larvae  of  36  species  display  some 
morphological  similarity  (snout-anus  length),  and  some  diver- 
sity (snout  length  and  body  shape).  Preflexion  larvae,  ca.  6-18 
mm  SL,  are  elongate  with  large  pectoral  fins  and  moderate  to 
wide  finfolds.  Channichthyid  larvae  have  well  developed  pelvic 
fins  at  this  stage  and  more  elongate  snouts  than  larvae  of  other 
notothenioids.  Some  species  have  large  yolk  sacs  which  persist 
after  notochord  flexion  has  begun.  Preanal  lengths  range  from 
slightly  under  to  slightly  over  50%  of  body  length. 

Dunng  flexion  and  postflexion  stages  most  larvae  maintain 
their  elongate  shape  (Fig.  301).  However  the  larvae  of  the  har- 
pagiferid  genera  Artedidraco  and  Pogonophryne  become  very 
robust  (North  and  White,  1982;  Efremenko,  1983a).  Notochord 
flexion  occurs  between  9  and  42  mm  with  larval  Harpagiferidae 
and  Nototheniidae  flexing  at  the  shortest  lengths,  Channichthyi- 
dae  at  the  longest,  and  Bathydraconidae  at  intermediate  lengths 
(Table  141).  Size  at  transformation  to  the  juvenile  stage  also 
spans  a  wide  range  with  Harpagifer  hispinis  settling  at  18.3  mm 


'  Efremenko  and  Yefremenko  are  alternative  transliterations  of  the 
name  of  the  same  author. 


(Everson,  1 968)  and  pelagic  larvae  of  other  species  reaching  24- 
60  mm  (de  Ciechomski  and  Weiss,  1976;  North  and  White, 
1982;  Yefremenko,  1979b,  c). 

Larvae  of  all  species  develop  pelvic  fins.  Channichthyid  larvae 
retain  their  elongate  snouts  and  develop  teeth  and  preopercular 
and  rostral  spines  not  reported  for  other  notothenioids  (Fig. 
301). 

Pigmentation.  — Pi%menx  patterns  of  all  known  larvae  are  highly 
specific  and  are  useful  identification  criteria.  The  amount  and 
location  of  pigment  varies  within  families  and  the  amount  usu- 
ally increases  with  development.  A  few  species  have  general 
body  pigment,  but  most  exhibit  patterns  in  one  or  more  of  the 
following  areas:  dorsal  body  margin,  ventral  body  margin,  body 
midline,  peritoneum,  gut,  and  along  the  myosepta.  The  occipital 
and  parietal  areas  typically  are  pigmented;  many  species  have 
snout,  opercular,  and  jaw  pigment  as  well.  The  paired  fins  are 
usually  pigmented.  Pigment  is  found  at  the  base  of  the  caudal 
fin  in  most  species,  associated  with  the  posterior  margin  of  the 
hypural  elements  or  the  bases  of  the  caudal  rays. 

Meristics.  —  Meristics  are  from  counts  given  for  adults  by  Regan 
(1913d,  1916),  Norman  (1937,  1938a),  Nybelin  (1947,  1951), 
Andriashev  (1959),  and  DeWitt  (1970)  (Table  142).  Vertebral 
counts  are  especially  useful  diagnostic  features  within  and  be- 
tween families.  The  dorsal,  anal,  pectoral,  and  vertebral  counts 
have  been  the  most  significant  characters  linking  larvae  to  adults 
(Yefremenko,  1979b,  c).  The  sequence  of  fin  formation  is  the 
same  in  Nototheniidae,  Harpagiferidae,  and  Bathydraconidae 
with  pectoral  and  caudal  fins  forming  first,  followed  by  pelvics, 
with  dorsal  and  anal  last  to  ossify.  In  Channichthyidae  the  pel- 
vics are  precocious  and  are  present  in  yolk-sac  larvae. 

Relationships 

Knowledge  of  the  early  life  history  of  Notothenioidea  has  not 
contributed  to  understanding  relationships  between  Blennioidei 
and  other  perciform  suborders,  but  does  ofler  some  clues  to 
relationships  within  the  suborder.  The  lengthy  ovarian  egg  de- 
velopment (Dearborn,  1965;  Everson,  1970)  is  probably  related 
to  the  cold  environment.  In  other  aspects  of  spawning,  i.e., 
nesting  behavior,  long  incubation  time,  and  laying  of  demersal 
adhesive  eggs,  this  infraorder  resembles  other  cold-water  blen- 
nioids.  The  well  developed  state  of  newly  hatched  larvae  and 
the  sequence  of  fin  development  as  well  as  the  general  lack  of 
specialized  larval  structures  are  also  blennioid  features.  Further 
study  of  developmental  characters,  such  as  the  sequence  of  os- 
sification, might  contribute  to  better  understanding  of  the  re- 
lationships among  the  Blennioidei.  Superficial  morphological 
and  meristic  resemblances  exist  among  notothenioid  larvae  and 
those  of  other  blennioid  species,  for  example,  the  notothenioid 
Patagonotothen  larscni  (Fig.  301),  the  trachinoid  Trichodon 
trichodon  (see  Trachinoidea,  this  volume)  and  the  blennioid 
Heterostichus  rostratus  (see  Blennioidea,  this  volume).  As  re- 
lationships among  the  blennioids  become  better  known,  their 
relationship  with  other  perciforms  might  become  clearer. 

(E.G.S.)  National  Marine  Fisheries  Service,  Southwest 
Fisheries  Center,  PO  Box  271,  La  Jolla,  California 
92038;  (W.W.)  Marine  Ecological  Consultants,  531 
Encinitas  Boulevard,  Encinitas,  California  92024; 
(A. CM.)  National  Marine  Fisheries  Service,  North- 
west AND  Alaska  Fisheries  Center,  2725  Montlake 
Boulevard  East,  Seattle,  Washington  981 12. 


Blennioidea:  Development  and  Relationships 
A.  C.  Matarese,  W.  Watson  and  E.  G.  Stevens 


THE  Blennioidea  is  composed  of  16  families  with  about  182 
genera  and  759  species  (Table  143).  The  families  discussed 
here  are  those  included  in  the  infraorder  Blennioidea  by  Nelson 
(1976),  as  amended  by  the  current  literature.  For  convenience 
we  divide  the  infraorder  into  a  tropical  and  a  northern  group. 
The  tropical  group  is  similar  to  Gosline's  ( 1 968)  superfamily 
Blennioidae  except  for  the  following:  1)  Ophiclinidae  and  Per- 
onedysidae  are  synonymized  with  the  Clinidae  (George  and 
Springer,  1980);  2)  Dactyloscopidae  is  included  (George  and 
Springer,  1980);  and  3)  Congrogadidae  is  excluded  (Winterbot- 
tom,  1982').  The  northern  group  is  similar  to  Gosline's  (1968) 
superfamily  Zoarceoidae  except  that  we  include  the  Bathymas- 
teridae  (Anderson,  1984).  The  Zoarcidae  is  treated  separately 
(Anderson,  this  volume). 

The  majority  of  species  (80%)  belong  to  four  tropical  families: 
Tripterygiidae,  Clinidae,  Labrisomidae,  and  Blenniidae.  Of  the 
northern  forms,  only  the  family  Stichaeidae  represents  a  sig- 
nificant percentage  (8%)  of  the  species.  Tropical  Blennioidea 
inhabit  primarily  the  Indo-West  Pacific  south  to  Australia,  while 
northern  fishes  inhabit  the  North  Atlantic  and  North  Pacific 
(Table  143).  Occasionally,  representatives  of  mainly  tropical 
families  occur  in  boreal  waters  (e.g.,  Clinidae  and  Blenniidae), 
and  members  of  northern  families  may  occur  in  the  subtropics. 
Some  dactyloscopids  inhabit  fresh  water.  Four  families  are 
monotypic  and  three  of  these  are  endemic  to  the  northeast  Pa- 
cific. 

As  a  group  most  of  the  16  families  in  Blennioidea  are  not 
well  understood,  probably  due  to  their  lack  of  commercial  im- 
portance, small  size  and  cryptic  habits.  In  general,  the  tropical 
and  more  speciose  families  (e.g.,  Blenniidae)  are  better  known 
than  the  northern  families.  Monotypic  families  are  quite  poorly 
known.  Although  sparse  and  incomplete,  some  early  life  history 
information  is  available  for  II  of  the  16  families  (Table  143). 
In  most  cases,  however,  the  data  on  few  species  may  not  be 
representative  of  the  family.  Among  the  families  in  the  infraor- 
der, the  Blenniidae  has  the  greatest  number  of  species  (22)  de- 
scribed; but  with  about  319  species  in  the  family,  this  amounts 
to  fewer  than  10%.  Morphology,  pigment,  and  meristics  of  lar- 
vae in  the  infraorder  are  diverse  (Figs.  302,  303). 

Development 

Eggs 

Fishes  in  this  infraorder  spawn  demersal  eggs  (Table  144), 

except  for  some  clinids.  Clinids  of  the  tribe  Ophiclinini  are 

ovoviviparous  (George  and  Springer,  1980),  while  those  of  the 

tribe  Clinini  are  viviparous  (Penrith,  1969;  Hoese,  1976). 

Most  blennioid  eggs  are  spherical  to  somewhat  flattened,  pos- 
sess one  to  several  oil  droplets,  are  attached  to  one  another  (and 
often  to  a  substrate)  by  filaments  or  other  adhesions,  and  have 


'  Winterbottom,  R.  1982.  The  perciform  fish  family  Congrogadidae— 
biogeography  and  evidence  for  monophyly.  Amer.  Soc.  Ich.  Herp.,  oral 
paper,  62nd  annual  meeting. 


a  smooth  unsculptured  chorion.  Sizes  range  from  among  the 
smallest  offish  eggs  (Blenniidae,  0.50  mm)  to  among  the  largest 
(Anarhichadidae,  8.0  mm).  Incubation  periods  range  from  6  to 
70  days.  Eggs  are  unknown  for  four  families:  Xenocephalidae, 
Ptilichthyidae,  Zaproridae,  and  Scytalinidae. 

Parental  care  is  common  among  most  families;  e.g.,  in  sti- 
chaeids.  males  or  females  may  guard  egg  masses  (Shiogaki  and 
Dotsu,  1972a;  Shiogaki,  1981,  1982).  In  an  extreme  example 
of  parental  care,  male  dactyloscopids  incubate  eggs  in  ball-like 
clusters  carried  beneath  the  pectoral  fins  (Dawson,  1982). 

Larvae 

Morphological  characters— ^Xenmoidea  larvae  hatch  at  sizes 
ranging  from  as  small  as  2.0  mm  (Blenniidae)  to  as  large  as  1 7.0 
mm  (Anarhichadidae)  (Table  145).  Larvae  of  the  northernmost 
families  hatch  at  more  than  twice  the  size  of  larvae  of  the  more 
tropical  families  (i.e.,  averaging  ca.  1 1.5  mm  versus  ca.  4.5  mm). 
Size  at  which  notochord  flexion  is  complete  is  also  variable,  but 
tropical  larvae  are  usually  fully  flexed  by  ca.  10.0  mm  whereas 
northern  larvae  do  not  complete  flexion  until  ca.  20.0  mm.  At 
least  three  families  have  larvae  with  an  extended  pelagic  exis- 
tence: Blenniidae,  Cryptacanthodidae,  and  Zaproridae.  Mem- 
bers of  the  blenniid  tribe,  Salariini,  have  the  only  well-docu- 
mented, prejuvenile  pelagic  stage  (Miller  et  al..  1979;  Leis  and 
Rennis,  1983).  This  has  been  termed  the  "ophiblennius""  stage 
and  usually  occurs  between  4.6  and  26.0  mm  (Fig.  302).  At  least 
two  families,  Cryptacanthodidae  (Shiogaki,  1982)  and  Zapror- 
idae (Haryu  and  Nishiyama,  1981),  have  heavily  pigmented 
larvae  and  prejuveniles  that  are  extensively  collected  in  surface 
nets  suggesting  an  extended  pelagic  existence  (Fig.  303C,  G). 
Most  blennioid  larvae  do  not  undergo  a  marked  metamorphosis. 
Transformation  is  usually  complete  in  tropical  forms  by  26.0 
mm,  but  may  begin  as  early  as  10.0  mm  in  some  families  (Trip- 
terygiidae and  Blenniidae).  Larvae  in  the  more  northern  families 
transform  at  a  slightly  larger  size,  ca.  30.0-40.0  mm,  although 
Ptilichthys  transforms  at  ca.  1 14.0  mm. 

Among  the  tropical  families,  larval  Tripterygiidae,  Clinidae, 
and  Labrisomidae  share  many  similar  morphological  features. 
They  are  moderately  elongate,  have  a  preanal  length  about  50% 
BL  (slightly  less  in  labrisomids),  possess  a  large  swimbladder, 
and  usually  lack  preopercular  spines  (Figs.  302A,  B,  C,  D). 
Heads  are  small,  sometimes  rounded,  with  a  short  snout.  Mouths 
extend  just  beyond  the  anterior  eye  margin.  In  tripterygiid  and 
clinid  larvae,  the  gut  is  initially  straight  but  coils  during  flexion. 

The  blenniids  include  many  larval  forms  with  diverse  mor- 
phological features.  According  to  Leis  and  Rennis  (1983),  how- 
ever, larvae  are  more  similar  within  tribes  than  between  tribes. 
Most  species  are  moderately  elongate  (Nemophini  includes  both 
slender  and  robust  forms),  becoming  either  more  slender  (Nem- 
ophini) or  more  robust  (Salariini)  with  development.  Heads  are 
short,  rounded,  and  broad  becoming  more  elongate  with  de- 
velopment (except  Salariini  larvae  in  which  the  snout  elongates 
early  in  the  preflexion  stage).  The  gut  is  short  to  moderate  (usu- 
ally <  50%  BL),  and  eventually  coiled  if  not  so  initially.  L-arval 
preopercular  spination  may  be  elaborate:  spines  can  be  numer- 


565 


566 


ONTOGENY  AND  SYSTEMATICS  OF  HSHES-AHLSTROM  SYMPOSIUM 


Table  143.    General  Summary  and  Early  Life  History  (ELH)  Information  in  Blennioidea. 


Early  life  history 


Tax  on 


Approx. 
Number  of      number  of 
genera  species 


Primary  Dislnbution 


Number      Number 

of  of 

genera         species 


Number 
of 

species 
illus- 
trated 


Pnmao'  early  life  history  sources 


Blennioidea 
Dactyloscopidae  6 

Xenocephalidae  1 

Notograptidae  2 

Tripterygiidae  18-19 


Clinidae  26 

Chaenopsidae  1 0 

Labrisomidae  1 4 

Blenniidae  S3 


20  Atlantic,  Pacific 

(tropical) 
1  New  Ireland,  New  Guinea 

3  Australia 

75-95         Atlantic,  Pacific. 
Indian  (tropical) 


85  Atlantic,  Pacific, 

Australia 


55 
100 
289-319       Indo-Pacific 


Atlantic.  Pacific 

(tropical) 
New  World  (tropical) 


0 

0 

2 

3 

7 

22 

Bathymasteri 

dae 

3 

7 

North  Pacific 

2 

3 

Stichaeidae 

37 

54 

North  Atlantic,  Pacific 

14 

18 

Cryptacanthodidae 

3 

4 

North  Atlantic,  Pacific 

3 

3 

Pholidae 

4 

13 

North  Atlantic,  Pacific 

3 

5 

Anarhichadidae 

2 

6 

North  Atlantic,  Pacific 

2 

3 

Ptilichthyidae 

I 

1 

Northeast  Pacific 

1 

1 

Zaproridae 

1 

1 

Northeast  Pacific 

1 

1 

Scytalinidae 

1 

1 

Northeast  Pacific 

0 

0 

Dawson  1982 


7        Graham,  1939;  Leis  and  Rennis,  1983; 
Miller  etal.,  1979;  Ruck,  1973a, 
1980;  Shiogaki  and  Dotsu,  1973; 
Watson,  unpubl.;  Wirtz,  1978 

14  Bamhart.  1932;  Padoa,  1956h;  Shio- 

gaki and  Dotsu,  1972b;  Sparta, 
1948;  Stevens,  unpubl.;  Watson,  un- 
publ. 
Bohlke,  1957;  Stephens  et  al.,  1966 

3         Breder,  1939;  Breder,  1941;  Springer. 
1958;  Watson,  unpubl. 
27         Cipria,  1934.  1936;  Dotsu.  1982;  Dot- 
su and  Monuchi,  1980;  Dotsu  and 
Oota,  1973;  Dutt  and  Rao,  1960; 
Eggert,  1932;  Fishelson,  1963,  1976; 
Fives,  1970a;  Ford,  1922;  Fritzsche, 
1978;  Hildebrand  and  Cable,  1938; 
Lebour,  1927;  Leis  and  Rennis. 
1983;  Lippson  and  Moran,  1974; 
Miller  etal,.  1979;  Mito,  1954; 
Munro,  1955;  Peters.  1981;  Qasim, 
1956;  Rao.  1970;  Russell,  1976;  Ste- 
vens and  Moser,  1982;  Thomson  and 
Bennett.  1953;  Watson,  1974.  un- 
publ.; Wickler.  1957 

1  Breder  and  Rosen,  1966;  Fitch  and 

Lavenberg,  1975;  Matarese,  unpubl. 

15  Breder  and  Rosen.  1966;  Faber,  1976; 

Hart,  1973;  Marliave,  1975;  Ma- 
tarese, unpubl.;  Peppar,  1965;  Rass, 
1949;  Russell.  1976;  Shiogaki.  1981; 
Shiogaki.  1983;  Shiogaki  and  Dotsu, 
1972a;  Tokuya  and  Amaoka,  1980; 
Wourms  and  Evans,  1974 

2  Hart,  1973;  Matarese,  unpubl.,  Shio- 

gaki, 1982 

3  Breder  and  Rosen,  1 966;  Marliave, 

1975;  Rass,  1949;  Sawyer,  1967; 

Tokuya  and  Amaoka,  1980 
2         Andriyashev,  1954;  Barsukov,  1959; 

Breder,  1941;  Kobayashi,  1961a; 

Marliave,  1975;  Rass,  1949 
1         Kobayashi,  1961b;  Richardson  and 

Denhart,  1975 
I         Chapman  and  Townsend,  1938;  Haryu 

and  Nishiyama,  1981 


ous  or  large  (Blenniini  and  Omobranchini)  or  completely  lacking 
(Nemophini).  Teeth  develop  early  in  most  species;  these  become 
large  (Nemophini)  or  hooked  (Salariini)  (Fig.  302E).  Cirri  may 
develop  at  the  end  of  the  larval  period.  Members  of  the  Salariini 
have  elongate  pectoral  fins  (Fig.  302F). 

Larvae  of  the  northern  families  have  an  elongate  body  shape. 


but  they  range  from  moderately  elongate  (Zaproridae)  to  ex- 
tremely long  and  thin  (Ptilichthyidae)  (Fig.  303).  Heads  are 
small,  and  initially  pointed  or  rounded  but  become  more  point- 
ed with  development.  Most  species  have  a  short  to  moderate 
snout.  Preanal  length  is  highly  variable.  Generally,  preanal  length 
is  at  least  50%  BL,  but  it  ranges  from  short  (<50%  BL  in  pre- 


MATARESE  ET  AL.:  BLENNIOIDEA 


567 


Table  144.     Summary  of  Egg  Characters  in  Blennioidea.  Blanks  indicate  data  are  unavailable. 


Egg  Single 

lype'  or  mass 


Egg  diameter 
(mm) 


Number  of 
oil  globules 


Altachmenl 

processes  or 

ornamentation 


Pigmentation 


Incubation 
(days) 


Primary'  sources 


Blennioidea 
Dactyloscopidae 
Xenocephalidae 
Notograptidae 
Tripterygiidae 


Clinidae 


D 


D        Mass 


D^ 


0.90-1.40 


0.96-1.7 


Few  to 
numerous 


Several 


Filaments  at 
one  pole  or 
everywhere 


Filaments 


Embryo,  yolk 
with  "pig- 
ment 
spheres'" 


Chaenopsidae 

D 

Labnsomidae 

D 

Mass 

1.15-1.33 

1-6 

Attach  to  each 
other, 
strands 

Embryo,  yolk 

Blenniidae 

D 

Mass 

0.58-1.6 

X 

0.4-0.96 

0-several 

Adhesive  disk 
or  pedestal 

Embryo,  yolk 

Bathymasteridae 


Stichaeidae 


D 


D 


Mass 


Mass 


0.99-1.1 


1.37-2.5 


Non-adhesive 
mass 


Adhesive 


Cryptacanthodidae 
Pholidae 


Anarhichadidae 


P^ilichthyidae 

Zaproridae 

Scytalinidae 


D 
D 


D 


Mass 


Loose  or 
clumps 


1.8 
1.4-3.0 


4.0-8.0 


Adhesive 


Dawson,  1982 


16-22      Graham,  1939;  Miller 
et  al.,  1979;  Ruck, 
1973a;  Ruck.  1980; 
Shiogaki  and  Dotsu, 
1973 

12-40      Bamhart,  1932;  Pa- 

doa,  1956h;  Shiogaki 
and  Dotsu,  1972b; 
Sparta,  1948;  Ste- 
vens, unpubl. 
Stephens  et  al.,  1966 
10         Breder,  1939 


6-6  P     Cipria,  1934,  1936; 

Dotsu,  1982;  Dotsu 
and  Moriuchi, 
1980;  Dotsu  and 
Oota,  1973;  Dutt 
and  Rao,  1960;  Eg- 
gert,  1932;  Fishel- 
son  1963,  1976; 
Fritzsche,  1978; 
Hildebrand  and  Ca- 
ble, 1938;  Lebour, 
1927;  Mito,  1954; 
Munro,  1955;  Peters, 
1981;  Qasim,  1956; 
Rao,  1970;  Stevens 
and  Moser,  1982; 
Thomson  and  Ben- 
nett, 1953;  Watson, 
unpubl.;  Wickler, 
1957 
13-15      Breder  and  Rosen, 
1966;  Fitch  and 
Lavenberg,  1975; 
Matarese,  unpubl. 
2 1  Breder  and  Rosen, 

1966;  Hart,  1973; 
Marliave,  1975; 
Matarese,  unpubl.; 
Peppar,  1965;  Shio- 
gaki, 1983;  Wourms 
and  Evans,  1974 

Hart,  1973 
42-70      Breder  and  Rosen, 
1966;  Marliave, 
1975;  Matarese,  un- 
publ.; Sawyer,  1967 

Barsukov,  1959;  Bred- 
er and  Rosen,  1966; 
Matarese,  unpubl. 


I  D  =  demersal. 

^  Ophichni-ovoviparous  George  and  Spnnger  (1980).  Clmmi-viviparous  (Pennth,  1969:  Hoese.  1976). 

'Usually  7-14  days. 


568 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  145.     Summary  of  Larval  Size  at  Selected  Developmental  Stages  in  Blennioidea  (mm  SL).  Blanks  indicate  data  are  unavailable. 


Taxon 

Hatching 

Notochord  flexion 

Special  prejuvenile 

Juvenile 

Blennioidea 

Dactyloscopidae 

Xenocephalidae 

Notograptidae 

Tripterygiidae 

2.7-6.1 

4.8-<9.4 

None 

>11.0 

Clinidae 

5.5-6.7 

by  11.4-14.3 

None 

> 16-25 

Chaenopsidae 

>16-17 

Labrisomidae 

4.1 

4.9-6.9 

None 

> 19-25 

Blenniidae 

2.0-5.4 

3.6-10.75 

4.6-26.0  Salariini 

6.4-26.0 

Bathymasteridae 

5.5-6.0 

<10.0 

None 

ca.  40.0 

Stichaeidae 

6.5-12.5  (8-9.5) 

12.0-20.0(13-15) 

None 

>25.0 

Cryptacanthodidae 

10.0-11.0 

<18.0 

Neustonic 

ca.  30.0 

Pholidae 

10.0-12.5 

ca.  19.0-30.0 

None 

>30.0 

Anarhichadidae 

17.0-18.0 

<20.0 

None 

ca.  40.0 

Ptilichthyidae 

<20.0 

None 

ca.  114.0 

Zaproridae 

<12.0 

<17.0 

Neustonic 

Scytalinidae 

flexion  Stichaeidae)  to  very  long  (Pholidae)  (Fig.  303B.  D).  The 
family  Anarhichadidae  includes  one  genus  with  a  long,  thin 
bodied  larva  (Anarrhtchthys)  and  another  with  only  a  moder- 
ately elongate  larva  (Anarhichas)  (Fig.  303E).  The  monotypic 
family  Ptilichthyidae  has  a  unique  larval  form  —  it  is  highly  elon- 
gate with  a  small  head  and  extended  postanal  body  (Fig.  303n. 

Pigmentation  characters.  —  Pigmentation  is  typically  sparse  for 
most  families  within  this  infraorder,  and  tends  to  be  added 
subcutaneously  with  development.  However,  four  families  [Za- 
proridae, Anarhichadidae,  Cryptacanthodidae,  and  some  Nem- 
ophini  (Blenniidae)]  have  larvae  with  dense  body  pigment  that 
is  not  typical  of  the  Blennioidea  (Figs.  303C,  E,  and  G).  Im- 
portant pigment  areas  are  along  the  ventral  body  midline  and 
in  the  gut  area  (Table  146). 


Head. —Eyes  are  pigmented  prior  to  hatching  in  all  known  groups. 
Pigment  is  generally  absent  or  light  during  notochord  flexion 
but  usually  increases  dorsally,  over  the  brain,  by  the  time  flexion 
occurs.  Additionally,  postflexion  larvae  may  have  pigment  on 
the  snout,  mouth,  and  in  the  opercular  area. 

Gut.  — Preflexion  larvae  in  most  species  have  peritoneal  and 
some  dorsolateral  pigment.  In  families  that  have  a  gas  bladder 
(e.g.,  Tripterygiidae,  Labrisomidae,  and  Blenniidae),  pigment  is 
present  on  its  dorsal  surface  (Fig.  302A,  D).  Ventral  pigment 
may  or  may  not  be  present.  During  notochord  flexion,  pigment 
increases  on  the  lateral  gut  surface,  and  becomes  subcutaneous 
in  postflexion  larvae. 

Trunk  and  tail.  — This  is  the  most  important  pigment  area  for 


Table  146.    Summary  of  Some  Pigmentation  Characters  in  Larvae  of  Blennioidea.  Key:  D,  dorsal;  A,  anal;  P,  pectoral;  V,  ventral;  C, 
caudal;  +,  present;  O,  absent;  An,  anterior;  Po,  posterior;  \.  increasing;  j,  decreasing;  -,  with  development;  and  O  -  +,  unpigmented  initially, 

becoming  pigmented  with  development. 


Head 

Gut 

Taxon 

Eye  at 
hatch- 
ing 

Brain 

Jaw 

Snout 

Oper- 
cle 

Isthmus 

Nape 

Antenor 

Doi^al 

Ventral 

Lateral 

Tripterygiidae 

+ 

+ 

O 

o 

O 

o 

O 

o 

+ 

O,  +Po 

o 

Clinidae 

+ 

o 

o 

o 

O 

o 

O  -  + 

o 

+ 

+  Po 

o 

Chaenopsidae 

o 

o 

o 

O 

o 

O 

o 

o 

O 

o 

Labrisomidae 

o 

o 

o 

o 

o 

o 

o 

+ 

+ 

o 

Blenniidae 

+ 

o, -h-T 

o 

o 

o 

o 

o 

o 

+  T 

0-+1 

+  -I 

Bathymasteridae 

+ 

O-  -1- 

o 

o 

o 

+ 

o 

o 

+ 

+ 

o 

Stichaeidae 

+ 

o,  + 

o,  + 

o 

o 

-1- 

o 

o 

+  -i 

-1- 

o 

Cryptacanthodidae 

-1- 

+  T 

+ 

+ 

+  [ 

o-  -^ 

+ 

o 

+  -1 

o 

o 

Pholidae 

-1- 

o-i 

o 

o 

o 

+ 

o 

o 

+ 

-1- 

o 

Anarhichadidae 

-1- 

+  1 

-t- 

+ 

+  1 

-1- 

o  -  + 

o 

+  T 

o-  + 

+  -! 

Ptilichthyidae 

-1- 

+ 

o 

o 

+ 

o 

-1- 

+ 

+ 

O  -  +An 

Zaproridae 

-1- 

+  T 

+ 

+ 

+  1 

4- 

-1- 

o 

+ 

o 

+  1 

MATARESE  ET  AL.:  BLENNIOIDEA 


569 


identifying  specific  groups  within  this  infraorder.  Except  for  the 
densely  pigmented  famihes  hsted  above,  pigment  along  the  dor- 
sal body  midline  is  rare  in  preflexion  larvae.  With  development, 
pigment  may  increase  along  the  dorsal  midline  or  on  the  nape. 
Initially,  lateral  pigment  is  either  absent  or  consists  of  a  few 
spots  internally  along  the  notochord.  After  notochord  flexion, 
internal  and  external  pigment  can  increase  ventrolaterally.  or 
above  and  below  the  notochord  (Stichaeidae,  Bathymastendae, 
and  Pholidae).  Typically,  a  series  of  ventral  midline  melano- 
phores  occurs  in  preflexion  larvae.  Although  these  melano- 
phores  may  be  absent  in  some  families  (Chaenopsidae,  some 
Tripterygiidae),  a  number  of  families  have  larvae  with  up  to  50 
melanophores  here  (e.g..  Blenniidae).  The  number,  size,  and 
shape  of  these  melanophores  can  be  very  important  when  iden- 
tifying groups.  These  spots  may  change  shape  with  development 
(becoming  y-shaped  in  Tripterygiidae  and  some  Blenniidae), 
decrease  in  number  (some  Blenniidae  and  Stichaeidae),  or  be- 
come subcutaneous  (Stichaeidae). 

Fins.  — With  the  exception  of  zaprondsand  some  blenniids,  fins 
are  rarely  pigmented  in  preflexion  larvae.  After  notochord  flex- 
ion pigment  develops  on  the  various  fins  of  blenniids,  anarhich- 
adids,  and  ptilichthyids  (Table  146). 

Hypural  margin.  — Pigment  in  the  caudal  area  is  usually  lacking 
in  preflexion  larvae,  and  in  postflexion  larvae  its  presence  is 
limited  to  a  few  families  (Table  146). 

Meristic  characters.— The  number  of  dorsal  fins  varies  from  one 
to  three  and  in  most  families  some  combination  of  spines  and 
rays  is  present,  with  spines  predominating.  Tripterygiids,  clin- 
ids,  and  labnsomids  may  have  up  to  three  dorsal  fins,  the  first 
two  composed  of  spines.  The  total  number  of  dorsal  elements 
is  highly  variable  but  in  some  groups  (stichaeids,  anarhichadids, 
and  ptilichthyids)  well  over  100  elements  are  present.  The  anal 
fin  in  most  groups  may  include  1-2  spines.  Stichaeids  may  have 
up  to  5  anal  spines.  Information  on  the  caudal  fin  is  incomplete. 
In  addition,  from  data  available  in  the  literature,  principal  rays 
and  branched  rays  are  not  consistently  distinguished.  Most  groups 


have  between  9  and  15  (usually  about  12-13)  principal  caudal 
fin  rays  and  about  25-30  total  caudal  fin  rays.  All  possess  a 
pectoral  fin  with  as  few  as  3  (labrisomids  and  clinids)  or  as  many 
as  25  (zaprorids)  fin  rays.  Pelvic  fins  can  be  present  or  absent. 
The  northern  families,  except  some  stichaeids  and  pholids,  lack 
pelvic  fins.  Tropical  families  usually  possess  thoracic  pelvic  fins 
with  1  spine  and  fewer  than  5  rays  (mostly  2-3  soft  rays). 

Vertebral  counts  are  unknown  for  many  blennioids  or  are 
based  on  few  specimens.  The  number  of  vertebrae  is  highly 
variable  within  some  families  (e.g.,  stichaeids,  blenniids,  an- 
archichadids).  In  general,  tropical  families  have  a  lower  verte- 
bral count  than  do  northern  families. 

The  order  of  fin  ray  development  is  highly  variable  in  the 
Blennioidea.  Information  available  on  this  is  also  inadequate, 
since  in  most  studies  reviewed  here  larvae  have  not  been  cleared 
and  stained  to  determine  the  onset  of  ossification.  In  the  tropical 
families  where  notochord  flexion  occurs  as  early  as  3.6  mm,  fin 
ray  development  may  begin  as  early  as  2.5  mm.  Caudal  fin  rays 
develop  first  in  clinid  and  labrisomid  larvae,  followed  by  the 
remaining  fin  rays  soon  after  notochord  flexion  is  complete. 
Typically,  pectoral  fin  rays  develop  first  in  blenniid  larvae  (Blen- 
niini  and  Salariini).  In  Ombranchini  larvae  (Blenniidae),  the 
pectoral  fin  rays  and  caudal  fin  rays  develop  simultaneously. 
Among  the  northern  families,  data  are  insufficient  to  allow  any 
generalizations.  Fin  rays  begin  forming  at  9-1  5  mm  in  stichaeid 
larvae  (usually  caudal  fin  rays  first)  but  may  not  be  complete 
until  larvae  are  30  mm  (Fig.  303B).  Zaprorid  and  cryptacan- 
thodid  larvae  begin  caudal  ray  development  about  the  time 
notochord  flexion  occurs.  Fin  ray  development  in  ptilichthyid 
larvae  begins  with  the  dorsals  and  second  anal  at  40  mm. 

Relationships 

Although  the  scope  of  the  available  egg  and  larval  data  within 
the  Blennioidea  is  limited,  early  life  history  characters  reviewed 
here  do  not  support  the  cohesiveness  of  this  group.  Due  to  a 
lack  of  unifying  characters,  the  infraorder  Blennioidea,  as  pres- 
ently arranged,  probably  does  not  form  a  monophyletic  group. 
Early  life  history  characters  appear  to  be  more  useful  in  clarifying 
relationships  between  families  or  within  families  rather  than 


Table  146.    E.xtended. 


Trunk 

Internal 

Dorsal 

Dorso- 

Medio- 

Ventro- 

Ventral 

noto- 

Hypural 

margin 

lateral 

lateral 

lateral 

margtn 

chord 

margin 

Fin  base 

Diagnostic 

+  Po 

o-  + 

O 

o 

-1- 

o 

o 

P 

Anus,  ventral  midline 

+  -T 

O-  -1- 

+  1 

o 

-t- 

+ 

o 

O 

Gut,  ventral  midline 

o 

o 

o 

o 

o 

o 

o 

O 

Lack  of  pigment 

O,  +Po 

o 

o 

o 

+ 

o 

o 

O 

Swimbladder,  ventral 
midline 

O  -  +  An 

o  -  + 

O-  +1 

o-o,  + 

+  -1 

-1- 

o  -o,  + 

o 

-  PV 

Gut.  ventral  midline 

O-  + 

o 

o 

O-  -H 

-t- 

-1- 

O,  -t- 

o 

Urostyle  or  lateral 
cross-checking 

o-  + 

o 

+  -T 

O-  H- 

+  -1 

+ 

o  -  o,  + 

o 

Gut,  anus,  ventral 
midline 

+  T 

+  T 

-t- 

+  T 

-1- 

- 

o 

o 

Dense  body 

O-O,  -1- 

o 

+ 

o 

-1- 

+ 

0--0,  + 

o 

Gut,  ventral  midline 

+  T 

+  1 

+  1 

+  1 

-1- 

+ 

o 

o 

-  DA 

Dense  body,  fins 

+ 

o 

o 

o 

+  T 

o 

o 

APO 

-C 

Gut,  dorsal  and  ventral 
margin,  caudal  fin 

+  I 

+ 1 

+ 1 

+  1 

+  1 

- 

+  T 

AP 

-TD 

Dense  body 

Fig.  302.  Blennioidea  larvae  (tropical  forms):  (A)  Enneaplerygius  atriceps  (Tripterygiidae),  5.8  mm  (from  Miller  et  al.,  1979  described  as 
Triplerygion  atriceps);  (B-C)  Heterostichus  rostralus  (Clinidae),  6.5  mm,  21.2  mm;  (D)  Parachnus  integripmnis  (Labrisomidae),  7.2  mm;  (E-F) 
Istiblennius  zebra  (Blenniidae),  .1.3  mm,  1 1.0  mm  (from  Miller  et  al.,  1979);  (G)  Enche/vunis  bnmneolus  (Blenniidae),  3.2  mm  (from  Miller  et 
al.,  1979). 


-^r.^^^rrr^r';^^-7r^^^P777:^^^c^^^,^^,^^.^ 


Fig.  303.  Blennioidea  larvae  (northern  forms):  (A)  /?OAi(?M//M.S7orrfaM/(Bathymasteridae),  10.4  mm;  (B)  -4«op/arf/H«pwrpwr«few.y(Stichaeidae), 
12.0  mm;  (C)  Lyconectcs  alcutensis  (Cryplacanlhodidac),  16.0  mm;  (D)  Pholis  sp.  (Pholidae),  23.0  mm;  (E)  Anarhichas  lupus  lupus  (Anarhi- 
chadidae).  24.5  mm  (from  Barsukov,  1959);  (F)  Pnlichlhys  goodei  (Ptilichlhyidae),  24.7  mm  (from  Richardson  and  Dehart,  1975);  (G)  Zaprora 
i(/f «Mi  (Zaproridae),  16.0  mm  (from  Haryu  and  Nishiyama,  1981). 


572 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES- AHLSTROM  SYMPOSIUM 

Table  147.     Summary  of  Selected  MERisTrcs  in  Blennioidea.  Blanks  indicate  data  are  unavailable. 


Fins 

Dorsal 

Anal 

Pectoral 

Pelvic 

Tax  on 

Spines 

Rays 

Spines 

Rays 

Blennioidea 
Dactyloscopidae 


0-VII  +  V-XVIII 


14-36 


22-41 


12-16 


1,3 


Xenocephalidae 

Notograptidae 

Tripterygiidae 

Clinidae 


III-VII  +  X-XXIV 
III  +  XXIV-LXXXIV 


7  -  10  21 

7-15  O-II  14-30  10-19 

1-14  II  14-62  3-18 


5 
1 
1,2-3 

1,2-3 


Chaenopsidae 


XVII-XXXIII 


10-34 


18-38  12-14 


1,3 


Labrisomidae 


III  +  I-IV  +  XX-LII 


7-14  I-Il 


15-28 


3-17 


O-I,  0-3 


Blenniidae 


III-XVII 


9-119  II 


10-119  10-lJ 


1,  2-4 


Bathymasteridae 
Stichaeidae 

II 
XXII-CXXI 

39-49 
0-43 

I-II 
I-V 

27-36 
24-95 

17-21 
8-21 

1,5 
Absent  or  I,  1-5 

Cryptacanthodidae 

Pholidae 

Anarhichadidae 

LX-LXXVII 

LXXIV-CCL 

LXX-CCL 

- 

II 
I-II 

O-I 

45-50 
29-53 
42-233 

11-15 
10-16 
18-23 

Absent 

Absent  or  I,  0-1 

Absent 

Ptilichthyidae 
Zaproridae 

LXXXIII-XC 
LIV-LVII 

115-148 

— 

179-196 
24-30 

13 

20-25 

Absent 
Absent 

Scytalinidae 

41-51 

- 

41-51 

8 

Absent 

between  Blennioidei  infraorders,  e.g.,  the  similarity  between 
labrisomid  and  clinid  larvae  and  the  differences  between  larvae 
in  the  various  blenniid  tribes. 

Many  of  the  families  in  Blennioidea  include  a  large  number 
of  intertidal  forms  and  many  of  the  similarities  (e.g..  demersal 
eggs,  parental  care,  and  advanced  state  of  newly  hatched  larvae) 
may  be  related  to  environmental  conditions  rather  than  to  a 
close  phylogenelic  relationship.  Additional  study  on  the  com- 
plete life  history  of  these  fishes  is  needed  to  identify  unifying 
characters,  if  any  exist.  Studies  at  the  family  level  will  improve 


our  knowledge  of  this  unsatisfactorily  defined  group  and  facil- 
itate outgroup  comparisons. 

(A. CM.)  National  Marine  Fisheries  Service,  Northwest 
AN15  Alaska  Fisheries  Center,  2725  Montlake  Boule- 
vard East,  Seattle,  Washington  981 12;  (W.W.)  Marine 
Ecological  Consultants,  531  Encinitas  Boulevard, 
Suite  110.  Encinitas.  California  92024;  (E.G.S.)  Na- 
tional Marine  Fisheries  Service.  Solithwest  Fisheries 
Center,  PO  Box  271,  La  Jolla,  California  92038. 


MATARESE  ET  AL.:  BLENNIOIDEA 

Table  147.    Extended. 


573 


Pnncipal  caudal 


Caudal 


Total 


Primary  sources 


9-11 


11-14 


25-41 


13-15 

10-13 

20-30 

10-15 

13-35 

25-63 

13,  19-23 

10-14 


10-15 


10-14 


9-16 


20-33 


19-119 


40-50 


39-57 


28-135 


14 

14-16 

35-39 

49-54 

3-8  +  3-9 

14-43 

29-72 

46-113 

13-15,  14 

24-27 

47-51 

72-78 

6-7  +  6-7 

80-107 

7-8  +  19-26 

24-39 

46-214 

72-250 

53-59 

170-181 

227-240 

30-31 

24-26 

61-62 

Bohlke  and  Caldwell,  1961;  Bohlke  and  Chaplin,  1968; 
Dawson.  1974a.  1975,  1976,  1982;  Kanazawa.  1952; 
Miller  and  Briggs.  1962;  Myers  and  Wade.  1946 

Munro,  1967;  Nelson,  1976 

Nelson,  1976 

Bath,  1973;  Leis  and  Rennis,  1983;  Rosenblatt,  1959, 
1960;  Wheeler  and  Dunn.  1975 

Bohlke,  1960b;  George  and  Springer,  1980;  Hoese,  1976; 
C.  Hubbs,  1952,  1953a;  Pennth.  1969;  Scott,  1955. 
1962,  1966,  1967;  Shen.  1 97 1 ;  Springer,  1955,  1970; 
Stevens  and  Springer,  1974 

Bohlke,  1957;  Greenfield,  1972;  Johnson  and  Green- 
field. 1976;  Robins  and  Randall.  1965;  Rosenblatt 
and  Stevens.  1978;  Smith-Vaniz  and  Palacio.  1974; 
Stephens,  1963,  1970;  Stephens  et  al..  1966 

Bohlke  and  Robins.  1974;  Bohlke  and  Sponger.  1961, 
1975;  C.  Hubbs.  1952,  1953b;  Rosenblatt  and  Parr, 
1969;  Rosenblatt  and  Taylor.  1971;  Smith,  1957; 
Sponger,  1954,  1955,  1958,  1959;  Springer  and  Go- 
mon,  1975b 

Bath.  1976,  1978;  Smith-Vaniz,  1975,  1976;  Smith- 
Vaniz  and  Springer.  1971;  Springer,  1967,  1968, 
1971,  1972a,  1972b,  1976;  Springer  and  Gomon, 
1975a;  Springer  and  Smith-Vaniz,  1972;  Springer  and 
Spreitzer,  1978;  Stephens.  1970 

NWAFC,  unpubl. 

Makushok,  1958;  NWAFC.  unpubl.;  Shiogaki,  1980, 
1981 

NWAFC,  unpubl.;  Shiogaki,  1982 

Makushok.  1958;  NWAFC.  unpubl. 

Barsukov,  1959;  Makushok,  1958;  NWAFC.  unpubl. 

Makushok.  1958 

Chapman  and  Townsend,  1938;  NWAFC,  unpubl. 

NWAFC,  unpubl. 


Ammodytoidei:  Development  and  Relationships 
E.  G.  Stevens,  A.  C.  Matarese  and  W.  Watson 


THE  suborder  Ammodytoidei  consists  of  one  family.  Am- 
modytidae,  with  5  genera  and  about  1 8  species.  These  are 
small  (less  than  100-350  mm  SL),  elongate  fish  occurring  in  the 
littoral  and  neritic  waters  of  the  Atlantic,  Indian,  Pacific,  and 
Arctic  Oceans.  Adults  form  schools  but  also  bury  themselves 
in  the  sand.  They  are  commercially  valuable  in  the  North  Sea 
and  off  Japan. 

The  systematic  position  of  Ammodytoidei.  reviewed  by 
Duncker  and  Mohr  ( 1 939).  is  unresolved,  although  the  suborder 
is  considered  a  perciform  derivative  by  Berg  ( 1 940),  Greenwood 
et  al.  ( 1 966),  and  Gosline  ( 1 97 1 ).  A  second  family.  Hypoptych- 
idae.  has  been  included  in  this  suborder  by  these  authors  and 
by  Robins  and  Bohlke  ( 1 970),  but  was  removed  to  the  suborder 
Gasterosteoidei  by  Ida  (1976),  who  considered  it  a  preperco- 
morph  family  on  the  basis  of  jaw  and  caudal  osteology,  egg  size, 
and  reproductive  behavior  (see  Fritzsche,  this  volume). 

Development 

The  ammodytid  genera,  Gymnammodytes,  Hyperoplus,  and 
Ammodytes  (11  species)  are  temperate  and  boreal;  Bleekeria 
and  Embolichthys  (7  species)  are  more  tropical  in  distribution. 
The  confused  nomenclature  of  the  North  Atlantic  species  was 
clarified  by  the  synonomies  in  Reay  (1970)  and  Russell  (1976), 
where  summaries  of  early  life  history  data  were  also  given.  Other 
larval  descriptions  were  given  by  Fage  (1918)  for  Gymnam- 
modytes; by  Altukhov  (1978),  Kobayashi  (1961c),  Norcross  et 


al.  (1961).  Richards  (1965).  Scott  (1972).  and  Senta  (1965)  for 
Ammodytes;  and  by  Macer  (1967)  for  North  Atlantic  species. 
To  date,  eggs  of  6  species  and  larvae  of  9  species  of  these  genera 
have  been  described.  No  early  life  history  data  are  available  for 
the  tropical  genera. 

Eggs 

Eggs  of  the  six  species  that  have  been  described  are  demersal 
and  adhesive,  forming  clumps  on  sandy  substrates  in  shallow 
water.  Eggs,  probably  loosened  by  tidal  currents,  have  been 
collected  in  plankton  nets  (Williams  et  al.  1964;  Senta,  1965). 
Russell  (1976)  summarized  studies  made  on  eggs  resulting  from 
artificial  fertilization.  Incubation  time  ranges  from  2.0  to  12.5 
weeks.  Eggs  are  irregularly  shaped,  but  generally  spherical,  from 
0.67  to  1.23  mm  in  diameter,  with  a  single  yellow  oil  globule. 
0.17  to  0.42  mm.  Embryos  develop  specific  dorsal  and  ventral 
pigment,  pigmented  eyes,  a  moderate  finfold.  and  pectoral  buds 
prior  to  hatching  at  about  3.6  mm. 

Larvae 

Morphology.  — harvdiC  of  Ammodytidae  typically  are  elongate, 
with  rounded  snouts  which  become  pointed  with  age,  and  pre- 
anal  length  slightly  more  than  50%  body  length  (Fig.  304).  Newly 
hatched  larvae  range  from  3.0  to  4.6  mm  body  length.  In  newly 
hatched  and  preflexion  larvae  the  anus  does  not  extend  to  the 
edge  of  the  moderately  wide  finfold  but  opens  to  the  side.  No- 
tochord  flexion  occurs  at  10  to  12  mm  body  length  in  most 


=====:----r^/f0<^^^ 


\\\W' 


■*  »  •■i,...jj-^..^.; 


:i^>;^i^;<g.^^ 


Fig.  304.     Larvae  of:  (upper)  Hyperoplus  lanceolalus.  16  mm,  redrawn  by  H.  Orr  from  Einarsson  (1951);  (middle)  Ammodytes  hexaplerus.  16 
mm;  and  (lower)  Ammodytes  marimis,  16  mm.  redrawn  by  H.  Orr  from  Einarsson  (1951). 


574 


STEVENS  ET  AL.:  AMMODYTOIDEI 


575 


Table  148.    Selected  Pigment  Characters  of  Ammodytidae  Larvae.  0  =  absent,  +  =  present, 

with  development,  po  =  posterior,  an  =  anterior. 


increasing  with  development,  i  =  decreasing 


Species 


Stage 


Body  length    

(mm)  Jaws 


Snout        Brain 


Nape 


Dorsal 
midline 


Fin- 
Caudal     fold 


Gymnammodyles  semisquamaliis  prefiexion 

flexion 
postflexion 

prefiexion 
postflexion 

flexion 
postflexion 


G.  cicerellus 


Hyperoplus  lanceolatus 


H.  immaculatus 


Ammodyles  tobianus 


A.   marinus 
dubius 
americanus 

A.  hexapterus 


preflexion 

flexion 

postflexion 

preflexion 

flexion 

postflexion 

preflexion 

flexion 

postflexion 

preflexion 

flexion 

postflexion 


4.8 

7.0 

11.8-38.0 


5.5-9.0 
13.0 
26.0 

4.0-5.0 

7.5-12.0 

16.0-27.0 

4.5-6.0 

7.5-11.0 

19.0-33.0 

7.0-8.0 
11.0-13.0 
16.0-31.0 


7.0  0 

11.0-25.0  0 

6.0  0 

11.0-25.0  0,  + 


0 
0 

+ 

0 
0 
0 

0 
0 
0 

0 
0 

OT 


0 
0 
OT 

0 

+  ,0 

+ 


0 
0 

+  .0 

0 

+ 


0  + 

0,  +  + 

0  01 

0  + 

+  + 

0  0 

0  + 

0  +,0 


0 

+ 
+ 

0 
0 
01 


+ 

+1 

0 
0 

+ 

+1 

an 
an 
an 


+  an 

+  ,0       an 
+  ,0        0 


an 
an 

+ 

an 
an 
an 


0 
near  tail 

T  + 


po  1/3 

+ 

po  1/4 


0 
near  tail 

I 

0 
po  1/4 

I 

0 
0 

01 


0 

+ 
+ 

+ 
+ 

0 

I 

0 

+ 
+ 

0 
0 

01 

+ 
+ 
+ 

0 

+ 
+ 


Cameron,  1959 
Macer,  1967 


0       Page,  1918 


Einarsson,  1951 
Macer,  1967 

Macer,  1967 


Einarsson,  1955 
Macer,  1967 

Einarsson.  1951 
Macer,  1967 

Kobayashi,  1961c 
NWAFC,  unpubl. 


species,  and  transformation  to  juveniles  occurs  at  about  40  mm. 
The  caudal  fin  is  the  first  to  ossify,  followed  by  the  pectorals, 
then  the  dorsal  and  anal.  The  median  fin  rays  form  in  the  pos- 
terior part  of  the  body,  and  ossification  proceeds  forward.  Dur- 
ing larval  development  the  body  thickens  somewhat,  but  main- 
tains its  elongate  shape.  All  adult  Ammodytidae  have  protrusible 
upper  jaws,  but  Gymnammodytes  semisquamatus  is  the  only 
species  in  which  this  character  is  reported  in  larvae  as  small  as 
9  mm  (Cameron,  1959).  Postflexion  larvae  oi  Hyperoplus  de- 
velop vomerine  teeth  which  persist  in  the  adult,  while  Gym- 
nammodytes postflexion  larvae  develop  both  vomerine  and  pre- 
maxillary  teeth  which  disappear  at  about  transformation.  During 
the  larval  period  Gymnammodytes.  Hyperoplus.  and  .Ammo- 
dyles are  pelagic.  Juveniles  and  adults  are  both  pelagic  and 
benthic. 

Pigment.  — Pigment  can  be  a  useful  diagnostic  feature  among 
the  larvae  of  Ammodytidae,  especially  the  location  and  devel- 
opment of  melanophores  on  the  ventral  gut  margin,  the  dorsal 
body  margin,  and  the  caudal  area,  i.e.,  the  tip  of  the  notochord 
and  the  edge  of  hypural  elements.  These  pigment  characters  are 
summarized  in  Table  148.  All  species  have  a  row  of  melano- 
phores dorsally  on  the  gut,  beginning  at  or  just  posterior  to  the 
cleithrum,  and  a  postanal  row  on  the  ventral  body  margin  from 
the  anus  to  the  tail.  The  dorsal  gut  pigment  becomes  obscured 
with  growth.  Specific  variations  in  pigment  patterns  can  be  seen 
in  the  16  mm  specimens  illustrated  in  Fig.  304.  At  this  length, 
dorsal  midline  pigment  forms  a  complete  row  in  H.  lanceolatus, 
but  occurs  only  on  the  posterior  quarter  in  .-i.  hexapterus  and 
.4.  marinus:  and  ventral  gut  pigment  extends  the  length  of  the 
gut  in  H.  lanceolatus  and  .4.  marinus  but  is  found  only  on  the 
anterior  ventral  gut  of  .4.  hexapterus.  Pigment  patterns  of  .1. 
marinus,  .4.  dubius  and  .4.  americanus  are  nearly  identical  (Ma- 
cer, 1976)  although  Richards  (1982)  has  noted  diflTerences  in  the 
ranges  of  melanophore  numbers,  especially  on  the  anterior  ven- 
tral gut  (stomach)  and  dorsal  midline  (supradorsal).  Pigment 


appears  variously  on  the  head,  increasing  with  age  in  all  species 
reported.  The  only  reported  decrease  in  pigmentation  is  on  the 
dorsal  and  ventral  margins  of  G.  cicerellus  (Page,  1918).  G. 
semisquamatus  has  pigment  on  the  ventral  finfold  margin,  the 
only  ammodytid  species  for  which  finfold  pigment  has  been 
noted  (Cameron,  1959). 

Aferistics.  — Fin  ray  and  vertebral  counts  for  the  family  Am- 
modytidae are:  Vert  54-78;  D  40-69;  A  14-36;  Pec.  10-16;  Pel. 
0-1,5;  and  C  15-17  prin.,  1 3  branched.  In  all  genera  the  number 
of  precaudal  vertebrae  exceeds  the  number  of  caudal.  Robins 
and  Bohike  (1970)  report  9  +  8  principal  caudal  rays  for  Em- 
bolichthys  sarissa.  but  all  other  ammodytid  species,  including 
E.  mitsukurit,  have  8  +  7.  Embolichthys  is  the  only  genus  with 
pelvic  fins,  which  are  thoracic.  The  caudal  fin  is  the  first  to  form, 
followed  by  the  pectorals,  dorsal,  and  anal.  Posterior  rays  of  the 
median  fins  form  first  and  development  proceeds  forward.  Fin 
formation  is  completed  by  30  to  40  mm  body  length. 

Relationships 

Although  early  life  history  data  of  the  suborder  Ammody- 
toidei  do  little  to  clarify  its  phylogenetic  position,  larval  pigment 
patterns  and  myomere- vertebrae  counts  are  useful  in  separating 
sympatric  species  (Macer,  1967).  General  characters,  such  as 
the  well  developed  state  of  newly  hatched  larvae  and  the  se- 
quences of  fin  development,  are  shared  with  other  perciform 
derivatives  and  relatives,  but  essentially  the  problem  of  the 
systematic  position  of  ammodytids  is  not  yet  resolved. 

(A. CM.)  National  Marine  Fisheries  Service,  Northwest 
AND  Alaska  Fisheries  Center,  2725  Montlake  Boule- 
vard East,  Seattle,  Washington  98112;  (E.G.S.)  Na- 
tional Marine  Fisheries  Service,  Southwest  Fisheries 
Center,  PO  Box  27 1 .  La  Jolla,  California  92038;  (W.W.) 
Marine  Ecological  Consultants,  531  Encinitas  Boule- 
vard, Slute  1 10,  Encinitas.  California  92024. 


Icosteoidei:  Development  and  Relationships 

A.  C.  Matarese,  E.  G.  Stevens  and  W.  Watson 


THE  suborder,  Icosteoidei,  consists  of  one  family,  Icosteidae, 
with  a  monotypic  species  Icosteus  aenigmaticus  (Nelson, 
1976).  Adults  inhabit  the  epi-  and  bathypelagic  areas  of  the 
North  Pacific  Ocean  from  southern  California  to  Japan.  Adults 
may  exceed  2  m  in  length  and  have  dark  flaccid  bodies;  a  char- 
acteristic implied  in  the  common  name  of  the  species,  raglish 
(Hart,  1973). 

The  systematic  position  of  this  group  and  its  designation  as 
an  order  or  suborder  is  not  well  established.  Greenwood  et  al. 
( 1 966)  consider  it  a  suborder  of  Perciformes  while  Gosline  (1971) 
elevates  it  to  an  order,  Icosteiformes,  a  probable  perciform  de- 
rivative. 

Development 

The  only  early  life  history  data  previously  published  is  a  brief 
description  of  the  egg  (Allen,  1968).  Icosteus  aenigmaticus  eggs 
are  commonly  collected  in  ichthyoplankton  surveys  off  the  Pa- 
cific coast  of  North  America  [California  Cooperative  Oceanic 
Fisheries  Investigations  (CalCOFI)  and  Northwest  and  Alaska 
Fisheries  Center  (NWAFC)],  but  larvae  (mostly  preflexion)  are 
infrequently  found  and  a  complete  size  series  from  hatching  to 
transformation  is  not  presently  available.  Larvae  may  move 
offshore  or  into  deeper  waters.  The  first  published  description 
and  illustration  of  the  larvae  from  pre-  to  postflexion  stages  are 
provided  here,  based  on  National  Marine  Fisheries  Service 
(NMFS)  collections.  Although  /.  aenigmaticus ]uven\\e^  undergo 
a  marked  transformation  to  the  adult  stage,  little  information 
is  available  concerning  this  change  (Hart,  1973). 

Eggs 

The  pelagic  egg  of  /.  aenigmaticus  ranges  in  diameter  from 
2.8  to  3. 1  mm  (Fig.  305).  A  large,  sometimes  irregular,  oil  glob- 
ule with  a  diameter  of  0.42  to  0.60  mm  is  present.  The  oil  globule 
usually  decreases  in  size  with  development.  The  chorion  is 
smooth,  sometimes  amber  or  rose  colored.  Early  stage  egg  yolks 
are  frequently  opaque,  although  later  stages  have  a  clear,  un- 
segmented  yolk.  During  the  middle  stage  of  development,  em- 
bryos have  pigment  along  the  dorsal  body  as  well  as  on  the  yolk 
and  oil  globule.  Late  stage  embryos  have  functional  mouths, 
pectoral  buds,  and  very  wide  finfolds.  Scattered  pigment  occurs 
on  the  eyes,  snout,  jaws,  and  dorsal  head.  The  dorsal  surface  of 
the  gut  is  pigmented.  Along  the  dorsal  and  anal  finfolds,  three 
or  four  clusters  of  melanophores  appear  at  each  distal  edge. 
Melanophores  also  appear  above  and  below  the  tail  in  the  caudal 
finfold.  An  irregular  double  row  of  melanophores  extends  the 
length  of  the  dorsal  body  margin.  A  few  mediolateral  spots 
appear  anteriorly.  Occasionally,  pigment  occurs  along  the  ven- 
tral body  margin. 

Larvae 

Morphology.— Nevj\y  hatched  larvae  of/,  aenigmaticus  are  6.5 
mm  NL;  yolk  material  may  persist  until  larvae  are  10  mm. 
Flexion  begins  at  about  1 1  mm  and  is  complete  at  about  1 7 
mm  SL.  The  size  at  transformation  is  not  known,  but  fin  de- 


velopment is  almost  complete  by  28  mm.  The  body,  surrounded 
by  a  wide  finfold,  is  very  soft.  Preflexion  larvae  have  small  heads 
with  rounded  snouts  and  long  tapering  bodies  (Fig.  306).  Dor- 
sal and  ventral  finfolds  are  wider  than  the  body.  During  flexion 
the  body  thickens  and  becomes  more  robust,  especially  ante- 
riorly. Postflexion  larvae  have  a  robust  head  and  gut  and  a 
tapering  trunk  (Fig.  306).  Preanal  length  is  less  than  50%  body 
length.  A  series  of  preopercular  spines  appears  during  late  flex- 


Pigment.  —  New\y  hatched  larvae  of  /.  aenigmaticus  display  es- 
sentially the  same  eye,  head,  gut,  body,  and  finfold  pigment  as 
the  embryos.  With  increasing  size  the  head  and  gut  usually 
become  increasingly  covered  with  discrete  spots.  Dorsal  body 
margin  pigment  is  present  throughout  larval  development,  while 
the  amount  of  lateral  and  ventral  body  margin  pigment  varies 
and  is  relatively  sparse.  The  characteristic  embryonic  caudal 
pigment  persists  in  the  developing  larvae,  becoming  less  prom- 
inent but  remaining  as  scattered  melanophores  on  the  hypural 
margin  and  fin  ray  bases.  In  general,  postflexion  larvae  are  less 
pigmented  except  on  the  head.  Pelvic  and  pectoral  fin  bases  and 
pelvic  rays  acquire  melanophores  during  postflexion. 

Meristics.  —  Icosteus  aenigmaticus  larvae  have  the  following  ver- 
tebral and  fin  ray  counts:  Vert.  66-68;  D  55;  A  39;  Pec.  21;  Pel. 
1,4;  and  C  9  -(-  8  =  17  (NWAFC  files).  These  counts  conform 


Fig.  305.     Egg  of  Icosteus  aenigmaticus:  2.8  mm,  drawn  by  H.  Orr. 


576 


MATARESE  ET  AL.:  ICOSTEOIDEI 


577 


Fig.  306.     Larvae  of  Icosteus  aenigrnaticus  from  top  to  bottom:  9.5  mm  SL;  10.2  mm;  and  28.5  mm  SL,  drawn  by  H.  Orr. 


to  those  for  the  adults  except  adults  lack  a  pelvic  fin  (Abe,  1954; 
Miller  and  Lea,  1972;  Hart,  1973).  The  caudal  fin  contains  the 
perciform  number  of  principal  rays,  1 7,  with  6-9  procurrent 
rays  on  each  side.  Pectoral  fin  blades  are  present  at  hatching 
and  rays  form  during  flexion.  Pelvic  fin  rays  begin  development 
during  flexion  and  are  complete  in  postflexion  larvae.  At  what 
size  the  pelvic  fins  disappear  is  not  known.  The  last  fin  rays  to 
form  are  the  dorsal  and  anal,  with  their  anlagen  appearing  in 
the  middle  of  the  posterior  half  of  the  finfolds  at  about  mid- 
flexion.  Formation  of  these  fins  proceeds  forward  and  toward 
the  body  margin  (Fig.  306).  The  largest  larva  available,  28.5 
mm,  has  the  complete  fin  ray  complement. 

Relationships 

The  foregoing  brief  description  of  the  eggs  and  larvae  of  /. 
aenigrnaticus  provides  some  additional  information  toward  the 
understanding  of  the  life  history  of  this  unique  but  poorly  under- 
stood fish.  Characters  discussed  here  (e.g.,  sequence  of  fin  for- 


mation and  meristics)  help  support  its  position  among  perciform 
relatives.  Sequence  of  fin  formation  and  reduced  number  of 
pelvic  fin  rays  are  blennioid-like  characters,  and  1 7  principal 
caudal  fin  rays  are  the  typical  percoid  number.  Eggs,  larvae,  and 
early  juveniles  superficially  resemble  stromateoid  fishes  but  ad- 
ditional data  are  needed  before  a  precise  relationship  can  be 
determined.  To  understand  this  family  more  fully,  we  need 
information  regarding  the  critical  juvenile  phase  as  well  as  a 
complete  osteological  examination  from  preflexion  larvae  to 
adults. 

(A. CM.)  National  Marine  Fisheries  Service,  Northwe.st 
AND  Alaska  Fisheries  Center,  2725  Montlake  Boule- 
vard East,  Seattle,  Washington  98112;  (E.G.S.)  Na- 
tional Marine  Fisheries  Service,  Southwest  Fisheries 
Center,  PO  Box  27 1 ,  La  Jolla,  California  92038;  (W.W.) 
Marine  Ecological  Consultants,  53 1  Encinitas  Boule- 
vard, Suite  1 10,  Encinitas,  California  92024. 


Zoarcidae:  Development  and  Relationships 
M.  E.  Anderson 


THE  eelpouts,  Zoarcidae,  comprise  a  monophyletic  group  of 
about  200  valid  species  of  marine  fishes  in  44  genera  (Table 
149;  Anderson,  1984).  About  20  additional  undescribed  species 
are  known  to  me  from  collections  around  the  world.  Most  zoar- 
cids  live  on  the  bottom  in  deep  water  in  boreal  seas,  but  1 1  are 
known  from  intertidal  areas,  especially  in  temperate  South 
America.  Twenty-two  species  are  known  from  tropical-sub- 
tropical areas  and  all  of  them  live  in  deep  water  (640-5,320  m). 


Fourteen  species  are  known  from  both  shallow  and  deep  waters 
of  Antarctica  and  subantarctic  regions.  Two  deep-living  genera, 
Lycodapus  and  Melanostignia,  are  coastal  or  thalassobathyal, 
deep-pelagic  forms  that  seem  to  occur  in  greatest  numbers  where 
their  zooplankton  prey  concentrate  (Belman  and  Anderson,  1 979; 
Anderson,  1981).  Thus  the  family  is  stenothermic  and  adapted 
to  very  low  temperatures  (mostly  below  about  8°  C). 


Table  149.     Distribution,  Ecology  and  Selected  Meristics  of  Zoarcidae. 


No.  ol 
species 

Distribution 

Genus 

Ecology 

D 

Aiakas 

1 

SW  Atlantic 

Benthic;  slope 

88-89 

Andriashevia 

1 

NW  Pacific 

Benthic;  slope 

144 

Austrolycichthys 

4 

E  trop.  Pacific;  W  trop. 
Atlantic;  Antarctica 

Benthic;  slope 

O-ll,  87-104 

Bilahria 

1 

NW  Pacific 

Benthic;  shelf 

110 

Bothrocara 

8-9 

NW  Pacific  to  Peru 

Benthic;  slope 

108-125 

Bothrocarina 

2 

NW  Pacific 

Benthic;  shelf-slope 

1,  106 

Cwssoslomus 

2 

SW  Atlantic 

Benthic;  shelf 

96-108 

Dadyanos 

I 

SW  Atlantic 

Benthic;  shelf 

104-116 

Davidijordania 

4 

NW  Pacific 

Benthic;  shelf 

85-118 

Derepodichthys 

1 

NE  Pacific 

Benthic;  slope 

110-116 

Exechodontes 

1 

W  trop.  Atlantic 

Benthic;  slope 

80-86 

Gymnelopsis 

4 

NW  Pacific 

Benthic;  shelf-slope 

I,  79-112 

Gymnelus 

5 

N  Pacific,  Arctic 

Benthic;  shelf 

I,  76-101 

Hadropareia 

1 

NW  Pacific 

Benthic;  shelf 

106-113 

Hadwpogonichthys 

1 

NW  Pacific 

Benthic;  slope 

126-128 

Iluocoetes 

2 

SE  Pacific-SW  Atlantic 

Benthic;  shelf 

84-98 

Krusensterniella 

4 

NW  Pacific 

Benthic;  shelf 

XLV-LVII,  II-XXVI 

37-64 

Lycenchelys 

40-41 

Worldwide,  except  Indo-Pac. 

Benthic;  slope-abyss 

0-1,94-132 

Lycodapus 

13 

NW  Pacific  to  subantarctic 

Deep  pelagic 

70-98 

Lycodes 

46-51 

Boreal  seas;  South  Africa 

Benthic;  shelf-abyss 

85-120 

Lycodichthys 

2 

Antarctica 

Benthic;  slope 

l-II.  83-89 

Lycodonus 

4 

N  and  S  Atlantic 

Benthic;  slope 

98-112 

Lycogrammoides 

NW  Pacific 

Pelagic  ? 

62 

Lyconema 

NE  Pacific 

Benthic;  shelf-slope 

100-107 

Lycozoarces 

NW  Pacific 

Benthic;  shelf 

II,  62-67 

Macrozoarces 

NW  Atlantic 

Benthic;  shelf 

92-103,  XVI-XXIV, 

16-30 

Maynea 

SW  Atlantic-SE  Pacific 

Benthic;  shelf 

119-127 

Melanosligma 

7 

Worldwide 

Deep  pelagic 

76-95 

Nalbantichthys 

N  Pacific 

Benthic 

slope 

143-152 

Notolycodes 

SW  Atlantic 

Benthic 

slope 

85-88 

Oidiphorus 

SW  Atlantic 

Benthic 

slope 

56-61 

Ophthalmolycus 

3 

Chile  to  Antarctica 

Benthic 

slope 

O-I,  87-103 

Pachycara 

9 

Worldwide 

Benthic 

slope-abyss 

0-1,95-113 

Phucocoetes 

SW  Atlantic-SE  Pacific 

Benthic 

shelf 

101-107 

L 

Piedrabuenia 

SW  Atlantic 

Benthic 

slope 

108-113 

Pogonolycus 

SW  Atlantic 

Benthic 

shelf 

86-88 

1 

Puzanovia 

NW  Pacific 

Benthic 

slope 

135-147 

Taranetzella 

N  Pacific 

Benthic 

slope-abyss 

84-89 

Zoarces 

3 

NE  Atlantic;  NW  Pacific 

Benthic 

shelf 

72-94,  0-XIX,  14- 

-27 

Genus  A. 

Coast  of  California 

Benthic 

shelf-slope 

97-107 

Genus  B. 

Scotia  Sea 

Benthic 

slope 

_ 

Genus  C. 

Bering  Sea 

Benthic 

slope 

— 

1 

Genus  D. 

SW  Atlantic 

Benthic 

slope 

77-83 

Genus  E. 

2 

SW  Atlantic 

Benthic 

slope 

92-96 

578 


ANDERSON:  ZOARCIDAE 


579 


Development 

As  far  as  known,  almost  all  eelpouts  are  oviparous,  laying 
relatively  few.  large  eggs.  The  exception  is  the  genus  Zoarces. 
which  is  viviparous.  There  are  three  species  of  Zoarces,  the 
common  European  eelpout,  Z.  viviparus  (Linnaeus),  and  two 
little  known,  northwestern  Pacific  species,  Z.  gillii  Jordan  and 
Starks  and  Z.  ekmgatus  Kner.  Viviparity  in  the  European  eel- 
pout  has  been  known  since  the  Middle  Ages  (Schonevelde.  1624). 
but  of  the  two  Pacific  species,  females  with  embryos  are  known 
only  in  Z.  gillii  (Anderson.  1984). 

Among  benthic,  oviparous  species,  nest  building  with  parental 
guardianship  is  probably  common.  Nesting  has  been  directly 
observed  in  Macrozoarces  americanns  (Olsen  and  Merriman, 
1946),  Gymnelns  viridis  (Emery,  1973),  Lycodes  pacificus  (Lev- 
ings,  1969)  and  Phucocoetes  latitans  and  Ihiocoetes  effusus 
(Gosztonyi,  1977).  Probably  most,  if  not  all,  the  other  South 


American  intertidal  zoarcids  discussed  by  Gosztonyi  also  build 
and  guard  nest  sites.  Pelagic  spawning  occurs  in  Lycodapus  and 
Melanostigma.  Markle and  Wenner(  1 979) found  Melanostigma 
ailanticum  may  utilize  the  sea  bottom  as  a  concentration  in- 
terface for  group  spawning.  Bottom  trawl-caught  ripe  individ- 
uals had  parasite  loads  more  typical  of  deep-demersal  fishes  in 
the  western  North  Atlantic.  However,  Anderson  (1981)  reported 
Lycodapus  mandihularis  to  have  a  parasite  fauna  similar  to 
other  midwater  fishes  in  Monterey  Bay,  California.  Early  ju- 
veniles were  caught  in  midwater  at  all  depths  inhabited  by  adults. 
This  suggests  L.  mandihularis  does  not  shoal  near  the  bottom 
for  spawning. 

Eggs 

Spawned  zoarcid  eggs  have  been  described  from  field  obser- 
vations for  only  seven  species  (Table  1 50).  Egg  descriptions  are 


Table  149.    Extended. 


Fin  rays 

Venebrae 

A 

P 

c 

Precaudal 

Caudal 

Sources 

67-69 

18-19 

10 

26 

65-66 

This  report;  Gosztonyi  (1977) 

123 

Absent 

— 

22 

125 

Fedorov  and  Neyelov  (1978) 

70-89 

15-19 

9-10 

20-25 

72-87 

This  report 

93-94 

15-16 

10 

23 

95-96 

This  report;  Lindberg  and  Krasyukova  (1975) 

92-109 

13-17 

0-13 

18-24 

93-108 

This  report 

95 

10-11 

11 

19 

94 

This  report 

68-78 

16-17 

10 

28-32 

67-76 

This  report;  Gosztonyi  (1977) 

89-95 

16-17 

10 

21-24 

84-93 

This  report;  Gosztonyi  (1977) 

68-90 

12-17 

10 

20-23 

77-97 

This  report;  Lindberg  and  Krasyukova  (1975) 

94-101 

10-11 

8-9 

22-26 

92-98 

Anderson  and  Hubbs  (1981) 

73-79 

13-15 

10 

19-21 

72-78 

This  report;  DeWitt  (1977) 

73-97 

9-12 

5-8 

16-23 

73-95 

Anderson  (1982) 

69-85 

9-14 

9-12 

17-26 

65-84 

Anderson  (1982) 

86-92 

13-15 

7-8 

24-28 

83-89 

This  report 

112-114 

19 

11 

23-24 

109-110 

Fedorov  ( 1 982) 

65-82 

15-19 

7-9 

19-24 

62-79 

This  report;  Gosztonyi  (1977) 

71-103 

11-12 

5-7 

19-25 

80-97 

This  report;  Lindberg  and  Krasyukova  (1975) 

80-112 

13-21 

9-13 

20-30 

77-118 

This  report;  Andriashev  (1955b) 

58-86 

5-9 

8-12 

13-19 

59-85 

Peden  and  Anderson  (1978,  1981) 

67-92 

14-24 

10-12 

19-26 

65-104 

This  report 

66-75 

15-17 

11 

23-24 

68-70 

This  report;  DeWitt  (1962a) 

83-93 

14-17 

7-9 

21-25 

85-105 

This  repon 

52 

9 

8^ 

14 

53 

This  report 

90-96 

15-17 

12 

20-21 

86-93 

This  report;  Gotshall  (1971) 

49-54 

13-15 

13-15 

15-17 

50-55 

This  report;  Toyoshima  (1981) 

103-125 

17-20 

9-10 

25-28 

105-118 

This  report 

95-103 

14-16 

7 

29-30 

89-98 

This  report;  Gosztonyi  (1977) 

62-80 

6-9 

8-10 

18-23 

62-81 

This  report;  Parin  (1977) 

121-127 

6 

7-10 

25 

119-125 

Schultz(1967) 

69-72 

18-21 

11 

23-26 

66-69 

This  report;  Gosztonyi  (1977) 

45-54 

16-19 

7-9 

15-17 

43-50 

This  report;  Gosztonyi  (1977) 

69-87 

14-18 

10 

22-23 

72-88 

This  report 

77-97 

14-19 

10-12 

25-32 

74-91 

This  report 

78-85 

14-16 

10 

24-27 

75-83 

This  report;  Gosztonyi  (1977) 

98-104 

17-18 

8 

24-25 

95-101 

This  report;  Gosztonyi  (1977) 

72-74 

17 

9 

20 

71-73 

This  report;  Gosztonyi  (1977) 

115-128 

9-12 

9-12 

22-24 

110-125 

Fedorov  (1975);  Amaoka  et  al.  (1977) 

71-76 

15 

8 

20 

69-74 

This  report;  Andriashev  (1952) 

64-90 

16-21 

9-11 

21-26 

80-106 

This  report;  Schmidt  (1917) 

83-93 

13-14 

10 

27-28 

79-84 

This  report;  Cailliet  and  Lea  (1977) 

— 

— 

— 

— 

— 

Tomoet  al.  (1977) 

— 

— 

— 

— 

— 

Bond  and  Stein  (in  prep.) 

64-69 

17-19 

9-10 

19-21 

66-73 

This  report 

68-73 

9-14 

7 

24-27 

67-75 

This  report 

580 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  150.    Data  on  the  Eggs  and  Larvae  of  Zoarcidae  Known  to  Date. 


Eggs 

rv 

Size 
range,  mm 

Oil 
globule 

Incubation 
(months) 

Hatch  size, 
mm 

Species 

Descnption 

Illustration 

Sources 

Gymnelus  viridis 

3.2-4.6* 

_ 

_ 

-20-25 

X 

X 

Rass(1949) 

Macrozoarces  americanus 

-6.0-7.0 

2.5-3.5 

28-31 

X 

X 

White  (1939);  Olsen 
and  Merriman 
(1946) 

Zoarces  viviparus 

2.8-3.2* 

— 

4 

32-40 

X 

X 

Soin(1968); 

Altukhov(1979) 

"Maynea"  californica 

4.0-5.3* 

— 

— 

— 

— 

— 

Kliever(l976) 

Melanostigma  atlanticum 

-3.0-3.8* 

— 

— 

— 

— 

— 

Markle  and 
Wenner(1979) 

Lycodes  pacificus 

5.0* 

— 

— 

— 

— 

— 

Levings(1969) 

Lycodes  palearis 

6.0-7.0 

Multiple 

— 

— 

— 

— 

Slipp  and  DeLacy 

(1952) 

Lycodes  pallidus 

— 

— 

— 

-10-12 

— 

— 

Altukhov(1979) 

Derepodichthys  alepidotus 

-2.1-2.4* 

— 

— 

— 

— 

— 

Anderson  and  Hubbs 
(1981) 

Bolhrocara  hoUandi 

9.2 

— 

— 

35-36 

X 

X 

Okiyama  (1982a) 

Bolhrocara  sp. 

7.0 

1  (1.6  mm) 

— 

— 

X  (embryos) 

X  (embryos) 

Kendall  et  al.  (1983) 

Lycodapus  mandihularis 

1.7-1.9* 

1 

— 

-15-17 

— 

— 

Anderson  (1981) 

Auslrolycus  laticmctus 

7.5-8.4* 

— 

— 

-17 

— 

— 

Gosztonyi  (1977) 

Dadyanos  insigms 

5.0 

— 

— 

— 

— 

— 

Gosztonyi  (1977) 

Phucocoetes  latilans 

4.5 

— 

— 

20 

— 

— 

Gosztonyi  (1977) 

Iluocoetes  effusus 

5.0-5.5 

— 

2 

— 

— 

- 

Gosztonyi  (1977) 

'  Maximum  ovanan  diameters. 


generally  cursory  (except  that  of  Kendall  et  al.,  1 983).  In  general, 
zoarcid  eggs  are  large  (about  4-9  mm,  except  in  some  diminutive 
species),  spherical  and  usually  with  a  single  oil  globule  that  may 
have  coalesced  from  a  few  smaller  globules.  Spawned  eggs  are 
orange-yellow  or  purple  with  a  somewhat  darker  orange  or  yel- 
low oil  globule  (Anderson,  1981;  Kendall  et  al.,  1983)  and  have 
a  narrow  perivitelhne  space.  Benthic  egg  masses  are  held  to- 
gether by  a  sticky,  gelatinous  mass  that  is  not  especially  thick- 
ened. Incubation  times  are  known  for  only  three  species.  Eggs 
"hatch"  in  the  ovary  of  Zoarces  viviparus  afler  two  months  and 
embryos  develop  for  another  two  months  therein  (Fig.  307). 
Embryos  develop  a  dense  vitelline  vascular  network  that  aids 
in  yolk  resorption,  respiration  and  assimilation  of  nutrients  from 
the  mother's  ovarian  fluid  (Soin,  1968).  Olsen  and  Merriman 
( 1 946)  found  that  eggs  of  southern  populations  of  Macrozoarces 
americanus  had  an  average  incubation  time  of  2.5  months, 
whereas  eggs  of  northern  populations  took  about  3.5  months  to 
hatch.  Gosztonyi  (1977)  observed  the  eggs  of  Iluocoeles  effusus 
(as  /.  elongatus)  from  the  Patagonian  intertidal  to  require  two 
months  to  hatch  during  the  austral  autumn. 

Larvae 

As  with  observations  on  eggs,  zoarcid  "larvae"  are  not  well 
known,  if  this  stage  is  developed  at  all.  Although  a  few  early 
stages  have  been  collected  during  ichthyoplankton  surveys  (Rass, 
1949;  Mattson  and  Wing,  1978;  Altukhov,  1979),  they  are  no- 
tably absent  in  collections  of  other  surveys  where  adults  are 
abundant,  such  as  the  Bering  Sea  (Musienko,  1963;  Waldron 
andVinter,  1 978)  and  offOregon  (Richardson  and  Pearcy,  1977). 
This  is  probably  due  to  their  short  planktonic  time.  Early  life 
history  stages  of  only  five  zoarcid  species  have  been  illustrated 
(Kendall  et  al.,  1983)  and  all  these  are  reproduced  here  (Figs. 
307,  308). 

Females  of  Lycodapus  mandihularis,  Gymnelus  viridis  and 


"Maynea'''  californica^  are  known  to  spawn  larger  eggs  at  in- 
creasingly larger  adult  sizes,  thus  zoarcid  hatching  sizes  vary. 
In  large  eelpouts,  like  Macrozoarces  and  Zoarces.  young  hatch 
at  about  30-40  mm,  but  diminutive  species,  like  Melagoslignia 
and  Derepodichthys  are  probably  only  about  1 0  mm  at  hatching. 
At  hatching,  the  yolk  sac  is  rapidly  taken  into  the  gut.  White 
(1939)  reported  "the  complete  external  disappearance  of  the 
yolk"  to  occur  in  about  20  seconds  in  Macrozoarces  americanus 
that  were  stimulated  to  hatch  in  a  pan  of  cold  sea  water.  I  have 
observed  a  similarly  rapid  internalization  of  the  yolk  in  larvae 
of  the  liparidid  Carcproclus  sp.  (Anderson  and  Cailliet,  1974). 
Perhaps  rapid  yolk  uptake  is  typical  of  fishes  with  a  protracted 
developmental  period. 

Newly  hatched  zoarcids  strongly  resemble  adults.  The  major 
difl^erences  are  the  larger  eyes  and  more  rounded  snout  in  the 
young  (Fig.  308).  At  the  free-swimming,  yolk-ingestion  stage, 
all  fin  rays  have  formed.  The  stage  and  direction  of  fin  formation 
in  embryos  is  unknown  in  Zoarcidae.  Most  of  the  cephalic  lat- 
eralis pores  were  formed  in  larvae  of  Bolhrocara  hollandi  (as 
Allolepis  hollandi)  noted  by  Okiyama  (1982a).  Post-hatching 
Macrozoarces  that  I  examined  for  this  study  had  not  developed 
all  their  lateralis  pores,  a  case  similar  to  that  of  Gymnelus  spp. 
(Anderson,  1982).  These  planktonic  young  Macrozoarces  had 
absorbed  their  yolk  and  measured  33.8-36.0  mm  SL.  The  young 
fish  were  generally  well  ossified,  except  central  regions  of  the 
neurocranium  and  suspensorium.  Jaw  and  pharyngeal  teeth  were 
developed  and  a  few  had  eaten  copepods.  In  the  smallest  spec- 
imen, the  pectoral  actinosts,  scapula  and  coracoid  were  a  fused 
mass  of  cartilage,  but  these  were  separated  and  ossified  in  just 
slightly  larger  specimens.  Vertebrae  were  square  in  shape  (rect- 


'  This  species  properly  belongs  in  an  unnamed,  monotypic  genus 

(Anderson.  1984). 


ANDERSON:  ZOARCIDAE 

A 


581 


Fig.  307.  Zoarces  viviparus.  (A)  egg  and  newly  hatched  embryo;  (B) 
developing  embryo  from  mother's  ovary;  and  (C)  newly  emerged  young; 
all  from  Soin  (1968). 


angular  in  adults)  and  all  neural  arches  were  fused,  as  in  adults. 
In  the  caudal  skeleton,  all  fin  rays  and  plerygiophores  were 
present,  as  in  adults,  but  the  neural  arches  of  the  first  ural  and 
first  preural  centra  were  poorly  developed,  with  some  sections 
free  of  the  urostyle.  Typical  of  many  zoarcids,  the  caudal  of 
Macrozoarces  has  two  epural.  four  upper  hypural  and  3-4  lower 
hypural  fin  rays. 

There  are  no  specialized  larval  pigment  patterns.  The  larvae 
of  Gymnelus  viridis  and  Bolhrocara  hollandi  appear  to  be  mono- 
tone, as  are  most  adults  (Rass,  1949;  Okiyama,  1982a).  The 
larva  of  Macrozoarces  illustrated  by  White  (1939)  and  those 
examined  by  me  bore  the  typical  criss-cross  pigment  pattern  of 
older  stages. 

Meristic  characters  ofA/arroroarcfi  early  juveniles  examined 
fit  within  the  range  reported  for  adults  (Table  I  50).  However, 
Soin  (1968)  and  Kendall  et  al.  (1983)  showed  that  developing 
embryos  oi  Zoarces  viviparus  and  Bolhrocara  sp.  had  myomere 
counts  well  below  that  of  adult  populations.  Although  large 
sample  sizes  of  most  zoarcid  genera  are  lacking  for  satisfactory 
statistical  analysis  of  meristic  characters,  the  important  thing  to 
note  is  that  myomere  addition  seems  to  be  a  slow  process  in 
zoarcids  and  that  the  full  adult  complement  may  not  be  reached 
until  embryos  are  very  close  to  hatching.  Alternatively,  zoarcid 
embryos  and  larvae  may  have  differentiated  myomeres  with  the 
adult  counts,  but  their  small  size  and  tight  packing,  particularly 
near  the  tail  tip,  may  make  it  difficult  to  observe  them  with  a 
conventional  light  microscope. 


Fig.  308.  Early  stages  of  Zoarcidae.  (A,  B)  Bolhrocara  sp.,  after 
Kendall  et  al.  (1983);  (C)  Bolhrocara  hollandi.  after  Okiyama  (1982a); 
(D)  Gymnelus  viridis.  after  Rass  (1949);  and  (E)  Macrozoarces  ameri- 
canus.  after  White  (1939). 


Relationships 

The  relationships  of  the  zoarcids  to  other  living  fishes  has 
been  confused  in  the  literature.  Greenwood  et  al.  (1966)  and 
Rosen  and  Patterson  (1969)  allied  the  zoarcids  to  the  gadiform- 
ophidiiform  lineages.  Two  of  the  four  characters  they  used  to 
suggest  this  relationship,  the  presence  of  a  basisphenoid  bone 
and  free  second  ural  centrum,  both  illustrated  by  Yarberry  ( 1 965), 
were  shown  to  be  erroneous  by  Anderson  and  Hubbs  (1981). 
Anderson  (1984)  suggested  zoarcid  relationships  are  within  Gos- 
line's  (1968)  Blennioidei,  especially  his  superfamily  Zoarceo- 
idae.  Eight  of  Gosline's  1 1  zoarceoid  families  were  recognized 
by  Anderson  (1984),  with  Lycodapodidae  and  Derepodichthyi- 
dae  synonymized  under  Zoarcidae  and  Stichaeidae  expanded  to 
include  Cryptacanthodidae  and  Neozoarcinae  (see  Makushok, 
1 96 1 ;  Peden  and  Anderson,  1978;  Anderson  and  Hubbs,  1981). 


582 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


A  definitive  phylogenetic  reconstruction  of  zoarceoid  rela- 
tionships is  not  presently  possible  without  a  more  thorough 
knowledge  of  the  anatomy  of  other  fishes  that  have  been  tra- 
ditionally allied  with  them.  Preliminary  phylogenetic  inferences 
were  made  by  Anderson  (1984),  who  also  discussed  relation- 
ships among  zoarcid  genera.  It  should  be  noted,  however,  that 
a  search  for  more  characters  is  still  in  progress.  Makushok  (1958) 
and  Springer  (1968)  suggested  zoarceoid,  or  "northern  blen- 
nioid"  relationships  were  not  close  to  the  "tropical  blennioids," 
a  fact  that  my  own  research  supports.  However,  for  the  con- 
venience of  the  reader,  information  on  the  early  life  history 
stages  of  zoarceoids,  excluding  Zoarcidae,  is  given  by  Matarese 
et  al.  (this  volume)  under  Blennioidea,  following  Nelson  ( 1 976). 

Since  there  is  a  dearth  of  knowledge  on  early  stages  and  since 
the  youngest  specimens  known  of  any  zoarcid  so  closely  resem- 
ble adults,  no  early  life  history  characters  have  helped  in  elu- 
cidating systematic  relationships  within  Zoarcidae,  or  the  zoar- 


cids  to  their  allies.  All  the  zoarceoids  are  characterized  by 
precocious  early  stages  (see  Matarese  et  al.,  this  volume),  but 
the  utility  of  these  forms  in  phylogeny  remains  untested.  Within 
Zoarcidae,  it  is  interesting  to  note  that  the  development  of  ce- 
phalic lateralis  pores  in  the  primitive  Gymnelus  viridis.  Melan- 
ostigma  pammelas.  and  Macrozoarces  americanus  takes  place 
over  a  much  longer  growth  period  (up  to  50-60  mm)  than  in 
the  more  derived  Bothrocara  (Okiyama,  1982a)  or  in  youngest 
stages  I  studied  of  Lycenchelys  (32  mm),  Lycodapus  (20  mm), 
or  Lycodes  (38  mm).  The  value  of  this  information  awaits  more 
complete  data  on  early  life  history  stages  of  all  zoarcids. 

Virginia  Institute  of  Marine  Science,  College  of  William 
AND  Mary,  Gloucester  Point,  Virginia  23062.  Present 
Address:  Department  of  Ichthyology,  California 
Academy  of  Sciences,  Golden  Gate  Park,  San 
Francisco,  California  941 18. 


Gobioidei:  Development 

D.  RUPLE 


GOBIOIDS  are  one  of  the  most  speciose  groups  of  fishes, 
comprised  of  approximately  2,000  species  or  ten  percent 
of  the  total  number  of  teleosts  in  the  world  (Cohen,  pers.  comm.). 
Various  workers  have  recognized  from  two  to  seven  major  fam- 
ilies of  gobioids,  based  on  adult  characters.  For  present  purposes 
I  will  recognize  seven  families'  Eleotridae,  Gobiidae,  Rhyacich- 
thyidae,  Kraemeriidae,  Gobioididae,  Trypauchenidae,  and  Mi- 
crodesmidae  after  Nelson  (1976). 

Development 

Larvae  are  known  for  less  than  5%  of  gobioid  species.  Eggs 
and  larvae  are  best  known  from  Japanese  waters  (e.g.,  Dotsu. 
1954,  1957,  1958,  1979;  Dotsu  and  Fujita,  1959;  Dotsu  and 
Mito,  1955,  1963;  Dotsu  and  Shiogaki,  1971;  Kobayashi  et  al., 
1973;  Shiogaki  and  Dotsu,  1971e,  1972c).  in  the  northeastern 
Atlantic  and  Mediterranean  Sea  area  (e.g.,  Petersen,  1917,  1919; 
Fage,  1918;  Lebour,  1919;  Sparta,  1934;  plus  summaries  in 
Padoa,  1956f;  and  Russell,  1976),  and  less  so  in  American  waters 
(e.g.,  Hildebrand  and  Cable,  1938;  Perlmutter,  1939;  Pearson, 
1 94 1 ;  and  Ruple,  in  prep.).  Most  of  these  descriptive  works  deal 
with  the  gobiids,  although  larvae  are  known  for  representatives 
of  all  families  except  Rhyacichthyidae  and  Kraemeriidae. 

Larvae  of  gobioids  are  fairly  distinctive  from  other  teleosts, 
but  considerable  variation  does  occur  within  the  suborder.  The 
diversity  of  characters  found  in  eggs  and  larvae  will  be  discussed 
in  the  following  section.  This  information  was  compiled  from 
published  literature  and  the  examination  of  gobioid  larvae. 


'  Hoese  (this  volume)  includes  Gobioididae  and  Trypauchenidae  in 
the  Gobiidae  subfamily  Amblyopinae  and  recognizes  Xenisthmidae  as 
a  distinct  family.  Eleotridae  is  changed  to  Eleotrididae. 


Eggs 

Eggs  are  known  for  eleotrids,  gobiids,  gobioidids,  and  micro- 
desmids  (Table  151).  Eggs  of  eleotrids  and  gobiids  are  generally 
ellipsoid  and  adhesive,  many  of  which  have  filamentous  strands. 
Eggs  range  in  size  from  as  small  as  0.40  x  0.32  mm  in  Eleotris 
avycep/jfl/a  (Eleotridae;  Dotsu  and  Fujita,  1959)  and  0.45  x  0.20 
mm  in  Evorthodus  lyricus  (Gobiidae;  Foster  and  Fuiman,  MS 
in  prep.)  to  3.8  x  1.3  mm  in  Pcrcottus  glehni  (Gobxiiisie:;  Kry- 
zhanovsky  et  al.,  1951)  and  5.5  x  0.9  mm  in  Acanthogobius 
Jlavimanus  (Gobiidae;  Dotsu  and  Mito,  1955).  Taenioides  ruh- 
icundus  (Gobioididae)  eggs  are  demersal,  adhesive  and  measure 
approximately  1.3  x  0.70  mm  (Dotsu,  1957)  whi\e  Gunnellich- 
ihys  (Microdesmidae)  eggs  are  spherical  (Smith,  1958a). 

Known  gobioid  eggs  usually  contain  numerous  small  oil  drop- 
lets within  the  yolk.  Newly  hatched  larvae  range  from  1.7  mm 
in  Aslerropteryx  semipunctatus  (Eleotridae;  Dotsu  and  Mito, 
1963)  to  7.0  mm  in  Chaenogobius  castanea  (Gobiidae;  Dotsu, 
1954). 

Larvae 

Gross  morphology.  —  Body  shape  of  gobioids  is  generally  slightly 
elongate  and  slender,  with  body  depth  usually  nearly  uniform 
rather  than  sharply  tapering  (Figs.  309-311).  Gobioidid  and 
microdesmid  larvae  are  moderately  elongate  and  slender  (Fig. 
311),  while  most  eleotrids  and  trypauchenids  are  only  slightly 
elongate  and  slender  (Fig.  311).  Microdesmids  have  the  most 
elongate  body  shape  of  any  known  gobioid  larvae.  Body  form 
within  the  gobiids  exhibits  the  greatest  variety,  ranging  from 
fairly  short  and  stout  (Gobiidae  Larva  1 ,  Fig.  309)  to  moderately 
elongate  and  slender  Luciogobius  elongatus  (Fig.  309).  These 
characteristic  body  shapes  are  usually  retained  from  the  larval 
through  adult  stages. 


RUPLE:  GOBIOIDEI 


583 


Fig.  309.  Larval  gobiids  from  top  to  bottom:  Gobiidae  Larva  1.  3.0  mm  NL  [AMS  (Australian  Museum  Sydney):  JML82/ 1-2-2];  Luciogohus 
elongalus.  1 2.0  mm  SL  (redrawn  from  Shiogaki  and  Dotsu  1 972c);  Gobiidae  Larva  2.  6.0  mm  SL (AMS:JML  1 6-10-7);  and  Microgobius  thalassinus 
8.4  mm  SL  [GCRL  (Gulf  Coast  Research  Uboratory):  02035]. 


584 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  310.     Larvae  of  gobiids  from  top  to  bottom:  E.xpedio  parvulns.  12.0  mm  SL  (redrawn  from  Shiogaki  and  Dotsu  I971e);  Astrabe  lactisella, 
11.1  mm  SL  (from  Dotsu  and  Shiogaki  1971);  Gobionellus  beleosoma.  8.6  mm  SL.  (GCRL:02038). 


The  gut  is  generally  straight  and  extends  to  about  midbody 
or  just  beyond  (~50%  to  65%  SL)  in  most  gobioids  (Figs.  309- 
311),  although  in  many  species  the  gut  is  slightly  looped  just 
anterior  to  the  vent  as  in  Microgohius  thalassinus  (Fig.  309).  In 
the  trypauchenid,  Trypauchen  microlepis  (Fig.  311).  the  gut  is 
considerably  shorter  (~39%  SL)  than  in  other  gobioids. 

A  prominent  feature  of  gobioid  larvae  is  a  large  gas  bladder, 
usually  situated  slightly  anterior  of  midbody  (Fig.  309).  The  gas 
bladder  is  located  just  posterior  to  the  pectoral  fin  in  Trypauchen 
microlepis  (Fig.  311)  and  is  smaller  and  less  pronounced  than 
in  most  other  gobioids.  In  small  larval  microdesmids  ( <4.0  mm) 
the  gas  bladder  is  located  at  about  mid-gut,  while  in  larger  larvae 
it  is  found  about  midbody,  near  the  posterior  portion  of  the  gut 
(Fig.  311).  The  prominent  gas  bladder  in  larvae  usually  disap- 
pears by  the  juvenile  stage,  but  is  retained  in  the  adults  of  some 
species  such  as  Gobiosoma  atronasum  (Colin,  1975). 

Eyes  of  known  gobioid  larvae  are  basically  round  or  slightly 


ovoid  in  shape.  The  elongate  gobioids  such  as  the  microdesmids 
and  gobioididshave  small  eyes  (<20%  HL)  while  most  eleotrids 
and  gobiids  have  somewhat  larger  eyes  (>20%  HL). 

The  head  is  of  moderate  length  (~  16%  to  34%  SL),  generally 
slightly  rounded  and  gently  sloping.  The  shape  of  the  head  changes 
drastically  in  many  species  as  they  transform  into  juveniles.  In 
microdesmids  such  as  Microdesmus  longipinms  and  Gunnelli- 
chthys  sp.  the  lower  jaw  becomes  hooked  and  protruding  during 
the  later  pelagic  larval  stages  (Fig.  311). 

The  lengths  of  dorsal  and  anal  fin  bases  vary  considerably 
and  are  useful  in  the  separation  of  gobioid  larvae  at  various 
taxonomic  levels.  The  lengths  of  the  fin  bases  are  related  to  the 
number  of  elements  and/or  the  spacing  between  the  individual 
elements,  which  varies  considerably.  Trypauchenids,  micro- 
desmids, and  some  gobioidids,  all  have  long  dorsal  and  anal  fin 
bases  (Fig.  311).  Some  eleotrids  (Eleolris  pisonis  and  Erotelis 
smaragdus)  and  various  gobiids  (Rhinogobius  similus.   Yono- 


RUPLE:  GOBIOIDEI 


585 


'^^^^$?;;j^:::^xs$^\\\\\\l\ 


Fig.  311.  Larvae  of  gobioids  from  top  to  bottom:  Trypauchen  microlepis  (Family  Trypauchenidae)  8.0  mm  SL  (ASMS:CFIT  2-11-78); 
Microdesmus  longipinms  (Family  Microdesmidae)  19.2  mm  SL  (GCRL:02036);  Gobioides  broussoniieti  (Family  Gobiodidae),  1 5.0  mm  SL  (GCRL: 
02037);  and  Dormilalor  macutatus  (Family  Eleotridae)  8.1  mm  SL  (GCRL:02039). 


gobius  boreus.  and  Luciogobiiis  elongatus;  Fig.  309)  have  shoin 
fin  bases,  with  few  closely  spaced  elements. 

Gobioids  transform  from  larvae  to  juveniles  over  a  wide  size 
range.  The  gobiids  Gobiosoma  bosci  and  G.  robusluin  begin 
transformation  at  ~7.0  mm,  while  some  microdesmids  main- 


tain a  pelagic  larval  existence  until  they  reach  lengths  of  ~25- 
35  mm. 

Meristics  and  fin   development. —Sequence  of  development, 
number  of  elements,  and  size  at  which  various  fins  develop  are 


586 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES- AHLSTROM  SYMPOSIUM 


Table  151.    Chae^acters  Useful  for  the  Separation  of  Gobioid  Larvae  to  the  Family  Level.  Characters  present  at  least  during  postflexion 
stage. 


Larvae 

Last  fin  to 

Eggs 

form  and  develop 

Body  shape 

Gul  length 

Eye  shape 

Continuous  or 
separated  dorsals 

Dorsal  and  anal 
fin  base  length 

full  complement 

General 

of  elements 

Eleotridae 

Most  ellipsoid 

Slightly  elon- 

Midbody to 

Round 

Separated 

Short  to 

First  dorsal 

and  adhesive 

gate  and 
slender 

slightly  be- 
yond 

long 

Gobiidae 

Most  ellipsoid 

Short  and 

Most  to  about 

Round  to 

Most  are 

Short  to 

Pelvic 

and  adhesive 

stout  to 
slightly  elon- 
gate and 
slender 

midbody  or 
slightly  be- 
yond 

slightly 
ovoid 

separated 

long 

Rhyacichthyidae' 

Separated 

Short 

Kraemeriidae' 

Continuous 

Long 

Gobioididae 

Demersal  and 
adhesive 

Moderately 
elongate  and 
slender 

Midbody  to 
slightly  be- 
yond 

Round 

Continuous 

Long 

Pelvic 

Trypauchenidae 

Slightly  elon- 
gate and 
slender 

1/3  body 
length  to 
slightly  be- 
yond 

Slightly 
ovoid 

Contmuous 

Long 

Pelvic 

Microdesmidae 

Spherical 

Moderately 
elongate  and 
slender 

Midbody  to 
slightly  be- 
yond 

Round 

Continuous 

Long 

Pelvic 

'  Larvae  unknown,  character  states  projected  from  adult  conditions. 


useful  in  distinguishing  gobioid  larvae  at  all  levels.  The  sequence 
of  fin  development  is  similar  in  most  gobioids.  although  it  vaiies 
somewhat  in  eleotrids  and  gobioidids.  Numbers  of  fin  rays  and 
spines  vary  greatly  among  the  gobioids  and  are  particularly  use- 
ful in  distinguishing  gobioids  at  the  family  and  species  levels 
(Hoese-).  Degree  of  fin  development  at  different  sizes  is  helpful 
in  separating  certain  species  of  larvae,  particularly  if  complete 
developmental  series  are  available. 

Median  finfolds  and  pectoral  fins  are  present  at  hatching  or 
develop  in  early  larvae  of  all  known  gobioids.  The  pelvic  fin  is 
the  last  fin  to  form,  usually  beginning  during  the  flexion  or 
postflexion  stages. 

The  caudal  fin  is  the  first  fin  to  form  diflTerentiated  rays,  be- 
ginning during  the  flexion  stage.  Gobioids  usually  have  17  seg- 
mented principal  caudal  rays  and  numerous  secondary  rays  that 
are  usually  all  differentiated  by  the  postflexion  or  transforming 
stages. 

The  second  dorsal  and  anal  fins  are  next  to  develop  in  eleo- 
trids, gobiids,  and  microdesmids.  The  anterior  and  middle  ele- 
ments are  first  to  form  and  development  generally  proceeds 
posteriorly.  It  is  difficult  to  distinguish  the  dorsal  spines  from 
rays  in  the  continuous  dorsal  fin  of  microdesmids  during  the 
larval  stages,  but  they  are  usually  shorter  than  the  rays.  The  first 
element  of  the  anal  fin  and  the  first  element  of  the  second  dorsal 
fin  initially  develop  as  rays,  but  later  transform  into  spines  in 
most  eleotrids  and  gobiids.  The  presence  of  a  continuous  (Afi- 
crodesmus  longipinnis.  Fig.  311)  or  separated  (Gobiidae  Larva 
2,  Fig.  309)  dorsal  fin  is  useful  in  family  diagnosis  of  gobioids. 
This  character  varies  considerably  from  the  continuous  dorsal 


■  Hoese  (this  volume)  provides  a  meristic  table  for  gobioid  families. 


in  microdesmids,  gobioidids,  and  trypauchenids  to  the  widely 
separated  fins  of  the  gobiid  Periophthalmus  cantonensis  (Ko- 
bayashi  et  al.,  1973). 

The  first  dorsal  and  pectoral  fins  are  usually  next  to  complete 
development  in  microdesmids  and  gobiids,  while  the  full  com- 
plement of  first  dorsal  spines  is  last  to  form  in  eleotrids.  While 
complete  developmental  series  of  eleotrids  are  sparse,  it  appears 
that  the  posteriormost  first  dorsal  spines  form  after  the  full 
complement  of  pectoral  and  pelvic  fin  elements  are  present  (e.g., 
Dormitator  maculatus,  Eleotris  ptsonis.  and  Erotelis  smarag- 
dus).  Inthegobioidid,  Taenoides cirratus (9 .3  mm),  Dotsu(1958) 
depicts  the  pectoral  fin  to  be  the  last  fin  to  obtain  its  full  com- 
plement of  elements.  First  dorsal  spines  range  from  2  or  3  in 
some  gobiids  (e.g.,  Claringer  cosnmrus)  to  28  in  some  micro- 
desmids (e.g.,  Microdesmus  longipinnis).  The  gobiid  Luciogo- 
bius  elongatus  (Fig.  309)  lacks  a  first  dorsal  fin  entirely  (Shiogaki 
and  Dotsu,  1972c).  Pectoral  fin  rays  range  from  3  to  26. 

In  known  gobiids,  trypauchenids,  and  microdesmids  the  pel- 
vic fin  is  last  to  form  and  complete  development.  Development 
of  the  pelvic  fins  in  gobioids  varies  greatly  between  families  and 
within  certain  families  such  as  the  gobiids.  Some  gobioids  have 
strongly  united  pelvics  that  form  a  cup-shaped  disc  (Fig.  309) 
at  a  very  early  age,  while  adult  Rhyacichthys  aspro  (Rhyacich- 
thyidae) have  widely  separated  pelvics.  Eleotrids,  trypauchen- 
ids, microdesmids,  kraemeriids,  and  gobioidids  usually  have 
separated  or  weakly  united  pelvics.  Pelvic  fins  in  gobiids  range 
from  strongly  united,  forming  a  disc  to  weakly  connected  at  the 
base  to  totally  separated  (usually  in  species  with  reduced  pelvics; 
Fig.  309).  The  size  at  which  pelvics  develop  is  an  important 
character  in  the  separation  of  some  gobiid  genera  and  species. 
One  pelvic  spine  and  2-5  rays  occur  in  gobioids.  Expedio  par- 
vulus  (Gobiidae)  lacks  pelvic  fins  entirely  (Fig.  310). 


RUPLE:  GOBIOIDEI 


587 


Table  151.    Extended. 


Larvae 

Pelvic  fin 

Gas  bladder 

General  bodv 

Prominent  ventral 

condition 

pigment 

pigment 

pigment 

Separated 

Dorsal  surface 

Moderate 

Present 

United  to  sepa- 

Dorsal surface 

Sparse  to 

Usually 

rated 

or  dorsal  and 

posterior 

surface 

heavy 

present 

Separated 

Usually  sepa- 

rated 

Usually  weakly 

Dorsal  and 

Sparse 

Absent 

united 

posterior 
surface  or 
lacking 

Usually  weakly 

Lacking 

Sparse 

Absent 

united 

Separated 


Dorsal  surface  Moderate  Present 


Various  other  meristic  characters  are  useful  in  the  separation 
of  gobioid  fishes.  Branchiostegal  rays  number  from  5  to  6.  Myo- 
meres/vertebrae  range  from  25  in  Eviola  infulata  (Eleotridae) 
to  76  in  some  of  the  microdesmids. 

Pigmentation.— V\%men\a\\on  on  the  gas  bladder  and  along  the 
ventral  surface  of  the  body  are  considered  to  be  characteristic 
of  most  gobioid  larvae.  Melanistic  pigmentation  in  gobioid  lar- 
vae varies  considerably,  from  the  heavily  pigmented  gobiid  .-15- 
trabe  lactisella  (Fig.  310)  to  the  sparsely  pigmented  gobioidids 
(Fig.  311)  and  trypauchenids.  Larval  gobiids,  eleotrids,  and  mi- 
crodesmids generally  have  a  moderate  amount  of  pigmentation. 
Pigmentation  patterns  are  especially  useful  in  separating  larvae 
at  the  generic  and  specific  levels. 

Trypauchenid  larvae,  Ctenotrypauchen  microcephalus  and 
Trypauchen  microlepis  (Fig.  311)  generally  lack  pigmentation 
except  for  two  spots  of  pigment  along  the  anterior  poilion  of 
the  caudal  fin,  while  the  gobioidids  Taenioides  cirratus  and  Go- 
bioides  hroussonneti  totally  lack  pigmentation  except  for  a  pig- 
mented gas  bladder  in  Gobioides  broitssonneti  (Fig.  311). 

Pigmentation  on  the  gas  bladder  is  a  prominent  feature  of 
most  known  gobioids.  The  most  common  condition  is  pigment 
on  the  dorsal  surface,  as  in  Microgohius  thalassinus  (Fig.  309) 
and  appears  in  most  known  gobioids.  Dorsal  and  posterior  gas 
bladder  pigment  is  known  only  in  Gobionclhis  species  and  Go- 
bioides broussonneti  (Figs.  310,  311).  Trypauchen  microlepis 
(Fig.  311),  Ctenotrypauchen  microcephalus.  and  Taenioides  cir- 
ratus are  the  only  known  gobioid  species  which  lack  gas  bladder 
pigmentation  entirely  during  their  early  development. 

The  most  pronounced  pigmentation  occurring  in  many  eleo- 
trids, gobiids,  and  microdesmids  is  that  found  along  the  ventral 


surface  of  the  body,  in  the  region  of  the  gut  and  anal  fin  base. 
Along  the  anal  fin  base,  this  pigmentation  often  occurs  on  in- 
ternal as  well  as  external  surfaces. 

Pigmentation  is  often  found  in  eleotrids,  gobiids,  and  micro- 
desmids on  the  caudal  peduncle,  along  the  dorsal  surface  of  the 
body,  on  the  otic  capsule,  on  the  tip  of  the  lower  jaw,  along  the 
mid-lateral  posterior  portion  of  the  body,  and  on  various  fins. 
While  pigmentation  often  appears  very  similar,  the  subtle  vari- 
ations are  frequently  useful  in  the  separation  of  larval  gobioids. 

Contribution  of  laiA'ae  to  systematics 

Gobioid  larvae  have  not  been  previously  examined  in  terms 
of  contributing  to  the  understanding  of  systematic  relationships, 
but  I  believe  they  will  be  of  great  use  in  the  future.  A  preliminary 
phenetic  overview  of  gobioids  based  on  characters  available  in 
larvae  (representing  less  than  5%  of  the  total  number  of  gobioid 
species),  presents  some  interesting  groupings.  Known  larvae  from 
three  eleotrid  genera;  Erotelis.  Eleotris.  and  Dormitator  seem 
to  form  a  cohesive  group.  Shared  characters  include:  gross  body 
and  head  shape,  short  dorsal  and  anal  fin  bases,  separated  pel- 
vies,  gut  length  (~55%  to  57%  SL),  dorsal  gas  bladder  and 
ventral  pigmentation,  late  development  of  the  first  dorsal  fin, 
and  separation  of  the  two  dorsal  fins.  Microdesmid  larvae  from 
the  genera  Microdesmus.  Cerdale.  and  Gunnellichthys.  likewise, 
all  appear  quite  similar  to  each  other,  based  on  the  following: 
gross  body  and  head  shape,  connected  dorsal  fins,  long  dorsal 
and  anal  fin  base,  high  vertebrae  number,  reduced  pelvic  and 
pectoral  fins,  gas  bladder  and  dorsal  and  ventral  body  pigmen- 
tation. 

While  the  above  mentioned  eleotrid  and  microdesmid  groups 
appear  fairly  cohesive  as  well  as  distinct  from  other  gobioids, 
the  family  Gobiidae  seem  to  be  in  some  respects  a  catch-all 
group.  Currently,  many  diverse  types  of  gobies  are  included 
within  the  family  Gobiidae  (some  250  genera  and  well  over 
1,000  species).  It  is  possible  that  larvae  may  present  us  with 
additional  characters  that  may  help  to  better  define  the  group. 
The  use  of  only  adult  characters  has  led  many  workers  to  debate 
the  rank  of  many  taxa,  for  example  Nelson's  (1976)  families 
Trypauchenidae  and  Gobioididae  have  been  relegated  to 
subfamilial  status  within  the  Gobiidae  (Hoese,  this  volume)  or 
lower  by  other  workers.  Although  the  use  of  larval  characters 
alone  will  not  define  gobioid  families  they  may  allow  a  better 
understanding  of  relationships.  Known  larvae  of  Trypauchen- 
idae and  Gobioididae  exhibit  characters  that  are  distinctive  or 
unique  to  these  taxa.  Trypauchen  microlepis  has  the  shortest 
and  most  acutely  looped  gut  of  any  gobioid  and  is  one  of  only 
three  species  that  lack  gas  bladder  pigmentation  (others  are  a 
trypauchenid  and  a  gobioidid).  Pectorals  are  also  more  reduced 
than  in  other  gobioids.  The  long  continuous  dorsal  fin  and  long 
anal  fin  base  are  not  shared  among  most  gobioids.  Gobioides 
broussonneti  also  has  a  long  continuous  dorsal  fin  and  long  anal 
fin  base.  It  is  one  of  only  two  known  gobioid  genera  with  dorsal 
and  posterior  gas  bladder  pigment  (the  other  being  the  gobiid 
genus  Gobionellus).  and  is  one  of  the  most  sparsely  pigmented 
gobioids  known. 

More  descriptive  work  needs  to  be  completed  on  the  taxo- 
nomic  level  of  both  adults  and  larvae  before  the  full  value  of 
ontogentic  characters  in  gobioid  systematics  can  be  adequately 
assessed. 

Gulf  Coast  Research  Laboratory,  East  Beach  Drive,  Ocean 
Springs,  Mississippi  39564. 


Gobioidei:  Relationships 
D.  F.  HoESE 


APPROXIMATELY  500  genera  and  2,000  species  of gobioid 
fishes  have  been  named.  Currently,  about  270  genera  are 
recognized,  and  it  is  estimated  that  the  group  contains  between 
1,500  and  2,000  species.  About  50  families,  subfamilies,  and 
tribes  have  been  named.  Gobioid  fishes  are  distributed  through- 
out much  of  the  tropical,  subtropical  and  temperate  regions  of 
the  world,  occurring  in  a  variety  of  habitats  in  fresh,  brackish, 
and  coastal  marine  waters  to  depths  of  about  220  meters  on  the 
continental  shelf  Of  the  six  extant  families  recognized  here, 
three  (Eleotrididae,  Gobiidae,  and  Microdesmidae)  are  world- 
wide, and  three  (Xenisthmidae,  Rhyacichthyidae,  and  Krae- 
mariidae)  are  restricted  to  the  Indo-Pacific.  Most  species  of 
gobioid  fishes  are  benthic,  but  some  are  pelagic,  many  are  bur- 
rowers,  and  many  live  in  burrows  constructed  by  other  organ- 
isms. 

Much  of  the  early  history  of  the  classification  of  gobioid  fishes 
has  been  summarized  by  Iljin  (1930),  Koumans  (1931),  and 
Miller  (1973).  Early  classifications,  based  on  external  features 
were  provided  by  Gunther  (1861),  Bleeker  (1874),  Jordan  (1923), 
and  Berg  ( 1 940).  Sanzo  (1911)  published  the  first  extensive  study 
of  the  lateralis-system  pores  and  papillae,  characters  which  have 
come  into  wide  usage  in  the  last  20  years  at  the  generic  and 
specific  levels.  Regan  (1911c)  presented  the  first  classification 
based  largely  on  osteological  characters.  He  established  the  fam- 
ily Psammichthyidae  (=Kraemariidae),  and  provisionally  placed 
it  with  the  gobioids,  a  placement  which  was  not  accepted  until 
relatively  recently.  The  study  of  Regan  was  largely  confined  to 
the  cranial  osteology,  pectoral  girdle,  and  vertebral  numbers. 
Gosline  (1955)  examined  the  osteology  of  a  few  representatives 
of  the  major  groups  of  gobioid  fishes,  and  gave  evidence  for  the 
placement  of  microdesmids  and  kraemariids  among  the  go- 
bioids. Takagi  (1950,  1953)  contributed  to  the  classification 
based  on  examination  of  scales  and  the  glossohyal,  and  later 
(1966)  published  an  extensive  paper  on  the  distribution  of  the 
group.  Akihito  (1963,  1967)  studied  the  scapula  in  a  number  of 
species,  and  later  (1969)  presented  one  of  the  most  detailed 
studies  of  the  higher  classification  of  gobioid  fishes,  dealing  with 
major  osteological  features  of  7 1  genera  and  85  species,  but  did 
not  present  a  classification.  Miller  ( 1 973)  described  the  osteology 
of  Rhyacichthys,  and  presented  a  largely  new  classification  of 
the  group.  Birdsong  (1975)  presented  information  on  the  prim- 
itive character  states  for  several  osteological  characters  and  in- 
dicated presumed  trends  for  each  character.  He  also  criticized 
the  classification  of  Miller  and  recommended  a  return  to  the 
traditional  classification. 

Dawson  ( 1 974b)  characterized  the  Microdesmidae  and  (1973) 
summarized  distributional  information  on  Indo-Pacific  species. 
Rofen  (1958)  reviewed  the  Kraemariidae.  Matsubara  and  Iwai 
(1959)  described  the  osteology  of  Kraemaria  scxradiata.  Obrhe- 
lova  (1961)  described  a  new  family  of  gobioid  fishes  (Pirsken- 
iidae)  from  Oligo-miocene  fossil  material  from  Europe. 

Few  studies  have  been  carried  out  on  the  relationships  of  the 
suborder  to  other  fishes.  Most  early  workers  considered  the 
group  related  to  perciform  or  scorpaeniform  fishes.  McAllister 


(1968)  and  Freihofer  (1970)  suggested  a  relationship  with  the 
Paracanthopterygii.  Other  workers  have  accepted  a  perciform 
derivation  (Miller,  1973:  Springer,  1983;  Gosline,  1955),  al- 
though Gosline  (1971)  suggested  that  the  group  might  eventually 
be  regarded  as  a  distinct  order  based  on  the  structure  of  the 
suspensorium  and  the  caudal  skeleton.  No  sister  group  has  been 
postulated. 

Gobioid  fishes  are  characterized  by  the  following  features:  no 
parietals;  a  pelvic  intercleithral  cartilage:  interhyal  displaced 
away  from  the  dorsal  end  of  the  symplectic:  a  gap  between 
symplectic  and  preoperculum:  no  orbitosphenoid  or  basisphe- 
noid:  lacrimal  typically  present,  extending  over  maxilla,  but  not 
forming  lower  margin  of  orbit:  only  one  other  suborbital  rarely 
present;  fourth  basibranchial  cartilaginous;  penultimate  verte- 
bra with  a  short  expanded  neural  spine  and  an  elongate  ex- 
panded hemal  fused  to  centra;  caudal  skeleton  with  one  to  three 
epurals,  a  small  free  parhypural,  an  enlarged  lower  hypural  plate 
articulating  with  and  sometimes  fused  with  urostyle.  an  enlarged 
upper  hypural  plate  fused  to  urostyle,  and  a  small  free  upper 
hypural;  procurrent  caudal  rays  articulate  with  cartilaginous 
plates;  lateral  line  usually  absent  on  body,  canals  often  devel- 
oped on  head,  suborbital  canal  and  mandibular  canal  usually 
absent;  first  spine  or  ray,  when  spine  absent,  associated  with 
proximal  elements  of  two  pterygiophores  (median  element  of 
first  pterygiophore  of  second  dorsal  fin  rarely  present);  last  two 
rays  of  second  dorsal  and  anal  fins  closely  spaced  and  articulating 
with  a  single  pterygiophore  in  each  fin.  Meristics  are  given  in 
Table  152. 

The  following  groups  are  recognized: 

Rhyacichthyidae. —  The  monotypic  family  Rhyacichthyidae  is 
the  most  primitive  gobioid  fish  in  the  following  features:  bran- 
chiostegals  6;  mesopterygoid  and  dorsal  postcleithrum  present; 
lateral  line  present  on  body;  an  anterior  sclerotic;  lacrimal  and 
one  additional  suborbital  present;  3  epurals;  interhyal  adjacent 
to  dorsal  end  of  symplectic;  3  posttemporals;  infraorbital  and 
mandibular  head  canals  present;  scales  with  multiple  rows  of 
ctenii.  Its  specializations  are  related  to  adaptations  to  fast  flow- 
ing rocky  streams  and  include:  thickened  muscular  pelvic  fins, 
small  mouth,  placed  ventrally  and  anteriorly. 

Eleotrididae— The  eleotridids,  largely  confined  to  freshwater 
and  estuarine  environments,  are  currently  definable  on  the  basis 
of  the  following  primitive  features:  branchiostegal  rays  6;  pelvic 
fins  widely  separate,  pelvic  girdle  with  a  short  post-pelvic  pro- 
cess, extending  well  beyond  last  pelvic  ray,  pelvic  rays  in  line 
with  pelvic  spine;  mesopterygoid  and  dorsal  postcleithrum  gen- 
erally present;  interorbital  normally  broad;  caudal  peduncle  long, 
generally  longer  than  second  dorsal  base;  palatine  normally  more 
or  less  L-shaped,  with  a  short  ethmoid  process,  articulating 
medially  with  lateral  ethmoid;  scapula  normally  completely  os- 
sified; anterior  sclerotic,  suborbital  (other  than  lacrimal),  post- 
temporals,  and  median  element  of  first  pterygiophore  of  second 
dorsal  fin  usually  absent;  first  basibranchial  cartilaginous,  ba- 


588 


HOESE:  GOBIOIDEI 


589 


Table  152.    Selected  Meristics  for  Gobioid  Families  and  Subfamilies. 


Group 


Branchi- 

oslegal 

rays 


Dl 


D2 


PI 


P2 


Epurals 


Segmented 
caudal  rays 


Vertebrae 


Rhyacichthyidae  6 

Eleotrididae  6 

Xenisthmidae  6 

Microdesmidae 

Microdesminae  5 

Oxymetopontinae  5 

Gobiidae 

Oxudercinae  5 

Amblyopinae  5 

Sicydiinae  5 

Gobiinae  5 

Kxaemariidae  5 


VII                1,8-9              1,8-9  21-22  1,5  3  17  12+16  =  28 

III-X              1,6-17            1,6-13  13-21  1,5  1-2  15-17  10-18+11-19  =  24-36 

0-VI  0-1,9-32            1,9-25  17-21       0-1.1-5  1-2  15-17  10-18+16-28  =  26-46 

XX-XXVIII  26-66              23-61  10-15  1.2-4         1  15-17  42-76 

VI                1,9-37            1,9-36  15-26  1,4-5         1  17  10-11  +  15-16  =  26 

V-VIII  0-1,10-30  0-1,10-30  10-21  1,5  2  17  10+16  =  26 

VI-VIIl           16-50  0-1,14-50  13-21  1,5  1-2  17  10+16-26  =  26-36 

VI                 1,9-11            1,9-11  15-23  1,5  1  17  10+16  =  26 

0-X  0-1,5-19  0-1,5-19  11-25  1,4-5  1-2  13,16-17  10-16+14-21=25-36 

IV-V             13-19               1,11-15  3-10  1,5  1  11  10-14+16-17  =  26-31 


sibranchials  2  and  3  present;  pterosphenoid  present,  corono- 
meckelian  bone  present;  pterygiophores  of  two  dorsal  fins  nor- 
mally continuous,  without  an  intemeural  gap.  One  group 
(Leplophilypnits  and  Gohiomorphus  and  relatives)  are  special- 
ized in  having  an  intemeural  gap  (an  intemeural  space  without 
a  pterygiophore)  between  the  two  dorsal  fins.  Members  of  the 
group  also  have  often  lost  several  eleotridid  primitive  features, 
such  as  the  mesopterygoid  and  dorsal  postcleithrum.  Some  are 
specialized  in  having  an  unossified  scapula  and  a  single  epural. 
Other  eleotridids  consistently  have  2  epurals  and  a  well  ossified 
scapula.  The  group  includes  about  40  genera  and  the  following 
named  taxa:  Butinae,  Belobranchinae.  Gobiomoridae.  Hypse- 
leotrini,  Milyeringidae,  Ophiocarinae,  and  Philypni.  Whether 
any  of  these  are  recognizable  must  await  further  study. 

Xenisthmidae— This  coral  reef  group,  restricted  to  the  Indo- 
Pacific,  is  treated  extensively  by  Springer  (1983)  and  is  distinc- 
tive in  the  following  specializations:  lower  lip  with  a  free  ventral 
margin;  ascending  process  of  premaxilla  absent  or  rudimentary; 
rostral  ossified  and  functionally  replacing  ascending  process  of 
premaxilla;  first  basibranchial  ossified;  basibranchials  2-4  ab- 
sent; no  pterosphenoid  or  coronomeckelian  bone;  intemeural 
gap  present  between  two  dorsal  fins.  The  two  genera  studied 
lack  the  dorsal  postcleithrum  and  the  mesopterygoid.  The  group 
includes  4  genera. 

Microdesmidae— The  group  possesses  the  following  primitive 
features:  maxilla  more  or  less  L-shaped,  with  a  very  short  inner 
process  articulating  medially  with  lateral  ethmoid;  usually  sep- 
arate pelvic  fins,  without  an  interspinal  membrane.  The  group 
is  uniquely  specialized  in  having  a  very  long  posterior  pelvic 
process.  Other  specializations  include  the  strongly  compressed 
head  and  body,  with  lateral  eyes;  5  branchiostegal  rays;  one 
epural;  dorsal  postcleithrum  and  mesopterygoid  absent.  Trends 
in  the  group  include  reduction  of  pelvic  rays,  the  tendency  for 
the  scales  to  become  nonimbricate,  and  the  development  of  a 
very  long-based  second  dorsal  fin.  Two  subfamilies  are  recog- 
nized here,  but  further  studies  may  show  both  to  be  distinct 
families. 

Microdesminae.— The  specializations  include:  maxilla  with  a 
long  strut-like  anterior  projection;  body  very  elongate,  with  a 
single  dorsal  fin  attached  to  or  reaching  near  caudal  fin;  dentary 


with  a  long  ventral  process  at  anterior  tip.  The  worldwide  group 
includes  5  genera  and  the  following  named  taxa:  Cerdalidae, 
Gunnellichthyidae,  and  Paragobioididae. 

Ptereleotrinae. — The  specializations  include:  mouth  almost  ver- 
tical; articular  process  of  premaxilla  absent  or  fused  with  as- 
cending process;  a  single  pterygiophore  precedes  the  first  hemal 
spine.  The  worldwide  group  includes  6  genera  (2  undescribed) 
and  the  following  named  taxa:  Nemateleotrinae.  Pogonoculinae, 
Oxymetopontinae.  In  addition  both  subfamilies  of  microdes- 
mids  contain  several  specializations  sometimes  found  in  Go- 
biidae, such  as  the  interlocking  of  the  anterior  preopercular 
process  with  the  dorsal  end  of  the  symplectic  and  the  expanded 
dorsal  flange  of  the  sphenotic  reaching  to  the  supraoccipital. 

Gobiidae.  — In  some  genera  primitive  features  are  found,  such 
as  the  ventral  postcleithrum.  2  epurals,  and  separate  head  canals 
between  the  eyes.  Specializations  include:  pelvic  fins  usually 
connected  to  form  a  cup-shaped  disc,  often  separate  in  coral 
reef  genera,  but  interspinal  membrane  usually  present;  pelvic 
spine  displaced  forward  and  ventrally,  not  in  line  with  rays; 
mesopterygoid  and  dorsal  postcleithrum  absent;  palatine  nor- 
mally T-shaped,  but  L-shaped  in  some  specialized  genera;  eth- 
moid process  of  palatine  extends  across  front  of  lateral  ethmoid, 
articulating  with  proximal  base  of  lateral  ethmoid  or  more  com- 
monly with  median  ethmoid;  maxilla  generally  without  an  an- 
terior process;  median  ethmoid  displaced  ventrally;  an  inter- 
neural  gap  present  between  two  dorsal  fins  (except  in  Trypauchen 
and  relatives.  There  may  be  one  or  two  dorsal  fins,  and  most 
genera  have  17  segmented  caudal  rays,  rarely  13  or  16.  Several 
subfamilies  have  been  recognized.  Four  are  recognized  here,  but 
further  studies  may  considerably  expand  the  number. 

Oxudercinae.— Tongue  fused  to  floor  of  mouth;  a  single  pte- 
rygiophore precedes  first  hemal  arch;  teeth  flattened;  second 
dorsal  fin  usually  long  based;  eyes  displaced  forward  and  up- 
ward; 2  epurals,  lateral  process  of  sphenotic  large  and  not  in 
contact  with  eye.  The  group  occurs  in  mud  and  mangrove  areas 
in  all  tropical  areas,  except  the  New  World.  The  group  contains 
about  10  genera  and  the  following  named  taxa:  Apocrypteidae, 
Boleophthalminae.  Periophthalmidae. 

Amblyopinae.— Tongue  fused  to  floor  of  mouth;  2  or  3  pteryg- 
iophores precede  first  hemal  spine;  a  single  dorsal  fin  reaching 


590 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


to  near  or  fused  with  caudal;  eyes  rudimentary  and  placed  for- 
ward in  orbit;  lateral  process  of  symplectic  large;  2  epurals.  The 
worldwide  group  occurs  in  estuaries  or  off  river  mouths,  and 
includes  about  10  genera  and  the  following  named  taxa:  Go- 
bioididae,  Taenioididae,  Trypauchenidae. 

Sicydiinae.— Tongue  fused  to  floor  of  mouth  or  free  only  at  tip, 
highly  modified  jaw  suspension;  thickened  and  highly  branched 
pelvic  rays  and  fleshy  pads  at  tips  of  pelvic  spines.  The  world- 
wide group  occurs  in  freshwater  and  includes  about  5  or  more 
genera  and  the  following  included  taxa:  Sicydiaphiinae  (in  part). 

Gobiinae.— The  worldwide  group  includes  about  200  genera 
and  is  not  easily  definable.  The  group  includes  the  following 
named  higher  taxa:  Aphyinae,  Austrolethopinae,  Benthophiliae, 
Brachygobii,  Calleleotrinae,  Chaeturichthyi,  Croilinae,  Crystal- 
logobiinae,  Doliichthyidae,  Gobiodontinae.  Gobiosomini, 
Gymnogobiini,  Latrunculini,  Lebetinae,  Leioterinae,  Luciogo- 
biinae,  Platygobii,  Rhinogobiinae,  Triaenophorichthyini,  Tri- 
dentigeriinae,  Valencienninae. 

Kraemariidae(=Psammichthyidae.)— The  family  agrees  in  most 
features  with  the  Gobiidae,  being  specialized  in  having  a  large 
amount  of  cartilage  in  the  skeleton  and  3  pectoral  radials.  The 
group  is  restricted  to  the  Indo-Pacific  and  includes  2  genera. 

Relationships 

Most  workers  have  generally  agreed  that  the  Rhyacichthyidae 
and  Eleotrididae  represent  the  most  primitive  gobioid  fishes 
characterized  by  6  branchiostegal  rays,  a  mesopterygoid,  and 
dorsal  postcleithrum.  In  addition  other  primitive  features,  not 
found  in  gobiids  are  sometimes  present,  such  as  an  anterior 
sclerotic,  lower  suborbital  (other  than  lacrimal),  and  extrascap- 
ulae.  Most  other  features  generally  retain  a  primitive  nature  in 
eleotridids,  such  as  2  epurals,  ossified  scapula,  head  canals,  when 
present,  separate  between  eyes,  and  a  ventral  postcleithrum. 
Gobiids,  microdesmids,  and  kraemariids  have  5  branchiostegal 
rays  and  lack  a  mesopterygoid  and  dorsal  postcleithrum  (with 
over  two  thirds  of  the  genera  examined).  These  differences  in 
organizational  grades  have  lead  some  workers  to  suggest  that 
the  advanced  gobiid  level  of  organization  may  be  polyphyletic 
(Springer,  1983). 

The  primary  innovate  character  defining  the  gobiid  fishes  is 
the  development  of  a  pelvic  cup-shaped  disc,  formed  by  mem- 
branes connecting  the  inner  pelvic  rays  and  two  pelvic  spines 
(interspinal  membrane  or  frenum);  with  the  forward  and  ventral 
rotation  of  the  pelvic  spines  on  the  pelvic  girdle.  It  has  been 
shown  that  reef  gobiids  often  have  secondarily  separate  pelvic 
fins  (Hoese,  1971),  although  most  species  retain  a  rudiment  of 
the  interspinal  membrane  and  the  typical  gobiid  pelvic  spine 
orientation.  Consequently,  the  question  of  whether  gobiid  fishes 
are  monophyletic  depends  in  part  on  whether  the  disc  has  evolved 
independently  in  various  gobiid  groups.  Studies  of  other  gobiid 
specializations,  although  incomplete,  have  not  indicated  that 
gobiids  are  polyphyletic.  For  example  Regan  (191 1  c)  first  noted 
that  the  eleotridids  have  an  L-shaped  palatine  and  gobiids  a 
T-shaped  palatine.  In  eleotridids  the  ethmoid  process  of  the 
palatine  is  short  and  articulates  directly  with  the  middle  of  the 
lateral  ethmoid,  while  in  gobiids  the  ethmoid  process  is  typically 
long,  extending  across  to  the  median  ethmoid,  which  is  displaced 
ventrally.  Similarly  in  gobiids  there  is  an  intemeural  gap  be- 


tween the  two  dorsal  fins  (a  space  between  two  neural  arches 
without  a  pterygiophore).  Primitively  in  eleotridids,  the  pteryg- 
iophores  of  the  two  dorsal  fins  are  continuous,  without  a  gap. 
From  the  relationship  between  the  pterygiophores  of  the  second 
dorsal  and  the  anal  fins,  it  appears  that  the  gap  in  gobiids  forms 
from  a  posterior  shift  of  the  second  dorsal  fin.  The  intemeural 
gap  also  occurs  in  Rhyacichthys  and  Xenisthmus,  and  several 
eleotridid  genera  from  New  Guinea,  Australia,  and  New  Zealand 
(Gobiomorphus,  Philypnodon,  Grahamtchthys,  and  two  new 
genera)  and  the  Central  American  genus  Leptophilypnus.  Struc- 
tural comparisons  indicate  that  Rhyacichthys  probably  obtained 
the  gap  by  loss  of  a  dorsal  spine  or  forward  shift  of  the  posterior 
dorsal  spines.  It  is  currently  unknown  whether  the  Xenisthmidae 
and  the  Gobiomorphus- Leptophilypnus  group  are  convergent 
with  gobiids  or  represent  sister  groups.  Both  groups  sometimes 
lose  primitive  eleotridid  features  such  as  the  mesopterygoid  and 
dorsal  postcleithrum.  In  general  body  form  the  Gobiomorphus- 
Leptophilypnus  group  most  closely  approach  the  gobiid  body 
form  expected  of  an  ancestral  gobiid.  Although  currently  placed 
with  the  eleotridids  further  studies  are  underway  to  determine 
the  relationships  of  the  genera  in  the  group. 

The  microdesmids  also  represent  a  gobiid  level  of  organiza- 
tion, in  lacking  several  primitive  features,  but  their  relationships 
to  other  gobioid  fishes  are  unclear.  The  group  is  characterized 
primarily  by  the  unique  specialization  of  having  an  elongate 
posterior  pelvic  process.  The  two  subfamilies  have  other  spe- 
cializations in  common  and  show  similar  trends,  with  the  Pter- 
eleotrinae  representing  the  primitive  sister  group.  The  micro- 
desmids retain  a  primitive  palatine-ethmoid  articulation,  and 
the  posterior  pelvic  process  probably  represents  an  elongation 
of  the  short  posterior  pelvic  process  of  eleotridids.  Microdes- 
mids share  with  gobiids  the  loss  of  the  anterior  branchiostegal 
ray.  The  strong  compression  of  the  head  may  have  lead  inde- 
pendently to  the  loss  of  the  anterior  branchiostegal  ray.  Unfor- 
tunately no  immediate  sister  group  is  known,  although  on  the 
basis  of  the  intemeural  gap,  the  Xenisthmidae  represent  a  pos- 
sible group.  Although  the  inner  rays  of  the  two  pelvic  fins  are 
sometimes  connected  in  microdesmines,  no  species  is  known 
with  an  interspinal  membrane.  The  microdesmines  have  pre- 
sumably secondarily  lost  the  intemeural  gap.  A  similar  situation 
occurs  in  the  gobiid  Trypauchen,  where  a  single  long-based  dor- 
sal fin  is  present. 

The  kraemariids  appear  closest  to  gobiids.  Whether  the  group 
will  remain  a  family  is  uncertain,  since  the  group  shows  some 
similarity  to  the  gobiid  Parkraemaria. 

Since  no  immediate  sister  group  has  been  postulated  for  go- 
bioid fishes,  relationships  of  the  more  primitive  groups  are  un- 
clear. Miller  (1973)  and  Springer  (1983)  have  recognized  only 
two  gobioid  families,  Rhyacichthyidae  and  Gobiidae.  Springer 
(1983)  has  suggested  that  the  Rhyacichthyidae  represents  the 
primitive  sister  group  of  all  gobioid  fishes.  It  is  clear  that  Rhy- 
acicithys  is  more  primitive  than  any  other  known  gobioid  (al- 
though arguably  only  marginally  more  primitive  than  some  eleo- 
tridid genera,  such  as  Micropercops),  and  at  the  same  time 
specialized.  However,  eleotridids  do  not  show  obvious  inno- 
vative specializations  in  relation  to  Rhyacichthys,  but  show  loss 
of  some  primitive  features.  Until  a  proposed  phylogeny  of  prim- 
itive genera  becomes  available,  the  eleotridids  can  only  be  re- 
garded as  a  primitive  stock,  which  gave  rise  to  one  or  more  lines 
leading  to  the  families  recognized  here.  While  most  eleotridid 
genera  may  well  have  evolved  before  the  xenisthmid-micro- 


HOESE:  GOBIOIDEI 


591 


desmid-gobiid  line  (or  lines)  evolved,  some  genera,  such  as  the 
Gobiomorphus-Leptophilypnus  group,  may  have  evolved  from 
a  common  ancestor  of  the  line  (or  lines). 

Irrespective  of  the  number  of  families,  or  subfamilies  of  go- 
bioid  fishes  recognized,  there  is  no  obvious  evidence  to  combine 
the  40  eleotridid  genera  with  any  particular  gobioid  group.  It  is 


clear  that  the  interrelationships  of  this  large  group  will  not  be 
fully  clarified  in  the  near  future. 

The  Australian  Museum,  6-8  College  Street,  Sydney  2000, 
Australia. 


Scombroidei:  Development  and  Relationships 

B.  B.  COLLETTE,  T.  POTTHOFF,  W.  J.  RICHARDS,  S.  UeVANAGI, 
J.  L.  RUSSO  AND  Y.  NlSHIKAWA 


THE  Scombroidei  is  a  suborder  of  the  Perciformes  containing 
6  families,  44  genera,  and  nearly  100  species.  All  species 
are  marine  although  at  least  one  (Scomheromonts  sinensis)  moves 
fairly  long  distances  into  fresh  water.  Most  species  are  pelagic, 
some  epipelagic  and  some  bathypelagic. 

The  first  modem  definition  of  the  scombroid  fishes  as  the 
suborder  Scombroidei  was  by  Regan  (1909).  He  clearly  sepa- 
rated the  scombroids  from  such  percoid  families  as  the  Car- 
angidae,  Rachycentridae,  Coryphaenidae,  Bramidae,  and  Men- 
idae.  Within  the  Scombroidei,  Regan  recognized  four  divisions: 
I.  Trichiuriformes  (Gempylidae  and  Trichiuridae);  II.  Scom- 
briformes  (Scombridae);  III.  Luvariformes  (Luvaridae);  and  IV. 
Xiphiiformes  (Istiophoridae,  Xiphiidae,  and  three  families 
known  only  as  fossils).  Regan's  Scombroidei  was  defined  by 
three  primary  characters:  premaxillae  beak-like,  gill  membranes 
free  from  the  isthmus,  and  epiotics  separated  by  the  supraoc- 
cipital.  To  include  Luvarus  in  the  Scombroidei,  reversals  must 
be  postulated  in  these  three  characters.  The  relationships  of 
Luvarus  lie  with  the  Acanthuroidei  (Regan,  1 902;  Leis  and  Rich- 
ards, this  volume;  Tyler  et  al.,  MS)  and  will  not  be  considered 
here.  Recent  workers  have  usually  recognized  a  suborder  Scom- 
broidei that  is  essentially  the  same  as  that  of  Regan  (1909) 
including  the  Luvaridae  (e.g..  Greenwood  et  al.,  1966)  or  have 
placed  the  billfishes  (Istiophoridae  and  Xiphiidae),  along  with 
the  Luvaridae,  in  a  separate  suborder,  the  Xiphioidei  (Gosline, 
1968;  Potthoff  et  al.,  1980),  or  have  removed  Xiphias  from  the 
group  and  placed  it  in  its  own  suborder,  Xiphioidei  (Potthoff 
and  Kelley,  1982). 

Scombroidei 

Perciform  fishes  with  epiotics  separated  by  the  supraoccipital, 
gill  membranes  free  from  the  isthmus,  premaxillae  beak-like, 
upper  jaw  nonprotrusile  (except  in  Scombrolabrax),  predorsal 
bones  lost  (except  for  a  small  one  in  Ruvettus,  Thyrsites,  and 
Tongaichlhys  and  three  well-developed  ones  in  Gasterochisma). 
second  epibranchial  extending  over  the  top  of  the  third  infra- 
pharyngobranchial  (except  in  Gasterochisma),  vertebrae  24  or 
more,  inlerorbital  commissure  of  the  supraorbital  canals  widely 
incomplete  or  absent  (Regan,  1909;  Gosline,  1968;  G.  D.  John- 
son, pers.  comm.). 

Six  families  are  recognized:  Scombrolabracidae  (monotypic; 
Potthoff  et  al.,  1980);  Gempylidae  (15  genera,  16  species;  Mat- 
subara  and  Iwai,  1952,  1958;  Russo,  1983);  Trichiuridae  (9 
genera,  about  20  species;  Parin  and  Bekker,  1972);  Xiphiidae 


(monotypic);  Istiophoridae  (3  genera,  about  1 1  species;  Naka- 
muraetal.,  1968;  Morrow  and  Harbo,  1969;  Nakamura,  1974); 
and  Scombridae  (15  genera,  49  species;  Collette,  1979,  1983). 
Fig.  3 1 2  is  a  Wagner  Tree  based  on  40  characters  considered 
significant  in  assessing  scombroid  relationships  (see  Appendix) 
generated  by  the  computer  program  (WAGNER  78)  written  by 
J.  S.  Farris  (following  Farris,  1970;  Farris  et  al.,  1970).  The  tree 
is  rooted  at  Scombrolabrax  which  is  considered  as  the  most 
primitive  scombroid  and  was  used  as  the  outgroup  for  com- 
parison with  the  other  scombroid  fishes.  Numbers  show  where 
a  character  changes  from  a  plesiomorphous  {Scombrolabrax 
condition)  to  a  derived-apomorphous  state.  The  gempylids  were 
grouped  together  on  this  cladogram  because  data  were  not  avail- 
able on  all  the  characters.  The  unresolved  areas  have  been  re- 
solved in  a  separate  study  by  Russo  (1983)  and  are  discussed 
in  the  section  on  the  Gempylidae.  The  cladogram  shows  several 
synapomorphies  of  the  billfishes  and  the  Scombridae:  pharyn- 
geal tooth  plate  stay  (character  3;  G.  D.  Johnson,  pers.  comm.), 
pair  of  small  lateral  keels  at  the  base  of  the  caudal  fin  (character 
12),  caudal  fin  rays  covering  hypural  plate  (character  14),  etc. 
Billfishes  show  many  character  reversals  and  independent  ac- 
quisitions. Within  the  Scombridae,  most  groups  seem  well-de- 
fined. 

Scombrolabracidae 

From  its  original  description  by  Roule  ( 1 922),  Scombrolabrax 
heterolepis  has  been  considered  as  related  to  gempylid  fishes 
(Grey,  1960;  Gosline,  1968;  Potthoff  et  al.,  1980).  In  most  in- 
stances wherein  Scombrolabrax  differs  from  the  gempylids  it 
differs  in  the  direction  of  the  percoids:  upper  jaw  protrusile, 
some  opercular  bones  spinous  or  serrate,  pelvic  girdle  relatively 
strong  and  attached  to  the  cleithra,  no  fusion  in  the  caudal 
skeleton,  one  fewer  vertebra  (17  +  13  =  30)  than  in  any  other 
scombroid  (except  the  billfishes)  and  procurrent  spur  present 
but  reduced  (Gosline,  1968;  Johnson,  1975;  Potthoff  etal.,  1980). 
The  stay  on  the  pharyngobranchial  of  the  fourth  gill  arch  that 
is  present  in  the  Scombridae,  Xiphiidae,  and  Istiophoridae  is 
absent  as  in  the  Gempylidae  and  Trichiuridae  (Potthoff  et  al., 
1980).  Roule  (1922)  originally  placed  Scombrolabrax  in  a  sep- 
arate suborder.  Bond  and  Uyeno  (1981)  also  recognized  a  sub- 
order Scombrolabracoidei  because  of  the  unique  specialization 
in  adults  of  the  5th  through  12th  vertebrae  which  are  expanded 
to  form  thin-walled  bullae  with  wide  ventral  openings  which 
accommodate  delicate  bubble-like  evaginations  of  the  gas  blad- 


592 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


-o 

a 
• — ^ 
o 


C/D 


< 

Q 

OS 
D 

X 
o 

5 


SCOMBRIDAE  Qo]  (g) 


01 


29113811^ 


Fig.  312.  Wagner  tree  of  scombroid  fishes  based  on  40  characters  (Appendix).  Numbers  are  inserted  where  a  character  changes  from  a 
plesiomorphous  (Scombrolabrax  condition)  to  an  apomorphous  state.  Numbers  in  darlc  circles  show  no  homoplasy,  those  in  light  circles  show 
reversals  in  character  state,  and  those  in  squares  show  independent  acquisitions. 


der.  The  presence  of  this  autapomorphy  is  not  sufficient  reason 
to  place  Scombrolabrax  in  a  monotypic  suborder.  Taxa  should 
be  grouped  based  on  shared  specializations. 

Development 

Scombrolabrax  heterolepis  (Fig.  313).  — Larval  development 
was  described  by  PottholTet  al.  (1980).  Early  larvae  from  3-4 


mm  NL  resemble  the  scombrid  Thunmis  in  pigmentation,  but 
Scombrolabrax  can  be  distinguished  from  Thunnus  in  having 
only  30  myomeres  as  opposed  to  39  myomeres  in  Thunmis. 
Larger  larvae  acquire  characteristic  melanophores  on  the  lower 
jaw  ramus  and  on  the  caudal  peduncle. 

Scombrolabrax  shares  characters  with  the  Gempylidae  and 
the  most  primitive  scombrid  tribe  Scombrini  (Scomber,  Ras- 


COLLETTE  ET  AL.:  SCOMBROIDEI 


593 


Fig.  313.     Lateral  views  of  scombroid  larvae,  (upper)  Scombrolahrax  heterolepis.  5.0  mm  SL.  from  Potthotfet  al.  ( 1 980);  (lower)  Lepidocybium 
JIavobrunneum  5.0  mm  SL,  western  Atlantic,  ATLANTIS  II,  Cr.  59,  Sta.  RHB2083,  Nov.  26,  1970,  drawn  by  J.  Javech. 


trelliger)  (Table  153).  Many  of  these  shared  characters  are  ple- 
siomorphous  and  so  are  not  useful  in  constructing  a  classifica- 
tion. 

Gempylidae 

Body  oblong  or  elongate,  compressed;  maxilla  exposed;  strong 
anterior  canine  teeth  present;  base  of  spinous  dorsal  fin  longer 
than  soft  dorsal;  three  anal  spines  except  Rexea  and  Nealotus. 
with  two  spines;  pelvic  fins  1,5  or  reduced  to  only  a  spine;  caudal 
fin  present;  vertebrae  32-58  (Tables  1  54  and  1  55);  anterior  pre- 
caudal  vertebrae  without  parapophyses,  with  sessile  ribs;  pos- 
terior precaudal  vertebrae  with  ribs  attached  at  the  extremities 
of  closed  haemal  arches  (Regan,  1909).  The  family  currently 
includes  16  species  in  15  genera  (Parin  and  Bekker,  1972;  Nak- 
amura  and  Fuji,  1983;  Russo,  1983). 

Russo  (1983)  divided  the  Gempylidae  into  six  monophyletic 
groups  (Fig.  3 1 4)  based  on  osteological  characters.  Three  groups 
are  monotypic:  Lepidocybium.  Rmettus.  and  Thyrsites.  The 
Epinnula  group  consists  of  four  genera  above  character  7:  Epin- 


nula,  Neoepinnula,  Tongaichthys,  and  Thyrsitops.  The  Nealotus 
group  is  composed  of  three  genera  above  character  2:  Nealotus, 
Promethichthys.  and  Rexea.  The  Gempylus  group  contains  five 
genera  above  character  3:  Thyrsitoides.  Nesiarchus.  Gempylus. 
Diplospinus,  and  Paradiplospinus.  Diplospinus  and  Paradiplo- 
spimts  should  probably  be  combined  under  Diplospinus. 

Development 

The  family  Gempylidae  is  characterized  by  the  following  lar- 
val and  adult  characters  when  compared  to  the  family  Scom- 
bridae:  known  gempylid  larvae  (except  Thyrsitops  with  smooth 
spines)  have  serrate  dorsal,  anal  and  pelvic  fin  spines,  scombrid 
larvae  have  smooth  spines  (Table  1 53).  Gempylids  initially  de- 
velop 3  epurals  (ontogenetic  fusion  in  Diplospimis),  scombrids 
develop  2  epurals.  Gempylids  develop  2  uroneurals  (we  were 
unable  to  confirm  this  on  all  gempylid  genera),  scombrids  de- 
velop one  uroneural.  In  gempylids  the  first  dorsal  pterygiophore 
inserts  in  the  second  intemeural  space;  in  scombrids  it  inserts 
in  the  third  space.  Most  gempylids,  except  Ruveltus  and  Neoe- 


594 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES -AHLSTROM  SYMPOSIUM 


i- 


■H5 


.1^ 


^ 

^ 


J 


^ 


^^ 


^ 

^ 

.^ 


«0 


-5f 


f             % 

=: 

o 

s? 

■*• 

-5f 

.•^ 

.5j 

N, 

V 

"*- 

^ 

<' 

^ 

=0 

^^ 

5j 

/^ 

.4) 

^"^ 

* 

-V 

CL^        i 

=0 


.'^ 


"if 


Fig.  314.     Most  parsimonious  cladogram  of  relationships  of  genera  in  the  Gempyhdae  (from  Russo.  1983;  fig.  47). 


pinnula.  have  more  precaudal  and  fewer  caudal  vertebrae;  in 
most  scombrids  the  reverse  is  tnae  (Table  1 54).  In  the  gempylids 
3  centra  support  caudal  rays;  in  scombrids  (except  Scombrini 
and  Grammatorcynus)  4  or  5  centra  support  the  caudal  rays. 

Lepidocybium  flavobrunneum  (Fig.  313).  — Lepidocybiuin  lar- 
vae, caught  in  the  Indian  Ocean,  have  been  described  by  Ni- 
shikawa  ( 1 982).  The  description  agrees  with  the  Atlantic  Ocean 
and  Gulf  of  Mexico  larvae  from  the  MCZ  collection  examined 
by  Potthoff,  except  for  the  vertebral  count  which  Nishikawa 
reported  to  be  16  +  15  =  31.  The  MCZ  specimens  had  17  + 
15  =  32  vertebrae.  Lepidocybium  larvae  and  juveniles  can  be 
distinguished  from  other  scorn  brid  and  gempylid  species  by 
meristics  (Tables  154,  155,  and  156),  pigmentation  and  shape. 
First  dorsal  fin  spine  count  is  the  lowest  for  all  gempylids  and 


scombrids.  The  first  dorsal  fin  is  intensely  pigmented  in  larvae 
of  Lepidocybium  and  the  individual  spines  have  serrations.  The 
height  and  pigmentation  of  the  fin  is  similar  in  larvae  of  Thunmis 
and  Euthynnus,  but  the  fin  spines  are  smooth  in  these  two 
scombrid  genera.  Neoepinnula  has  a  considerably  higher  first 
dorsal  fin,  also  with  serrate  fin  spines  and  also  intensely  pig- 
mented. The  low  total  vertebral  count  of  32  in  Lepidocybium 
is  similar  in  Scombrolabrax,  Scomber.  Rastrelliger  and  Thyr- 
sitops  and  the  count  is  the  same  in  Ruvellus,  Epinmila  and 
Neoepinnula.  The  small  projection  on  top  of  the  head  of  Lep- 
idocybium larvae  as  shown  in  Fig.  313  is  also  present  in  larvae 
of  Scomberomorus  and  Sarda.  Cranial  rugosities  (striations) 
observed  in  Lepidocybium  larvae  seem  to  be  unique  to  this 
genus.  A  very  stout  and  long  serrate  preopercular  spine  is  present 
in  Lepidocybium  larvae.  The  overall  intense  gut  pigmentation 


Fig.  315.  Lateral  views  of  gempylid  larvae  from  lop  to  bottom:  Neoepinnula  orientalis.  5.5  mm  NL,  Gulf  of  Mexico,  Flower  Garden  81-12, 
Sta.  379,  Nov.  8,  1981,  drawn  by  J.  Javech;  Epinnula  magistralis.  6.3  mm  NL,  modified  after  Gorbunova  (1982);  Thyrsitops  lepidopoides, 
5.5  mm  NL,  drawn  from  a  specimen  from  Sato's  (1983)  study  by  J.  Javech. 


COLLETTE  ET  AL.:  SCOMBROIDEI 


595 


596 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  316.     Lateral  view  of  the  gempylid  larva  Thyrsiles  atitn  6.0  mm  NL,  modified  after  Haigh  (1972a). 


in  Lepidocybium  is  probably  unique  among  scombrid  and  gem- 
pylid larvae.  In  all  known  scombiid  and  gempylid  larvae  the 
gut  is  intensely  pigmented  only  dorsad  with  little  lateral  and 
ventral  pigmentation. 

Lepidocybium  has  more  affinities  to  the  Gempylidae  than  to 
the  Scomberomorini.  With  the  Gempylidae  it  shares  caudal 
skeletal  characters  such  as  2  uroneurals  and  3  epurals:  scom- 
brids  have  1  uroneural  and  2  epurals.  Larval  Lepidocybium  have 
serrate  pelvic  and  first  dorsal  and  anal  fin  spines,  which  also  are 
present  in  known  gempylid  larvae,  except  Thyrsitops.  The  first 
dorsal  pterygiophore  in  Lepidocybium.  as  in  all  gempylids,  in- 
serts in  the  second  intemeural  space  in  Atlantic  specimens,  but 
in  Pacific  Lepidocybium  the  first  dorsal  pterygiophore  is  found 
in  the  third  intemeural  space;  in  all  scombrids  the  first  dorsal 
pterygiophore  inserts  in  the  3rd  space. 

Ruvettus pretiosus.—T\\t  larvae  of  Ruvellus are  not  known.  This 
lack  of  knowledge  is  surprising,  because  Ruvettus  is  caught  as 
by-catch  in  the  tuna  longline  fishery  (Nakamura,  1977).  The 
smallest  Ruvettus  known  to  us  is  209  mm  SL  and  has  the  features 
of  adults. 

Epinnula  magistralis  (Fig.  315).— The  larvae  of  Epinnula  are 
not  well  known.  Gorbunova  (1982)  reported  the  capture  of  3 
larvae  from  the  Gulf  of  Mexico  and  one  larva  from  the  Straits 
of  Yucatan.  In  larval  Epinnula,  the  first  dorsal  fin  is  not  as  high 
and  not  as  intensely  pigmented  as  in  Neoepinnula  and  the  first 
dorsal  fin  is  inserted  more  anteriorly  in  Neoepinnula  than  in 
Epinnula.  In  Epinnula  the  preopercular  spine  is  shorter  than  in 
Neoepinnula.  We  believe  that  the  17.8  mm  specimen  figured  in 
Belyanina  (1982b)  is  a  specimen  of  Epinnula  not  Neoepinnula 
because  of  the  more  posterior  first  dorsal  fin  insertion. 

Neoepinnula  orientalis  (Fig.  3\ 5).  — Neoepinnula  larvae  have 
been  described  by  Nishikawa  and  Nakamura  (1978)  and  one 
7.3  mm  specimen  was  figured  by  Gorbunova  (1982).  Belyanina's 
(1982b)  figure  of  a  17.8  mm  Neoepinnula  probably  is  an  Epin- 


nula as  mentioned  above.  The  larvae  of  Neoepinnula  are  very 
distinctive.  They  have  a  very  high  and  intensely  pigmented  first 
dorsal  fin  which  inserts  anteriorly  almost  on  top  of  the  head. 
This  causes  the  anteriormost  first  dorsal  pterygiophores  to  insert 
slanted  in  a  posterior  direction;  no  other  gempylid  or  scombrid 
larva  has  such  a  first  dorsal  fin. 

Tongaichthys  robustus. —Tht  larvae  of  this  recently  described 
genus  and  species  are  unknown  (Nakamura  and  Fuji,  1983). 

Thyrsitops  lepidopoides  (Fig.  315).— The  larvae  of  Thyrsitops 
were  recently  described  by  Sato  ( 1 983).  These  are  the  only  known 
gempylid  larvae  which  lack  serrations  on  the  fin  spines.  We 
believe  that  the  count  of  XVI-XXII  first  dorsal  fin  spines  for 
Thyrsitops  given  in  Parin  and  Bekker's  (1972)  Table  4  is  a 
misprint  and  should  be  XVI-XVII. 

Thyrsites  atun  (Fig.  316).  — Haigh  (1972a)  described  the  larvae 
of  Thyrsites  captured  in  plankton  tows.  Pigmentation  is  dis- 
tinctive with  2  to  3  dark  pigment  blotches  on  the  ventral  tail 
margin  unlike  any  other  known  gempylid,  but  similar  to  the 
trichiurid  Benthodesmus.  Haigh  (1972a)  gave  counts  for  Thyr- 
sites: XVIII-XXI  first  dorsal  fin  spines  and  34-35  vertebrae. 
Grey  (1953)  gave  XX  first  dorsal  fin  spines  and  37  vertebrae 
and  Parin  and  Bekker  (1972)  gave  XX-XXI  first  dorsal  fin 
spines. 

De  Jager  (1955)  fertilized  the  eggs  of  a  ripe  Thyrsites  female 
with  sperm  from  a  male  in  the  laboratory.  The  eggs  hatched 
and  the  larvae  were  fed  drops  of  human  blood.  After  9  days 
they  died,  visibly  undernourished.  De  Jager  illustrated  the  de- 
velopment of  the  eggs  and  very  early  stages  of  the  larvae.  The 
illustrations  are  not  helpful  for  identification  of  wild  caught 
gempylid  larvae  because  of  starvation  and  underdevelopment. 

The  larvae  figured  by  Regan  ( 1 9 1 6)  as  Thyrsites  are  probably 
Promethichthys  or  Rexea  because  the  first  dorsal  fin  in  the  figure 
of  the  largest  specimen  shows  XVIII  spines  and  no  pelvic  fin 
rays.  Regan  stated  in  the  text  that  total  vertebrae  were  35. 


Fig.  317.  Lateral  views  of  gempylid  larvae  from  top  to  bottom:  Promethichthys  promelheus.  8.5  mm  SL,  modified  after  Gorbunova  (1982); 
Rexea  solandri.  21.7  mm  SL,  Indian  Ocean,  DANA,  Cr.  3915II1,  from  a  cleared  and  stained  specimen  drawn  by  J.  Javech;  and  Nealotus  tripes, 
9.0  mm  SL,  modified  after  Strasburg  (1964). 


COLLETTE  ET  AL.:  SCOMBROIDEl 


597 


598 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


COLLETTE  ET  AL.:  SCOMBROIDEI 


599 


Table  153.    Comparison  of  Characters  among  Larvae  and  Juve- 
niles OF  Scombrolabrax,  of  known  Gempylidae  and  of  the  Scombrid 
Tribe  Scombrini. 


Scombrolabrax 


Gempylidae 


Tnbe  Scombrini 


First  dorsal  fin  de- 

after 

before 

after 

velops  before  or 

after  second 

First  dorsal  fin  pte- 

3 

2 

3 

rygiophore  inserts 

in  intemeural 

space  number 

First  dorsal,  anal 

smooth 

serrated,  ex- 

smooth 

and  pelvic  fin 

cept  smooth 

spines  of  larvae 

in  Thyrsi- 

serrated  or 

tops 

smooth 

Number  of  epurals 

3 

3 

2 

Number  of  uro- 

2 

2 

1 

neurals 

Procurrent  spur  of 

present 

absent  or  pres- 

absent 

caudal  fin  (John- 

ent reduced 

son,  1975) 

Hypural  fusion 

absent 

present 

present 

Dorsal  and  anal 

not  bifur- 

bifurcated 

bifurcated 

stay  of  posterior- 

cated 

most  pterygio- 

phore  bifurcated 

posteriorly  or  not 

Rexea  solandri  (Fig.  3\1).—Rexea  larvae  are  poorly  known. 
Parin  and  Bekker  (1972)  reported  a  10.5  mm  larva  but  did  not 
describe  it.  Six  cleared  and  stained  Rexea  (21.7-28.9  mm  SL) 
from  the  "Dana"  collections  were  identified  by  PotthofT  from 
meristics,  but  pigment  was  lost  due  to  the  age  of  specimens 
which  had  been  collected  in  the  1920s.  In  general,  Rexea  larvae 
look  similar  to  Promethichthys  larvae.  The  pelvic  spine  is  short- 
er and  does  not  reach  the  anus  in  Rexea,  whereas  in  Prometh- 
ichthys the  spine  reaches  past  the  anus.  Also,  Promethichthys 
has  3  elements  associated  with  the  first  anal  pterygiophore:  Rex- 
ea has  only  2.  All  our  Rexea  larvae  had  one  long  serrate  spine 
and  a  miniscule  vestige  of  a  ray  in  each  pelvic  fin.  The  larvae 
of  Rexea  promethoides  are  not  known.  Adult  R.  promelhoides 
have  a  fully  developed  pelvic  fin  with  a  count  of  1,5  rays. 

Promethichthys  prometheus  (Fig.  317).  — The  larvae  and  young 
of  Promethichthys  are  poorly  known.  Giinther  (1889)  described 
and  figured  two  larvae  5  mm  and  10  mm  as  Thyrsites  prome- 
theus (=Promethichthys  prometheus).  These  larvae  are  not  Pro- 
methichthys. The  smaller  one  cannot  be  positively  identified  but 
could  be  a  serranid  larva  because  of  body  shape  and  number  of 
myomeres.  The  larger  specimen  is  definitively  Diplospinus  mul- 
tistriatus.  Roule  and  Angel  (1930)  described  and  figured  two  P. 
prometheus  larvae  6  mm  and  10  mm.  From  their  description 
and  figures  it  is  impossible  to  confirm  their  identification.  We 


APHANOPODIN/e      DipLymus 


Fig.   319.     Relationships  of  the  Trichiuridae  and   the  gempylid 
subfamily  Gempylinae  (from  Tucker,  1956:  fig.  23). 


do  not  think  they  are  Promethichthys  because  the  pelvic  rays 
are  too  long  and  well  developed,  and  there  are  2 1  first  dorsal 
fin  spines  on  the  larger  specimen.  The  larger  specimen  may  be 
Nesiarchus.  Gorbunova  (1982)  described  and  figured  two  P 
prometheus  larvae  3.9  mm  and  8.5  mm  from  the  northern  Ca- 
ribbean Sea.  The  smaller  specimen  has  only  dorsal  gut  pigment, 
a  high  first  dorsal  fin  and  long  pelvic  spines  reaching  past  the 
anus.  The  larger  larva  has  a  very  high,  moderately  pigmented 
first  dorsal  fin  and  a  very  long  pelvic  spine  extending  to  the 
anterior  portion  of  the  anal  fin  and  a  distinct  pigment  patch 
near  the  hypurals.  In  body  shape  Promethichthys  larvae  resem- 
ble those  oi  Re.xea.  The  first  dorsal  fin  spine  count  in  Parin  and 
Bekker  ( 1972)  for  Promethichthys  is  a  printing  error.  Total  ver- 
tebral counts  for  this  genus  given  by  Grey  (1953)  and  Matsubara 
and  Iwai  (1958)  are  33  and  34  respectively,  but  ours  were  34 
or  35.  This  difference  between  our  counts  and  Grey's  and  Mat- 
subara and  Iwai's  is  probably  one  of  methodology.  We  counted 
the  urostyle  as  the  last  vertebra. 

Nealotus  tripes  (Fig.  3 1 7). —The  larvae  of  Nealotus  are  not  well 
known.  Liitken  (1880)  figured  two  advanced  Nealotus  larvae, 
but  we  are  not  certain  of  his  identification.  Strasburg  (1964) 
described  a  size  series  of  Nealotus  from  9  mm-41  mm.  The  9 
mm  specimen  had  fully  formed  fins  and  probably  had  attained 
some  juvenile  pigmentation.  Nealotus  has  a  very  long  posterior 
process  in  the  first  anal  pterygiophore  which  is  evident  in  larvae 
as  small  as  about  8  mm.  This  is  an  excellent  character  to  separate 
Nealotus  from  Nesiarchus.  Nealotus  and  Nesiarchus  can  be  dis- 
tinguished by  their  pelvic  fin  ray  count  and  by  the  number  of 
middle  radials  in  the  second  dorsal  and  anal  fins.  In  juvenile 
Nealotus  the  middle  of  the  three  anal  spines  fuses  lengthwise  to 
the  posterior  process  of  the  first  anal  pterygiophore.  Thus,  in 
adult  Nealotus  only  2  anal  spines  are  visible.  One  of  us  (PotthofT) 
obtained  many  vertebral  counts  from  post-larval  stages  and 


Fig.  318.  Lateral  views  of  gempylid  larvae  from  top  to  bottom:  Nesiarchus  nasutus.  7.5  mm  SL.  Gulf  Stream  off  Miami,  Virginia  Key,  Cr. 
Fl,  June  30,  1982,  drawn  by  J.  Javech;  Gempylus  serpens.  5.6  mm  NL,  Gulf  of  Mexico,  OREGON  II,  Cr.  117,  Sta.  34521,  May  22,  1981,  drawn 
by  J.  Javech;  and  Diplospinus  multistriatus,  7.1  mm  NL,  no  data,  drawn  by  J.  Javech. 


600 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  154.    Number  of  Total  First  Arch  Gillrakers  and  Number  of  Vertebrae  in  the  Genera  of  Scombroidei. 


Family  and  genus 


Number  of 
species 


Total  number  of  gillrakers 
on  first  arch 


Number  of  vertebrae 


Scombrolabracidae 

Scombrolabrax 

Gempylidae 

Lepidocybium 

Ruveltus 

Epinnula 

Neoepinnula 

Thyrsilops 

Thyrsites 

Rexea 

2 

Promethichlhys 

Nealolus 

Thyrsitoides 

Nesiarchus 

Tongaichlhys 

Gempylus 

Diptospinus 

Paradiptospinus 

Trichiuridae 

Aphanopus 

Benlhodesmus 

8 

Lepidopus 

2 

Emxyinetopon 

Assurger 

Eupleurogrammus 

2 

Tentortceps 

Trichiurus 

Lepturacanthus 

Xiphiidae 

Xiphias 

1 

Istiophoridae 

Tctrapturus 

5 

Makaira 

3 

Islwphorus 

1-2 

Scombridae 

Gasterochisma 

1 

Scomber 

3 

Rastrelliger 

3 

Grammatorcynus 

2 

Scomberomoms 

18 

Acanthocybium 

1 

Orcynopsis 

1 

Cybiosarda 

1 

Gymnosarda 

1 

Sarda 

4 

Allothunnus 

1 

Auxis 

2 

Euthynnus 

3 

Katsuwonus 

1 

Thunnus 

7 

4-5 


1.(6-8)* 


1,(2-4)*** 
4,(13,  14)*** 


1.(4-9)*** 
1.(10)*** 

1.(0-5)*** 

1.(0)*** 

1.(5,6)*** 


28 
Present 


10-16 
15-18 

10-22 


Absent,  small  teeth  found 
on  epibranchial  in  one 
juvenile. 

Rakers  modified  to  nu- 
merous toothpatches  in 
Istiophoridae. 

0 
29-51 
33-66 
14-24 

1-27 

0 

12-17 

12-15 

11-14 

8-27 
72-80 
36-47 
29-47 
51-63 
19-43 


13 


15,  16 


17 


10,  11 


12 

12 

11 

13 

12 

12 

20,21 

23 

13,  14 

17,  18 

13 

18 

12-14 

17-19 

16-32 

21-36 

30-32 

31-33 

17,  18 

19-21 

22-24 

23-26 

19 

19 

22-28 

20-27 

20 

19 

20 

19 

20 

17-19 

20 

21 

18,  19 

20,21 

30 


16 

15 

31* 

17 

15 

32** 

16 

16 

32** 

15 

17 

32** 

16 

16 

32** 

17 

16 

32** 

20 

15 

35* 

22 

15 

37** 

19,20 

14,  15 

34** 

18 

16 

34* 

20,21 

14 

34,  35** 

20-22 

14-17 

36-39** 

19,20 

14,  15 

34** 

20-23 

13-15 

33-37** 

17 

16 

33 

26 

23 

49* 

24-26 

23-25 

48-50**  Pacific 

26-29 

24-26 

51-55**  Atlantic 

22-28 

30-37 

58-61** 

33,34 

34 

67** 

42-44 

55-56 

98,99 

38-53 

65-103 

123-156 

41 

70-73 

111-113 

39,40 

63-65 

103-104** 

43 

86 

129 

32-41 

125-151 

157-192 

39,40 

123-128 

162-168 

32-35 

124-130 

159-162 

26 


24 
24 
24 

43,44 

31 

31 

31 
41-56 
62-64 
37-39 
47,48 

38 
43-55 

39 

39 
37-39 

41 

39 


•  From  Matsubara  and  Iwai  (1958). 
••  Vertebral  counts  mostly  from  PotihofTs  unpublished  data, 
*••  Numbers  m  parentheses  are  large  spmes  emerging  from  toothpatches.  The  number  one  outside  parentheses  represents  a  large  raker  in  the  epi-ceratobranchial  angle.  Numerous  toothpatches 
are  present.  Dunng  ontogeny  gillrakers  transform  to  toothpatches. 


COLLETTE  ET  AL.:  SCOMBROIDEI 


601 


Table  155.    Numbers  of  Spines  and  Rays  in  All  Fins  of  Scombroid  Genera.  Numbers  in  parentheses  denote  vestigial  rays  and  were  counted 

on  cleared  and  stained  specimens  only. 


Number  of  fin  spines 

ind  rays 

Firsl 

Second 

Dorsal 

Anal 

Caudal 

Dorsal 

Ventral 

Family  and  genus 

dorsal 

dorsal 

tinlels 

Anal 

tinlels 

Pccloral 

Pelvic 

secondary 

Principal 

secondary 

Total 

Scombrolabracidae 

Scombrolahra.x 

12 

11,  15-16 

0 

III,   16-17 

0 

18-19 

I,  5 

8-9 

9  +  8 

9-10 

34-36 

Gempylidae 

Lepidocybium 

8-12 

16-18 

4-6 

II.   10-14 

4-5 

15-17 

I,  5 

10 

9  +  8 

10 

37 

Rmettus 

13-15 

15-18 

2-3 

III.  12-16 

2-3 

14-15 

1.5 

10 

9  +  8 

9-10 

36-37 

Epinnula 

15-16 

I.  16-19 

0 

HI.  13-16 

0 

15 

I,  5 

10 

9  +  8 

10 

37 

Neoepinnula 

16 

I.  16-20 

0 

III.  17-20 

0 

13-16 

1,5 

9-10 

9  +  8 

9-10 

35-37 

ThyrsUops 

16-18 

I.  14-16 

4-5 

III,  14-15 

5 

14-16 

I,  5 

8-9 

9  +  8 

8-9 

33-35 

Thyrsiles 

20-21 

II,  10-11 

5-7 

II,  11-12 

6-7 

14-15 

I,  5 

Rexea 

16-19 

-II,  13-16 

2-3 

I,  12-16 

2-3 

12-15 

0-1,(1)71,2-3 

8-10 

9  +  8 

8-9 

33-36 

Promelhichlhys 

17-19 

I,  17-21 

2 

l-II.  15-17 

2 

14 

1(1-2) 

lO-U 

9  +  8 

10 

37-38 

Neatotus 

19-21 

I,  16-19 

2 

II.  15-19 

2 

12-14 

1(1-2) 

8 

9  +  8 

9 

34 

Thyrsitoides 

18-19 

11,8 

8 

III.  8 

1 

13-15 

I,  5 

9 

9  +  8 

9 

35 

Nesiarchus 

19-22 

I,  19-22 

2-3 

II 

-III.  15-17 

2-3 

13 

I,  5 

8-9 

9  +  8 

8-9 

33-35 

Gempylus 

26-32 

-11,  10-12 

5-7 

II 

+  1,  10-12 

5-7 

12-15 

1,3-4 

8-10 

9  +  8 

9-10 

34-37 

Diplospinus 

30-36 

1,35-41 

0 

II,  29-35 

0 

14 

I 

4 

9  +  8 

5 

26 

Paradiphspinus 

36-44 

28-33 

0 

II,  25-30 

0 

13-14 

I 

Tongaichlhys 

12 

I,  14 

6 

III,  14 

5 

17-18 

I,  5 

9 

9  +  8 

11 

37 

Trichiuridae 

Aphanopus 

38-41 

II,  53-57 

0 

II,  44-50 

0 

12 

I.  1  (juv.) 

Present 

Benthodesmus 

32-46 

70-109 

0 

11,65-101 

0 

12-13 

1,  1 

5 

9  +  8 

5 

27 

Lepidopus 

9 

90-97 

0 

11,61-64 

0 

1(1) 

Present 

Evoxymetopon 

10 

77-86 

0 

11,56 

0 

11-12 

1.(1-3) 

7 

9  +  8 

6-7 

30-31 

Assurger 

Total  122 

0 

11,80 

0 

12 

1 

Present 

17 

Eupleurogra  minus 

3 

123-147 

0 

11.  114-121 

0 

1,2 

Absent 

Tentonceps 

Total  12 

0 

0 

0 

Present 

Absent 

Trichiurus 

3 

120-140 

0 

II.  105-108 

0 

Absent 

Absent 

Lcpluracanlhus 

4 

105-134 

0 

11.72 

0 

Absent 

Absent 

Xiphiidae 

Xiphias  adults 

38-45 

4-5 

0 

(12- 

-16)  +  (3-4) 

0 

17-19 

0 

8-10 

9  +  8 

9-11 

34-38 

Xiphias 

juveniles* 

Total  44 

-49 

0 

16-19 

0 

16-19 

0 

8-10 

9  +  8 

9-11 

34-38 

Istiophoridae 

Tetrapturus 

38-55 

5-7 

0 

(11- 

-19)  -1-  (5-8) 

0 

16-22 

1,2 

Makaira  adults 

38-46 

6-8 

0 

(13- 

-18) +  (5-7) 

0 

18-23 

1,2 

13 

9  +  8 

13 

43 

Makaira 

juveniles* 

Total  5 1 

0 

25 

0 

18-23 

1,2 

13 

9  +  8 

13 

43 

Isliophorus 

adults 

37-49 

6-8 

,    0 

(8- 

-16)  +  (5-8) 

0 

17-23 

1,2 

11-12 

9  +  8 

11-12 

39-41 

Isliophorus 

juveniles 

Total  53 

-54 

0 

23-24 

0 

17-23 

1.2 

11-12 

9  +  8 

11-12 

39-41 

Scombridae 

Gasierochisma 

17-19 

9-10 

6-8 

10-13 

5-8 

19-22 

I,  5 

Scomber 

9-13 

11-12 

5 

I,  11-12 

5 

19-21 

1,5 

10-11 

9  +  8 

10-12 

37-39 

Rastretliger 

9-11 

12 

5 

I,  12 

5 

19-20 

1,5 

10 

9  +  8 

10 

37 

Grammalorcymis 

11-13 

10-12 

6-7 

11-13 

6-7 

22-26 

1,5 

Scombcromnnis 

12-22 

15-25 

6-11 

15-29 

5-12 

19-26 

I,  5 

11-13 

9  +  8 

11-13 

39-43 

Acanlhocybium 

23-27 

12-16 

7-10 

12-14 

7-10 

22-26 

1,5 

Orcynopsis 

12-14 

12-15 

7-9 

14-16 

6-8 

22-23 

1,5 

Cybiosarda 

16-18 

17-19 

8-10 

15-17 

6-7 

23-24 

1,5 

Gymnosarda 

13-15 

12-14 

6-7 

12-13 

6 

25-27 

I,  5 

Sarda 

16-23 

13-18 

7-9 

12-17 

5-8 

23-26 

1,5 

15-16 

9  +  8 

16-17 

48-50 

Allothunmis 

15-18 

12-13 

6-8 

13-14 

6-7 

24-26 

1,5 

Auxis 

10-12 

10-12 

7-9 

11-14 

7 

23-25 

I,  5 

15 

9  +  8 

16 

48 

Euthynnus 

13-17 

11-13 

8-9 

11-15 

7-8 

25-29 

I,  5 

15-16 

9  +  8 

14-16 

47-49 

Kalsuwomis 

14-16 

14-16 

7-8 

14-16 

6-8 

26-28 

I,  5 

16-17 

9  +  8 

17-18 

50-51 

Thunnus 

11-14 

12-16 

7-10 

11-16 

6-10 

30-36 

1,5 

15-17 

9  +  8 

15-17 

47-51 

•  Xiphias  and  Isliophondae  postlanae  and  juvenile  have  a  continuous  dorsal  and  anal  fin. 


602 


ONTOGENY  AND  SYSTEM ATICS  OF  FISHES- AHLSTROM  SYMPOSIUM 


Fig.  320.  Lateral  views  of  trichiurid  larvae  from  top  to  bottom;  BenOwdesmus  sp..  8.1  mm  NL.  Gulf  of  Mexico,  OREGON  11.  Cr.  1 13,  Dec. 
6,  1980,  drawn  by  J.  Javech;  Tnchiitrus  lepturus.  6.3  mm  NL  and  17.0  mm  NL,  Gulf  of  Mexico,  OREGON  H,  cruise  unknown,  Dec.  12,  1979, 
drawn  by  J.  Javech;  and  Lepidopus  caudatus.  9.0  mm  NL  and  12.0  mm  SL,  modified  from  Padoa  (1956a). 


COLLETTE  ET  AL.:  SCOMBROIDEI  603 

Table  156.    Characters  for  Gempyud  Larvae,  Juveniles  and  Adults  and  Our  Knowledge  of  Gempylid  Larvae  and  their  Occurrence. 


Species 

Elements 
associated 

with  first 
anal  pteryg- 

iophore 

Middle 
radials. 
Dorsal 
Ventral 

Dorsal  and  anal  slay, 
one  or  two  parts,  posteriorly  bifurcated 

Predorsal 
bones 

Larvae  known 
or  unknown 

Occurrence 

Lepidocybium  flavobrunneum 

3 

7 
5 

one  part  bifurcated 

0 

known 

worldwide 

Ruvcllus  prcliosus 

3 

2,3 
2,3 

1 

two  parts  bifurcated 

1 

not  known 

worldwide 

Epinnula  magistralis 

3 

two  parts 

0 

poorly  known 

worldwide 

Neoepinnuta  orienlalis 

3 

0 
0 
4 

two  parts  bifurcated 

0 

known 

worldwide 

Thyrsilops  lepidopoides 

3 

two  parts  bifurcated 

1 

known 

offcast  and  west 

4,5 

coasts  of  South 
America 

Thyrsites  atun 

— 

— 

— 

— 

known 

all  southern  oceans 
20"'-50°S 

Rexea  spp. 

2 

2,3 
2,3 

one  part  bifurcated 

0 

not  known 

Indian  and  West 
Pacific  oceans 

Promelhuhylhys  prometheus 

3 

2 
2 

two  parts  bifurcated 

0 

poorly  known 

worldwide 

Nealotus  tripes 

3* 

2 
2 

two  parts  bifurcated 

0 

poorly  known 

worldwide 

Thyrsiloides  marleyi 

- 

8 

7 

two  parts  bifurcated 

0 

not  known 

Indian  Ocean,  West 
Pacific 

Nesiarchus  nasutus 

3 

3 
3 

two  parts  bifurcated 

0 

known 

worldwide 

Tongaichthys  robuslus 

- 

6 
5 

two  parts  bifurcated 

1 

not  known 

Tonga  Ridge 

Gempylus  serpens 

3 

6,7 
6, 1 

one  part  bifurcated 

0 

known 

worldwide 

Diplospmus  mutlislnatus 

3 

1 

one  part  bifurcated 

0 

known 

worldwide 

Paradiplospinus  gracilis 

3 

1 
1 

one  part  bifurcated 

0 

poorly  known 

temperate  and 
arctic  waters  of 
Southern  Hemi- 
sphere oceans 

•  Nealotus  has  .^  fin  spine-ray  elements  associated  in  larvae  and  juveniles.  The  middle  element  gradually  fuses  to  the  long  posterior  process  of  the  first  anal  pterygiophore. 


juveniles  of  Nealotus.  The  great  variability  in  vertebral  counts 
suggests  that  more  than  one  species  exists  in  the  genus  Nealotus. 

Thyrsiloides  marleyi.  — Thyrsiloides  larvae  are  unknown.  This 
is  not  surprising  since  the  adults  are  considered  to  be  rare  (Nak- 
amura.  1980). 

Nesiarchus  nasutus  (Fig.  318).— The  larvae  of  Nesiarchus  are 
well  known.  Giinther  (1887)  described  a  33  mm  pre -juvenile 
specimen.  A  size  series  of  26  specimens  5.1-23.5  mm  NL  or 
SL  from  the  Atlantic  Ocean  was  described  by  Voss  (1954).  The 
5.1  mm  NL  specimen  shown  in  Voss  (1954:  fig.  3A)  and  iden- 
tified as  Nesiarchus  is  Gempylus  serpens  because  of  the  mid- 
lateral  pigment  stripe  and  the  large  number  of  myomeres  as 
indicated  by  the  close  spacing.  Nesiarchus  larvae  are  easily  iden- 
tified by  a  heavily  pigmented  gular  membrane.  Larvae  larger 
than  6  mm  develop  pigment  in  the  hypural  area  and  a  distinct 
pigment  stripe  from  the  tip  of  the  snout  to  the  eye.  One  of  us 


(PotthofT)  obtained  many  vertebral  counts  from  post-larval 
specimens  and  juveniles  of  Nesiarchus.  The  great  variability  in 
vertebral  counts  suggests  that  more  than  one  species  exists  in 
the  genus  Nesiarchus. 

Gempylus  serpens  (Fig.  318).— The  larvae  of  Gempylus  are 
known.  Liitken  (1880)  figured  four  post-larvae  and  juvenile  G. 
serpens.  We  believe  that  these  were  correctly  identified  because 
at  least  6  finlets  are  present  on  all  but  the  smallest  specimens. 
Voss  (1954)  described  2  series  of  Gempylus  larvae.  Her  Gem- 
pylus A  is  Diplospinus  and  Gempylus  B  is  G.  serpens.  Eight 
larvae  from  4.4  to  1 1.6  mm  were  described  and  the  5.1  mm 
specimen  in  fig.  3A  is  a  G.  serpens  not  a  Nesiarchus.  Gempylus 
serpens  larvae  can  be  distinguished  from  other  gempylid  larvae 
by  having  a  distinct  line  of  lateral  body  pigment  and  up  to  4 
rays  in  the  pelvic  fin.  The  preopercular  spines  of  Gempylus  are 
smooth,  but  the  first  dorsal  and  pelvic  fin  spines  are  serrate. 
Late  larvae  and  juvenile  Gempylus  develop  6  or  7  dorsal  and 


604 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


angustirostns       belone 


/ 

Ibidus  oudox        several  species'^/ mdica  nigricans 


Table    157. 


Comparison   of   Characters   among   Juveniles   of 

.\'/PH/4S  AND  IsTIOPHORIDAE. 


Fig.  32 1 .  Phylogenetic  relationships  within  the  family  Isliophoridae 
(from  Robins  and  de  Sylva  (I960:  fig.  5),  names  in  quotations  not 
employed  by  Robins  and  de  Sylva). 


anal  finlets  and  consequently  have  6  or  7  middle  radials.  One 
of  us  (Potthoff)  obtained  vertebral  counts  from  late  larval  and 
juvenile  specimens  from  the  Atlantic  and  Indo-Pacific  oceans. 
Total  counts  are  higher  for  the  Atlantic  and  lower  for  the  Indo- 
Pacific  with  a  definite  separation  which  may  indicate  that  there 
are  separate  species. 

Diplospinus  multisthatus  (Fig.  318).  — The  larvae  of  Diplospi- 
ntts  are  very  well  known,  but  some  earlier  researchers  described 
them  under  other  names.  The  1 0  mm  specimen  figured  as  Thyr- 
sites  prometheus  in  Gunther  (1889)  is  Diplospinus  because  of 
the  anteriorly  protruding  spines  on  the  tip  of  the  lower  jaw  and 
because  of  the  distinct  flatness  of  the  ventral  gut,  although  the 
first  dorsal  fin  spine  count  is  too  low  for  Diplospinus.  Voss  ( 1954 
and  1 957)  described  Gempylus  type  A  larvae,  which  definitively 
are  Diplospinus.  Strasburg  (1964)  and  Yevseyenko  and  Sere- 
bryakov  (1974)  correctly  identified  and  described  Diplospinus 
larvae. 

The  larvae  of  Diplospinus  (Fig.  318)  superficially  resemble 
those  o{  Gempylus.  but  the  larvae  of  Gempylus  lack  the  following 
characters  present  in  Diplospinus:  two  horizontal  spines  at  the 
lower  jaw  tip,  serrate  preopercular  spine,  absence  of  pelvic  fin 
rays,  flatness  of  ventral  gut  due  to  posterior  process  of  basip- 
terygium,  and  pigmented  gular  membrane.  Larval  Diplospinus 
lack  the  lateral  body  pigment  stripe  characteristic  oi  Gempylus. 
Diplospinus  juvenWes  lack  dorsal  and  anal  finlets  and  supporting 
middle  radials,  features  present  in  Gempylus. 

Paradiplospinus gracilis.— Thelarvae of  Paradiplospinus are  not 
well  known.  One  of  us  (Nishikawa)  has  an  unpublished  manu- 
script on  the  larval  description. 

Trichiuridae 

Body  elongate,  strongly  compressed;  maxilla  sheathed  by 
preorbital;  anterior  canine  teeth  strong;  spinous  dorsal  not  long- 
er than  soft  dorsal  (very  slightly  longer  in  occasional  specimens 
of  .Aphanopus):  two  anal  spines  immediately  posterior  to  the 
vent;  pelvic  fins  reduced  to  1,1  or  absent;  caudal  fin  greatly 


Xtphias 


Istiophondae 


Dorsal  and  anal  fin 
development,  ad- 
dition 

First  dorsal  fin  pte- 
rygiophore  inserts 
in  intcmeural 
space  number 

Dorsal  and  anal  stay 
posteriorly  bifur- 
cated or  not 

Middle  radial  pres- 
ent or  absent  for 
posteriormost  dor- 
sal and  anal  pte- 
rygiophore 

Number  of  post- 
cleithra 

Pelvic  fin  and  basip- 
terygium  present 
or  absent 

Caudal  fin  rays  sup- 
ported by  how 
many  rentra  includ- 
ing urostyle 

Number  of  autog- 
enous haemal 
spines  in  hypural 
complex 

One  pair  of  ribs  on 
centra 


from  a  center  in  an 
anterior  and  pos- 
terior direction 


not  bifurcated 


mostly  from  anterior 
in  a  posterior  di- 
rection 

1 


bifurcated 


absent 


present 


absent 


present,  fin  ray  num- 
ber reduced 


1-4  and  13-14 


1-12 


reduced  or  absent;  dorsal  spines  and  intemeurals  correspond  to 
vertebrae,  dorsal  soft  rays  correspond  to  or  are  slightly  more 
numerous  (in  .Aphanopus  and  Benihodesmus)  than  vertebrae 
(Table  155);  vertebrae  numerous,  98-99  (.iphanopus)  to  192 
(Eupleurogrammus)  (Table  1 54);  nbs  feeble,  sessile  (Regan,  1 909; 
Tucker,  1956).  The  family  contains  9  genera  and  at  least  18 
species  (Parin  and  Mikhailin.  1981).  Most  genera  have  only  one 
or  two  species;  Benihodesmus  has  at  least  8  valid  described 
species  (Parin  and  Bekker,  1972;  Parin,  1976,  1978). 

Tucker  (1956)  recognized  three  subfamilies  within  the  Trich- 
iuridae (Fig.  319);  Aphanopinae  (Aplianopus,  Benihodesmus, 
and  Diplospinus);  Lepidopinae  {Lepidopus.  .Assurger.  Tentori- 
ceps,  Evoxymeiopon,  and  Eupleurogrammus).  and  Trichiurinae 
{Trichiurus  and  Lepturacanthus).  Diplospinus  and  Paradiplo- 
spinus  have  been  transferred  from  a  primitive  position  in  the 
Trichiuridae  to  an  advanced  position  in  the  Gempylidae  by 
Russo  (1983). 

Development 

Information  on  larval  trichiunds  is  scarce.  Of  9  trichiurid 
genera  only  3  species  in  3  different  genera  have  been  described. 
The  known  trichiurid  larvae  are  characterized  by  very  long  bod- 
ies, more  than  100  myomeres,  pelvic  fins  reduced  or  absent, 
serrate  spines  in  the  first  dorsal  and  anal  fins  and  in  the  pelvic 
fin  if  present.  The  first  dorsal  fin  is  the  first  fin  to  develop.  The 


Fig.  322.     Lateral  views  of  istiophorid  larvae  from  Ueyanagi  (1963a)  rrom  top  to  bottom:  Istiophonis  platypterus.  5.1  mm  NL;  Telraplurus 
audax.  5.0  mm  NL:  T.  angiistirosths.  4.5  mm  NL;  and  Makaira  mazara.  4.4  mm  NL. 


606 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  323. 
Javech. 


Lateral  view  of  Xiphias  gladius  larva  6.1  mm  NL,  Gulf  of  Mexico,  OREGON  II,  Cr.  126,  Sta.  36784,  May  25,  1982,  drawn  by  J. 


gut  in  preflexion  larvae  is  visibly  short,  but  elongates  during 
flexion  and  post-flexion. 

Benthodesmus  (Fig.   320).— Gorbunova  (1982)  described   B. 


Evseenko  (1982b)  described  a  20.3  mm  SL  specimen.  Pigmen- 
tation in  these  larvae  is  strikingly  similar  to  Lepidopus  caudalus 
larvae  described  by  Padoa  (1956a).  However,  in  B.  elongatus 
simonyi  the  first  dorsal  fin  spine  is  not  more  elongate  than  the 


elongatus  simonyi  from  a  size  series  of  5  larvae  3.5-1 8  mm  and      other  spines  as  in  Lepidopus  and  Trichiurus  (Fig.  320).  Bely- 


D 

m 

E 

o 

n 

C 

3 

h 

CO 

0) 

o 

o 

E 

^ 
X 

t/t 

O) 

o 

u 

i/> 

y 

k. 

D 

^ 

o 

a 

y 
o 

E 

E 
E 

0) 
-Q 

E 

c 

o 

c 

o 
O 

o 

o 

o 

o 

o 

u 

to 

D 

(J 

< 

o 

o 


o 

15 

>- 

u 


D 

-a 
o 

CO 


o 

-o 


o 

c 

E 

>- 
O 


3 
C 
C 
3 


C9 


3 

c 
c 


X 
< 


C 

o 

3 

o 


3 
C 
C 
3 


SCOMBRINI  SCOMBEROMORINI 


SARDINI 


THUNNINI 


T 


GASTEROCHISMATINAE         SCOMBRINAE 


L-H 


Fig.  324.    Subfamilies,  tribes  and  genera  of  Scombridae  with  number  of  species  in  each  genus  indicated  (modified  from  Collcttc  and  Russo 
1979:  fig.  1). 


COLLETTE  ET  AL.:  SCOMBROIDEI 


607 


Fig.  325.     Cladogram  of  relationships  within  the  Scomberomorini  (from  Collette  and  Russo.  in  press). 


anina  (1982b)  repoiled  on  a  series  of  22  B.  elongatus  pacificus 
8.0-44  mm  and  on  a  B.  vityazi  3 1  mm  (Fig.  320).  Benthodesmus 
vityazi  lacks  dorsal  pigmentation  and  has  only  two  pigment 
blotches  ventrad,  and  the  pelvic  spine  is  reduced. 

Lepidopus  (Fig.  320).  — The  eggs  and  larvae  of  L.  caudal  us  were 
described  by  Padoa  (1956a).  The  larvae  are  strikingly  similar 
to  Benthodesmus  in  pigmentation;  in  Lepidopus  the  first  dorsal 
fin  spine  is  longer  than  the  following  spines.  Regan  (1916)  figured 
an  1 1  mm  larva  as  L.  caudatus.  It  is  impossible  to  determine 
from  the  drawing  and  from  the  brief  description  if,  in  fact,  it  is 
a  larva  of  L.  caudatus.  The  figured  specimen  is  alcohol  shrunk, 
body  and  trunk  pigments  are  absent,  and  the  first  dorsal  fin 
spine  is  shorter  than  the  following  spines. 

Trichtunis  lepturus  (Fig.  320).  — Delsman  (1927)  described 
Trichiurus  eggs  and  early  larval  stages  hatched  from  wild  caught 
eggs.  He  believed  that  his  descriptions  were  based  on  a  number 
of  Trichiurus  species.  Newly  hatched  and  early  Trichiurus  larvae 
have  a  dendritic  blotch  of  pigment,  usually  in  the  ventral  finfold. 
This  blotch  disappears  when  the  first  dorsal  fin  spines  begin  to 
form  anteriorly.  Cutlass-fish  larvae  were  also  described  by  Gor- 
bunova  (1982)  from  a  series  of  59  specimens  5.0-1  7.2  mm  and 
by  Tsukahara  (1961)  from  a  series  of  laboratory-reared  and  wild- 


caught  specimens.  Small  larvae  lack  pigment  on  the  ventral 
trunk  and  tail.  With  growth,  a  single  row  of  melanophores  ap- 
pears just  anterior  to  the  first  dorsal  fin  and  develops  posteriorly. 
Ventral  and  lateral  tail  pigment  is  conspicuously  absent  even  in 
larger  larvae.  Trichiurus  belongs  to  the  tail-less  trichiurids  and 
has  no  flexion  stage.  The  pelvic  fin  in  Trichiurus  is  absent. 

Istiophoridae 

Hypural  plate  mostly  covered  by  caudal  fin  rays;  caudal  fin 
supported  by  3  centra  (urostyle  and  preural  centra  2  and  3);  long 
rounded  rostrum  formed  by  united  premaxillae;  nasals  not 
forming  part  of  the  bill;  predentary  bone  present;  teeth  present; 
pectoral  fins  placed  low  on  body;  scales  present  on  body  through- 
out life;  pelvic  fins  consisting  of  one  spine  and  two  long  rays; 
vertebrae  few,  (11-12)  +  (12-13)  =  24;  neural  and  haemal  spines 
expanded  into  strong  overlapping  laminae;  ribs  sessile  (Regan, 
1909;  Gregory  and  Conrad,  1937).  Three  genera:  Tetrapturus. 
the  spearfishes  (six  species).  Makaira.  the  marlins  (three  species), 
and  Istiophorus.  the  sailfish  (one  or  two  species). 

A  diagram  of  relationships  within  the  Istiophoridae  was  pre- 
sented by  Robins  and  de  Sylva  (1960)  and  is  included  here  as 
Fig.  321.  Two  additional  species  of  Tetrapturus  have  been  val- 
idated since  then:  7'.  pfluegeri  Robins  and  de  Sylva  and  T 
georgei  Lowe.  The  former  is  most  closely  related  to  T.  angus- 


608 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  326.     Lateral  views  of  scombrid  larvae  from  top  to  bottom:  Scomber  japonicus.  5.0  mm  NL,  modified  after  Kramer  (I960);  and  Gram- 
matorcynus  bilineatus,  4.1  mm  SL,  modified  after  Nishikawa  (1979). 


tirostris  and  T.  helone.  the  latter  to  T.  albidus  (Robins,  1974). 
Nakamura  (1974)  recognized  two  species  of  blue  marlins,  the 
Atlantic  A/,  nigricans  and  the  Indo-Pacific  M.  mazura  instead 
of  only  M.  nigricans.  Morrow  and  Harbo  (1969)  considered 
Istiophorus  monotypic;  Nakamura  (1974)  recognized  the  At- 
lantic sailfish  /.  albicans  as  specifically,  or  subspecifically,  dis- 
tinct from  the  Indo-Pacific  /.  platypterus. 

Development 

Eggs.  — No  information  is  available  on  the  identification  of  is- 
tiophorid  eggs,  except  for  a  brief  account  of  eggs  identified  as 
Tetrapturus  belone  by  Sparta  (1953). 

Larvae.— Three  studies,  all  of  which  appeared  in  1974,  sum- 
marized the  identification  status  of  istiophorids  (Richards,  1974; 
Ueyanagi,  1974a,  b).  These  larvae  are  extremely  difficult  to 
identify.  Two  types  of  larvae  are  generally  recognized— those 
with  short  bills  and  those  with  long  bills.  The  short-billed  group 
is  generally  referable  to  Makaira,  the  long-billed  group  to  Is- 
tiophorus and  Tetrapturus  (Fig.  322).  Specimens  less  than  7.0 
mm  in  length  are  all  very  similar.  Other  useful  characters  include 
melanophore  distribution  on  thegularand  branch iostegal  mem- 
branes, relation  of  the  pterotic  and  preopercular  spines  with  the 
body  axis,  shape  of  the  orbit  and  position  of  the  eye. 

Meristic  factors  such  as  fin  ray  counts  and  vertebral  formula 
are  not  particularly  useful  in  distinguishing  istiophorid  species 


from  each  other  (Richards,  1974;  Tables  154and  155).  Vertebral 
counts  can  be  used  to  distinguish  Istiophorus  and  Tetrapturus 
(12  +  12  =  24)  from  Makaira  (11  +  13  =  24)  at  sizes  greater 
than  about  20  mm  (Richards,  1974).  Probably  the  most  useful 
character  is  head  morphology  (Ueyanagi,  1963a).  The  snout  is 
short  in  all  istiophorid  larvae  under  about  5  mm  in  body  length, 
but  in  larger  specimens  the  snout  lengthens  greatly  in  Istiophorus 
and  Tetrapturus.  At  lengths  greater  than  about  1 2  mm,  the 
elongate  snouts  of  Istiophorus  and  Tetrapturus  readily  distin- 
guish them  from  the  shorter-snouted  Makaira.  Thus,  in  ver- 
tebral numbers  and  relative  snout  length,  Istiophorus  and 
Tetrapturus  are  more  similar  to  each  other  than  to  Makaira, 
confirming  the  first  subdivision  in  the  family  shown  in  Fig.  321. 
For  Pacific  species,  larval  and  juvenile  stages  are  known  for 
all  species  except  juvenile  black  marlin,  M.  indica.  Makaira 
indica  larvae  have  a  characteristic  pectoral  fin  which  is  erect 
from  the  body  in  larvae  and  adults  and  presumably  juveniles, 
too.  Makaira  mazara  lai^ae  are  characterized  by  a  short  snout, 
large  eyes,  and  forward  placement  of  the  anterior  edge  of  the 
orbit.  The  characteristic  lateral  line  appears  in  juveniles  at  about 
30  mm  SL.  Tetrapturus  aiida.x  larvae  do  not  have  forward  pro- 
jecting orbits  and  the  center  of  the  eye  is  located  at  the  same 
level  as  the  tip  of  the  snout.  The  pterotic  spine  is  parallel  to  the 
body  axis,  and  the  preopercular  spine  is  inclined  sharply  down- 
ward, forming  a  large  angle  with  the  body  axis.  Melanophores 
occur  above  the  midline  of  the  gular  membrane  or  on  the  mid 


COLLETTE  ET  AL.:  SCOMBROIDEI 


609 


Fig.  327.  Lateral  views  of  scombnd  larvae  from  top  to  bottom;  Scomheromorus  cavalla.  5.0  mm  NL,  eastern  Atlantic.  ALBATROSS  IV.  Cr. 
7206,  Sta.  79,  capture  date  unknown,  drawn  by  J.  Javech;  Acanthocybium  solanderi.  7.2  mm  NL.  Gulf  of  Mexico,  OREGON  U,  Cr.  1 17,  Sta. 
34463,  drawn  by  J.  Javech;  Sarda  sarda.  6.4  mm  SL.  Atlantic.  GERONIMO.  Cr.  3.  Sta.  133.  capture  date  unknown,  drawn  by  J.  .lavech; 
and  Gyinnosarda  unicolor,  5.1  mm  NL,  modified  after  Okiyama  and  Ueyanagi  (1977). 


610 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  328.     Lateral  views  oi  Allolhunnus  fallai  larvae  (upper)  5.7  mm  NL  and  (lower)  6.9  mm  SL  modified  after  Watanabe  et  al.  (1966). 


anterior  portion  of  the  branchiostegal  membrane.  Istiophorus 
platypterus  has  an  elongate  snout,  small  eyes,  and  a  relatively 
small  head  depth.  Melanophores  appear  characteristically  on 
the  posterior  peripheral  area  of  the  gular  membrane.  However, 
there  are  some  sailfish  larvae  which  lack  gular  melanophores, 
and  these  are  thought  to  belong  to  a  different  population.  Tel- 
rapturus  angiistirosths  larvae  of  small  size  are  similar  to  M. 
mazara,  but  the  anterior  edge  of  the  orbital  rim  does  not  project 
forward  and  melanophores  occur  on  the  branchiostegal  mem- 
brane. 

In  the  Atlantic,  specific  differences  are  not  nearly  as  clear. 
Makaira  indica  larvae  have  not  been  identified  although  adults 
are  known.  Makaira  nigricans  larvae  are  indistinguishable  from 
Makaira  mazara  and  are  characterized  by  the  short  snout,  an- 
terior projection  of  the  orbital  rim,  and  a  lack  of  gular  mela- 
nophores. Tetrapturus  pjluegeri  larvae  are  very  similar  to  T. 
angustirostris  and  characteristically  possess  melanophores  on 
the  branchiostegal  membrane.  It  is  also  a  winter  spawning  species, 
whereas  the  others  are  spring  and  summer  spawners.  Tetrap- 


turus albidus  larvae  are  very  similar  to  T.  audax  in  the  profile 
of  the  head  and  possessing  melanophores  on  gular  membranes. 
Tetrapturus  georgei  larvae  are  unknown,  and  T.  he/one  have 
been  briefly  described  without  mention  of  the  presence  or  ab- 
sence of  branchiostegal  melanophores.  A  great  amount  of  time 
has  been  spent  attempting  to  separate  Atlantic  /.  platypterus 
and  T.  alhidus  with  no  success  (Richards,  1974).  Both  have 
relatively  long  snouts  and  pigmented  gular  membranes.  Vari- 
ation in  gular  and  branchiostegal  pigmentation  has  been  de- 
scribed in  Atlantic  T.  pfluegeri  (Ueyanagi,  1974b). 

The  elongate  upper  jaw,  a  characteristic  of  istiophorid  fishes, 
is  also  found  in  the  fossils  Palaeorhynchus  and  Blochius  which 
are  thought  to  be  the  ancestral  forms  (Fierstine,  1972);  hence, 
the  elongate  upper  jaw  may  have  phyletic  meaning.  When  the 
character  of  upper  jaw  length  compared  to  body  length  is  ex- 
amined during  the  larval  period,  clear  differences  were  observed 
(e.g.,  longer  upper  jaw  in  Istiophorus  and  Tetrapturus,  and  short- 
er in  Makaira).  Adult  T.  angustirostris  possess  an  especially 
short  snout  among  the  species  in  the  genus;  elongation  of  the 


Fig.  329.  Lateral  views  of  scombrid  larvae  from  top  to  bottom:  all  drawn  by  J.  Javech,  Gulf  of  Mexico,  OREGON  II.  Cr.  1 17,  Auxis^p-.  5.0 
mm  NL,  Sta.  34463;  Euthynnus  allelleralus,  6.2  mm  SL,  Sta.  34463;  Kalsuwonus  pelanus.  5.9  mm  SL,  Sta.  34448;  and  Thuimus  thynnus.  6.0 
mm  SL,  Sta.  34497. 


COLLETTE  ET  AL.:  SCOMBROIDEI 


611 


612 


ONTOGENY  AND  SYSTEM ATICS  OF  FISHES- AHLSTROM  SYMPOSIUM 


Class   intervals    of   numbers  of   coincident    characters 
state   between  pairs   of    genera 


•:,(omber.   !;  ,    i,,-lh^:.  , 

„ ,„„., 

Kuthynrtus;  Katsuuoni.  -. 

'. 

Sa  rda 

Acinlhocybiu"! 

Group   B 


Fig.  330.     Dendogram  depicting  larval  relationships  among  12genera 
of  Scombrinae  (from  Okiyama  and  Ueyanagi,  1978:  fig.  2). 


snout  is  greatest  in  the  larval  period.  In  the  case  of  T.  audax, 
however,  the  snout/body  length  value  is  not  so  high  among 
Tetraplurus,  and  it  is  thought  to  be  intermediate  between  Ma- 
kaira  and  Tetraptums.  The  pattern  of  morphological  change  in 
snout  length  with  ontogeny  in  various  genera  corresponds  to  the 
classification  of  adult  fishes  and  is  thought  to  reflect  phylogeny 
among  the  genera  of  istiophorid  fishes  (Ueyanagi,  1963b). 


Table  1 58.     Present  Status  of  the  Larval  Fish  Taxonomy  of  the 
Family  Scombridae  (Modified  after  Okiyama  and  Ueyanagi,  1978: 

TABLE  1). 


Table  159.  Presumed  Phylogenetically  Important  Larval 
(Chiefly  Advanced  Postlarval  or  Early  Juvenile  Stages)  Char- 
acters as  Coded  State  for  Comparison  of  12  Genera  of  the 
Subfamily  Scombrinae.  (After  Okiyama  and  Ueyanagi,  1978:  table  2). 


Char- 

Coded  stale 

index 

Character 

1 

2 

3 

1 

Supraoccipital 

absent 

— 

present 

spine 

2 

Head 

small;  less 
than  Vi 
ofSL 

large;  more 
than  '/,  of 
SL 

3 

Viscera  and 

compact 

— 

elongated. 

vent 

with  wide 
space 
from  anal 
fin 
rounded 

with  vent 
just  in  front 
of  anal  fin 

4 

Snout 

pointed 

elongated 

5 

PremaxiUary 
teeth 

minute 

large 

large;  some 
fang-like 

6 

Jaw 

equal  size 

equal  or  un- 
equal size 

unequal  with 
distinct  up- 
per jaw  pro- 
jection 

7 

Preopercular 
spine 

absent 

— 

present 

8 

Spiny  supraor- 
bital crest 

absent 

present  or 
absent 

present 

9 

Pterotic  spine 

absent 

— 

present 

10 

Cartilaginous 
pad  on  lower 
jaw 

absent 

present  or 
absent 

present 

11 

Dorsal  body 
pigmentation 

heavier 

— 

lighter 

12 

Post  vent  pig- 

present, ex- 

absent or  a 

absent 

mentation 

tensive 

few  dots 

13 

Myotome 
counts 

low;  30-3 1 

middle;  38- 
41 

high;  40-65 

Subfamily,  tnbe. 
and  genus 


Present  state  of  the  lai^al  fish  taxonomy 


Genenc  level 


Species  level 


Scombrinae 

Scombrini 

Scomber  well  established 

RastreUiger  well  established 

Scomberomorini 

Grammatorcynus     well  established 
Scomberomorus       well  established 


Acanthocybium 
Sardini 

Orcynopsis 
Cybiosarda 
Sarda 


Gymnosarda 
Thunnini 

Allothunnus 

AlLXlS 

Euthynnus 

Katsuwonus 

Thunnus 


well  established 

no  information 
no  information 
rather  well  estab- 
lished 

well  established 

well  established 
well  established 
well  established 
well  established 
well  established 


well  established 
well  established 


well  established 
incomplete  or  none 
for  many  species 
well  established 


no  information 

no  information 

no  problem  in  identi- 
fication but  poor 
information 

well  established 


well  established 

incomplete 

well  established 

well  established 

established  for  most, 
but  identification 
very  difficult  to  ac- 
complish 


Xiphiidae 

Hypural  plate  mostly  covered  by  caudal  fin  rays;  caudal  fin 
supported  by  only  two  centra  (urostyle  and  preural  centrum  2); 
long  depressed  rostrum  formed  only  by  united  premaxillae;  na- 
sals not  forming  part  of  the  bill;  predentary  bone  absent;  su- 
pratemporal  absent;  one  postcleithrum;  anteriormost  dorsal  pte- 
rygiophore  inserts  in  second  (rather  than  third)  intemeural  space; 
no  teeth  in  adult;  pectoral  fins  placed  low;  scales  lost  in  adult; 
pelvic  fins  and  pelvic  girdle  absent;  vertebrae  few,  15-1-  11  = 
26;  neural  and  haemal  spines  not  expanded;  ribs  present  on  only 
the  first  four  centra  and  the  last  two  precaudal  vertebrae  (Regan, 
1909;  Gregory  and  Conrad,  1937;  Potthofl'and  Kelley,  1982; 
G.  D.  Johnson,  pers.  comm.).  Monotypic,  contains  only  Xiphias 
gladius. 

Development 

Xiphias  gladius  (Fig.  323).  — Eggs  and  larvae  of  Xiphias  have 
been  described  by  a  number  of  authors  during  the  1 9th  and  20th 
centuries  as  summarized  by  Richards  (1974).  The  most  recent 
and  complete  description  is  by  Arata  (1954)  and  drawings  of  a 
developmental  series  are  by  Taning  (1955).  Osteological  devel- 
opment was  studied  by  Potthoffand  Kelley  (1982). 

Early  larvae  of  swordfish  are  distinguished  by  having  overall 
body  pigmentation  and  lacking  the  strong  pterotic  and  pre- 
opercular spines  so  characteristic  of  the  istiophorids  (Richards, 
1974).  Late  preflexion  larvae  to  juveniles  acquire  prickly  squa- 


I         3        5        7       9        II       13       15      17       19      21      23     25     27      29     31       33     35     37  39 

I     I     I     I     I     I     I     I     I     I     I     I     I     I     I     I     I     I     I     I     1     I     I     I     1     I     I     I     I     I     I     I     I     I     I     I     I     1  I 


3  7  mm  NL 


-^ 


\j  u  M  y  u  u  y 


Mtii 


4  Omm  NL 


-^^QS^ 


0  0  0  0  a 


l/VUUUUl/UUUV""" 


i\    i\   r\    (\    n    r\ 


4-4  mm  NL 


■^e:^^ 


'(?  Po 


UULI-UULXjUUui 


3        5        7        9        II        13      15       17       19      21      23     25      27     29     31       33     35     37  39 

Fig.  331.     Osteological  development  of  Thunnus  atlanlicus.  family  Scombridae. 


614 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 

2  4  6  8  10  12  14  16  18  20  22 


T r 


-| r 


~i r 


"1        n     I        I       I        I       r 


-| 1 1 1 r— 

3  0  mm  NL 


24 

— I 


4- 


4 — e- 


40mm  NL 


V/^ 


iXXXXXLXXXXx^^ 


4  3  mm  NL 


0  0  ff  I)  ff  II  ^  ffO  d 


4-5  mm  NL 


s  /)  0  a  ^  s  n  a  ii  0  o  a 


4  9  mm  NL 


-\ — r-^^ — « — ^ 


\~i 


0 0 €  ff  0 oj  0  (I  II  0  tP  /?  ^/j 


L11UUU±SXS±±jU  i  I  /  iJTl 


4  9  mm  NL 


■VTT^V"^^""^-^^^^ 


1^ 


6  6  mm  NL 


oOQooocdpOjo a  (]  ff  ff  a/j^ouoo 0  0 u  g  s  a  o  « «„„„„. 


^  ,j  /,  ,. ■  . 


J^ OL. a^^^ i^ 


tXtWTTX^ 


4  6  8  10  12  14  16  18  20  22 

Fig.  332.    Osteological  de\elopinent  of  Isliophonds  platypterus,  family  Istiophoridae. 


COLLETTE  ET  AL.:  SCOMBROIDEl 


615 


Table  160.    Developmental  Features  for  the  Scombroid  Families  and  Morone.  a  Primitive  Perctform  Fish. 


Scorn bro- 

labracidae 

Gempylidae 

{Scombro- 

(Gempvius. 

Gempylidae 

Thunnini 

Scorn  bnni 

Isnophondae 

Xiphiidae 

Percichihyidae 

lahrax) 

Sestanhiis) 

{Diplosptnus) 

( Thunnus) 

(Scomber) 

(Isuophonts) 

{Xtphia^) 

i  Morone)* 

Developing  neural 

Antenor  col- 

Entire col- 

Entire col- 

Anterior col- 

Not known. 

Entire  col- 

Entire col- 

Anterior 

and  haemal 

umn:  pos- 

umn: pos- 

umn:  pos- 

umn: pos- 

umn: pos- 

umn: pos- 

column: 

spines  and  arch- 

teriorly. 

tenorly. 

teriorly. 

teriorly. 

teriorly. 

teriorly. 

posterior- 

es and  hypural 

Center 

Hypural 

Hypural 

Center 

Hypural 

Hypural 

ly.  Center 

complex  parts 

column: 

complex: 

complex: 

column: 

complex: 

complex: 

column: 

are  added  in  a 

anteriorly 

anteriorly 

antenoriy 

anterioriy 

anteriorly 

anterioriy 

anterioriy 

direction. 

and  poste- 

and poste- 

and poste- 

and poste- 

and  poste- 

and poste- 

and poste- 

riorly.  Hy- 

riorly. 

riorly. 

rioriy.  Hy- 

riorly. 

riorly. 

norly. 

pural 

pural 

Hypural 

complex: 

complex: 

complex: 

anteriorly 

anteriorly 

antenorly 

and  poste- 

and poste- 

and  poste- 

riorly. 

riody. 

rioriy. 

Developing  pter- 

First  dorsal: 

First  dorsal: 

First  dorsal: 

First  dorsal; 

First  dorsal: 

Entire  dor- 

Entire dor- 

First dorsal: 

ygiophores  and 

antenorly 

posterior- 

posterior- 

posterior- 

very few 

sal:  very 

sal:  ante- 

anteriorly 

fin  spines  and 

and  poste- 

ly. Second 

ly.  Second 

ly.  Second 

antenorly. 

few  ante- 

riorly and 

and  poste- 

rays are  added 

riorly. 

dorsal:  an- 

dorsal: 

dorsal:  an- 

most pos- 

rioriy. 

posterior- 

riorly. 

in  a  direction. 

Second 

tenorly 

some  an- 

tenorly 

tenoHy. 

most  pos- 

ly.  Anal: 

Second 

dorsal:  an- 

and poste- 

teriorly. 

and  poste- 

Second 

tenorly. 

very  few 

dorsal: 

teriorly 

norly. 

most  pos- 

riorly. 

dorsal:  an- 

Anal: very 

antenorly. 

anteriorly 

and  poste- 

Anal: an- 

teriorly. 

Anal: 

terioriy 

few  ante- 

most pos- 

and  poste- 

riorly. 

teriorly 

Anal:  few 

some  an- 

and poste- 

norly. 

terioriy. 

riorly. 

Anal:  an- 

and poste- 

anteriorly. 

teriorly. 

riorly. 

most  pos- 

Anal: an- 

teriorly 

norly. 

most  pos- 

most pos- 

Anal: an- 

teriorly. 

teriorly 

and  poste- 

teriorly. 

teriorly. 

tenorly 

and  poste- 

riorly. 

and  poste- 
riorly. 

rioriy. 

Sequence  of  fin 

1.  Second 

1 .  First  dor- 

1.   First  dor- 

1.   First  dor- 

1. Second 

1.   First  dor- 

1. Second 

1.  Second 

and  associated 

dorsal 

sal. 

sal. 

sal. 

dorsal 

sal. 

dorsal 

dorsal 

pterygiophore 

and  anal 

2.  Second 

2.  Second 

2.  Second 

and  anal 

2.  Second 

and  anal 

and  anal 

development. 

concur- 

dorsal 

dorsal 

dorsal 

concur- 

dorsal 

concur- 

concur- 

rently. 

and  anal 

and  anal 

and  anal 

rently. 

and  anal 

rently. 

rently. 

2.  First  dor- 

concur- 

concur- 

concur- 

2. First  dor- 

concur- 

2.  First  dor- 

2. First  dor- 

sal. First 

rently. 

rently. 

rently. 

sal. 

rently. 

sal.  First 

sal.  Sepa- 

dorsal 

First  dor- 

First dor- 

First dor- 

First dor- 

dorsal 

ration  or 

separated 

sal  sepa- 

sal only 

sal  sepa- 

sal nol 

and  anal 

continui- 

from sec- 

rated 

briefly 

rated 

separated 

nol  sepa- 

ty of  first 

ond  dor- 

from sec- 

separated 

from  sec- 

from sec- 

rated 

and  sec- 

sal during 

ond  dor- 

from sec- 

ond dor- 

ond dor- 

during 

ond  dor- 

part of 

sal  during 

ond  dor- 

sal during 

sal  dur- 

develop- 

sals not 

develop- 

part of 

sal  dunng 

part  of 

ing 

ment. 

given. 

ment. 

develop- 
ment. 

develop- 
ment. 

develop- 
ment. 

develop- 
ment. 

First  anteriormost 

Dorsal  from 

Dorsal  from 

Dorsal  from 

Dorsal  from 

Dorsal  from 

Dorsal  and 

Variable, 

Dorsal  and 

dorsal  and  anal 

one  piece. 

one  piece. 

one  piece. 

one  piece. 

one  piece. 

anal  from 

dorsal  and 

anal  de- 

pterygiophore 

anal  from 

anal  from 

anal  from 

anal  from 

and  anal 

two 

anal  may 

velop 

develop  from 

two 

two 

two 

two 

from  one 

pieces. 

develop 

from  two 

one  or  two 

pieces. 

pieces. 

pieces. 

pieces. 

piece. 

from  one 

pieces. 

pieces  of  carti- 

or two 

lage. 

pieces. 

Centra  develop 

Yes 

Yes 

Yes 

Yes 

Yes 

No 

No 

No 

from  saddle- 

shaped  ossifica- 

tions at  bases  of 

neural  and  hae- 

mal spines. 

•  Data  from  Fntzsche  and  Johnson  (1980) 


mation  (Arata,  1954;  Yabe  et  al.,  1959;  Potthoff  and  Kelley, 
1982),  and  the  scales  are  supposedly  lost  in  adults  (Palko  et  al., 
1981).  One  of  us  (Potthoff)  has  found  that  scales  are  retained 
in  adult  swordfish,  at  least  on  some  parts  of  the  body.  Larval 
and  juvenile  swordfish  differ  from  istiophoinds  in  a  number  of 
characters  (Table  157). 


Adult  Xiphias  have  two  dorsal  and  two  anal  fins  but  larvae 
and  juveniles  have  single  continuous  dorsal  and  anal  fins  (Yabe 
et  al.,  1959:  fig.  9).  During  development,  the  fin  rays  in  the 
center  of  the  fins  stop  growing  and  the  rays  become  subcuta- 
neous. The  subcutaneous  rays  and  their  pterygiophores  are  pres- 
ent in  the  adults  (Potthoff  and  Kelley,  1982).  In  three  scombrid 


616  ONTOGENY  AND  SYSTEMATICS  OF  FISHES- AHLSTROM  SYMPOSIUM 

Table  161.    Osteological  Features  and  Counts  for  the  Scombroid  Families  and  Morone.  a  Primitive  Perciform 


Fish. 


Scombro-  Gcmpylidae 

labracidae  {detnpylits. 

{Scombrolahrax)  Ncsiarchits) 


Thunnini  Scombnni  Istiophondae  Xiphiidae 

iThimnus)  {Scomber)  {hliophonis)  (Xiphtas) 


Percichthyidae 
i.\forone)* 


Predorsal  bones: 

Present  or  absent      absent 


Number 


0 


usually  ab- 

sent# 
Oor  1 


First  anteriormost  dorsal  pterygiophore; 
Supports  how  2  2 

many  spines 
Inserts  in  inter-         3  2 

neural  space 

number 

First  anteriormost  anal  pterygiophore: 
Supports  how  3  2  or  3t 

many  spines 

Middle  radials: 

Present  or  absent      present  present 


Dorsal  and  anal  stay: 
Present  or  ab- 


sent, ossifies  to         part 
one  or  two 
parts,  poste- 
riorly 
Bifurcated  or  non-bifur- 

non-bifurcated  cated 


present:  one       present:  usu- 


ally 2 
parts® 


bifurcated 


absent 
0 

2 
2 

2  or  3*** 


present  or 
absent 


present:  one 
part 


bifurcated 


absent 
0 

2 
3 


absent 
0 

2 
3 


absent 
0 

3 
1 


present 


present 


present 


absent 
0 

1  to  3, 
mostly  2 

2 


1  to  3, 
mostly  2 

absent 


present 
3 

3 
3 


present:  one      present:  one      present:  one      present:  one 
part  part  part  part 


bifurcated 


bifurcated 


bifurcated 


non-bifur- 
cated 


present 


present:  one 
part 


non-bifurcat- 
ed 


Pelvic  fin: 
Spine,  ray  count 

1,5 

I,  5;  I,  4;  I, 
2;  I,  1:1 

1,  3;  I,  2;  1, 
1;  I 

1,5 

1,5 

1,2 

0 

1,5 

Preural  centrum  3: 

Neural  spine 
with  or  with- 
out cartilage 
tip 

Haemal  spine  au- 
togenous or 
nonautogenous 

with 
autogenous 

with 
autogenous 

with 
autogenous 

with 
autogenous 

with 
autogenous 

with 

autogenous 

without 

non-autog- 
enous 

with 
autogenous 

Epurals: 

Number 

Anterior  epural 
fused  with 
neural  arch  of 
PU, 

3 

No 

3 
No 

1  to  3" 
No 

2 
Yes 

2 

3 
No 

3 
No 

3 

No 

Uroneural: 

1  or  2 

2 

2 

1 

1 

1 

1 

1 

2 

Hypural  5: 

Present  or  absent 
Fused  or  separate 

present 
separate 

present 
separate 

absent 
fused  early 

present 
separate 

present 
separate 

absent 

present 
separate 

present 
separate 

Ontogenetic  hypural 
fusion: 

Present  or  absent 

absent 

present 

present 

present 

present 

present 

present 

absent 

Procurrent  spur  (Joh 
Present  or  absent 

nson,  1975): 
present 

present,  re- 
duced or 
absent 

absent 

absent 

absent 

absent 

absent 

present 

Vertebrae  inclusive  of  urostyle  supporting  caudal 
Number                     3                         3 

rays: 
3 

4 

3 

3 

2 

3 

COLLETTE  ET  AL.:  SCOMBROIDEI 


617 


Table  161.    Continued. 


Scombro- 

labracidae 

(Scomhrolahro-X) 

Gempylidac 
IGempylus. 
Nestanhns) 

Tnchiundae 

Thunnini 
( Thitnnus) 

Scombnni 
{Scomber) 

Isliophondae 
(Isriopfiorus) 

Xiphiidae 
(Xiphias} 

Percichlhyidae 
(Morone)* 

Number  of  vertebrae: 

Precau-                       13+  17  = 

usually 

usually  few- 

usually few- 

13. 14  +  17, 

12+12  = 

15  +  11  = 

12+  13  =  25 

da!  +  caudal  =          30 

more  pre- 

er  precau- 

er  precau- 

18  =  31 

24 

26 

11  +  14  =  25 

total 

caudal, 

dal,  more 

dal,  more 

11  +  13  = 

16+10  = 

fewer 

caudal. 

caudal. 

24 

26 

caudal.  3 1 

58-192 

39-41 

to  53 

Stay  on  4lh  phar>ngobranchial  (G. 

D.  Johnson,  pers 

comm.): 

Present  or  absent      absent 

absent 

absent 

present 

present 

present 

present 

absent 

"  Data  from  Fnlzsche  and  Johnson  ( 1 980)  and  G.  D.  Johnson,  pers.  comm. 

#  Rtivelltis.  Thyrmops  and  Tongaichthys  have  one  predorsal  bone. 

t  Rexea  has  2  spines,  Nealolus  ontogenelically  has  3  spmcs  but  center  spine  fuses  to  basipterygium  dunng  development. 

Si  Leptdocybiuw.  Rexea  and  Gempyltts  have  a  one-pan  stay. 

••  Diplospinus  ontogenelically  has  3  epurals;  poslenor  2  epurals  are  fused  to  one  in  adults,  some  Diplospinus  develop  only  2  epurals. 
*•*  Trichiunts  has  only  2  spines  on  the  first  anal  pterygiophore. 


genera.  Scomber,  Rastrelliger.  and  Auxis,  there  is  a  separation 
between  the  first  and  second  dorsal  fins  similar  to  that  in  adult 
Xiphias,  except  that  in  these  scombrids  the  two  fins  are  separate 
initially  even  though  the  first  and  second  dorsal  fin  pterygio- 
phores  are  continuous  (Kramer.  1960), 

Scombridae 

Hypural  plate  mostly  covered  by  caudal  fin  rays;  caudal  fin 
rays  supported  by  3  centra  (Scombrini  and  Grammatorcynus) 
or  4  to  5  centra  (all  other  tribes)  (urostyle  and  preural  centra 
2  to  4);  premaxillae  beak-like,  free  from  the  nasals  which  are 
separated  by  the  ethmoid;  no  canine  teeth;  pectoral  fins  placed 
high  on  the  body,  with  19-36  rays;  pelvic  fins  1.5;  vertebrae  31- 
64  (Tables  154  and  155);  5-12  finlets  follow  the  second  dorsal 
and  anal  fins.  The  family  contains  1 5  genera  and  49  species 
(Collette,  1979,  1983). 

The  family  Scombridae  can  be  divided  into  two  subfamilies 
(Fig.  324);  the  Gasterochismatinae,  which  contains  only  the 
distinctive  Gasterochisma  nielampus,  and  the  Scombrinae.  There 
are  problems  with  the  placement  of  Gasterochisma.  To  be  in- 
cluded in  the  Scombridae,  it  must  have  lost  the  extension  of 
the  cartilaginous  tip  of  the  second  epibranchial  that  extends  over 
the  top  of  the  third  infrapharyngobranchial  (Fig,  312,  character 
1;  G,  D,  Johnson,  pers,  comm.),  regain  well  developed  predorsal 
bones  (character  2),  and  lose  the  pharyngeal  tooth  plate  stay 
characteristic  of  all  other  higher  scombrids  except  Gasterochis- 
ma and  Grammatorcynus  (G.  D,  Johnson,  pers.  comm.).  How- 
ever. Gasterochisma  agrees  with  the  billfishes  and  other  scom- 
brids in  several  caudal  skeletal  characters  (Fig.  312.  characters 
8.  12,  14). 

The  Scombrinae  is  composed  of  two  groups  of  tribes.  The 
primitive  mackerels  (Scombnni— Scomber  and  Rastrelliger)  and 
Spanish  mackerels  (Scomberomorini  — 5cow/)<'roworj<.s,  Acan- 
thocyhium,  and  Grammatorcynus)  have  a  distinct  notch  in  the 
hypural  plate,  lack  any  bony  support  for  the  fleshy  keels  on  the 
caudal  peduncle,  and  do  not  have  preural  centra  two  and  three 
greatly  shortened.  The  more  advanced  bonitos  (Sardini)  and 
tunas  (Thunnini)  form  a  monophyletic  group  showing:  loss  of 
the  notch  between  the  fused  lower  and  fused  upper  hypural 
bones  (Fig.  312,  character  33),  bony  support  for  the  medial 
caudal  peduncle  keel  (character  1 6),  anterior  corselet  of  enlarged 


scales  (character  22)  and  have  preural  centra  two  and  three 
greatly  shortened.  The  Scomberomorini.  like  the  two  more  ad- 
vanced tribes,  have  a  median  fleshy  keel  on  the  caudal  peduncle 
between  the  pair  of  small  keels  (character  1 1).  However,  there 
is  no  bony  support  for  this  keel  as  there  is  in  the  bonitos  and 
tunas.  Grammatorcynus  shares  this  character  state  with  the  oth- 
er two  genera  of  Scomberomorini  but  has  only  three  centra 
supporting  the  caudal  fin  (reversal  at  character  8),  as  in  the 
Scombrini,  rather  than  four  or  five  as  in  the  Scomberomorini, 
Sardini  and  Thunnini.  The  Sardini  (Orcynopsis,  Cybtosarda, 
Sarda,  Gymnosarda,  and  Allothunnus:  Collette  and  Chao.  1 975) 
differ  from  the  Thunnini  (Auxis.  Euthynnus,  Katsuwonus.  and 
Thunnus)  in  lacking  any  trace  of  the  subcutaneous  vascular 
system  (Fig.  312.  character  23)  that  permits  the  members  of  the 
Thunnini  to  be  warmer  than  the  water  around  them.  Instead  of 
being  considered  as  a  bonito.  Allothunnus  can  better  be  inter- 
preted as  the  most  primitive  member  of  the  Thunnini.  sharing 
the  presence  of  a  prootic  pit  on  the  skull  with  the  higher  tunas 
(character  26)  but  lacking  their  subcutaneous  vascular  system. 
Allothunnus  also  has  an  autogenous  second  epibranchial  carti- 
lage as  in  the  Thunnini  (G,  D,  Johnson,  pers.  comm.)  and  shares 
a  common  parasitic  copepod,  Elytrophora  brachyptera.  with  six 
of  the  seven  species  of  Thunnus  (Gibbs  and  Collette,  1967; 
Cressey  et  al.,  1983). 

The  Scomberomorini  is  the  most  speciose  group  within  the 
Scombroidei  and  so  merits  further  attention.  After  comparing 
the  18  species  of  Scomberomorus  with  each  other  and  with 
Acanthocybium  and  Grammatorcynus  (Collette  and  Russo,  in 
press),  characters  that  differentiated  among  species  or  genera 
were  listed.  Grammatorcynus  clearly  is  more  primitive  than 
Scomberomorus  and.  therefore,  it  was  used  as  the  outgroup  for 
comparison  with  Scomberomorus.  Character  polarities  were  de- 
termined by  considering  the  character  stale  present  in  Gram- 
matorcynus to  represent  the  plesiomorphous  condition.  Of  the 
72  characters  that  differentiated  at  least  one  taxon  from  the 
others,  14  were  autapomorphies  oi  .Acanthocybium.  These  can- 
not contribute  to  an  understanding  of  relationships  within 
Scomberomorus  and  were  omitted  from  the  analysis.  The  re- 
maining 58  characters  were  employed  to  generate  a  cladogram 
(Fig.  325)  using  a  computer  program  (WAGNER  78)  written 
by  J.  S.  Farris  (following  Farris,  1970  and  Farris  et  al.,  1970). 


618 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


The  numbers  at  the  nodes  indicate  characters  that  are  discussed 
by  Collette  and  Russo  (in  press). 

Scomberomorus  differs  from  Acanthocybium  at  character  17 
and  from  all  other  scombrids  in  possessing  a  spatulate  vomer 
that  projects  anteriorly  well  beyond  the  neurocranium.  Scom- 
beromorus differs  from  both  Acanthocybium  and  Grammator- 
cynus  in  12  osteological  characters.  In  three  more  characters, 
Scomberomorus  differs  from  both  genera  but  is  closer  to  Acan- 
thocybium. Scomberomorus  and  Acanthocybium  share  17  os- 
teological synapomorphies  at  character  18  but  differ  from 
Grammatorcynus.  There  are  six  species  groups  within  Scom- 
beromorus: sinensis,  commerson,  munroi.  semifasciatus.  gut- 
tatus.  and  regalis  (Fig.  325;  Collette  and  Russo,  in  press). 

The  young  stages  of  scombrids  are  difficult  to  identify  to  genus 
and  particularly  to  the  species  level  (Richards  and  Potthoff, 
1974).  Young  stages  are,  for  the  most  part,  easily  identified  to 
family,  but  the  eggs  are  unknown  except  for  a  few  species.  To 
give  some  indication  of  the  amount  of  work  already  directed  to 
these  problems,  a  recent  bibliography  of  young  scombrids  cov- 
ering the  years  1880-1970  listed  170  papers  dealing  with  iden- 
tification of  eggs,  larvae,  and  juveniles  (Richards  and  Klawe, 
1972).  Where  no  specific  references  are  indicated,  information 
is  from  papers  listed  by  Richards  and  KJawe  (1972)  or  Fritzsche 
(1978). 

Development 

Scombrid  eggs  are  very  difficult  to  identify  because  they  re- 
semble the  vast  majority  of  perciform  eggs  characterized  by  0.8- 
1 .9  mm  in  diameter,  smooth  shell,  usually  a  single  oil  globule 
(several  in  Sarda),  narrow  perivitelline  space,  homogenous  yolk 
and  a  variety  of  distribution  patterns  of  pigment  cells  including 
melanophores  and  other  pigments,  usually  yellow,  white  or  green. 
The  latter  three  colors  are  lost  upon  preservation  and  are  only 
useful  for  identifying  living  eggs.  Because  of  the  great  interest 
in  rearing  scombrids  from  eggs,  several  papers  have  appeared 
which  describe  living  eggs,  but  not  enough  species  have  been 
described  to  shed  light  on  relationships.  The  following  works 
should  be  consulted:  Harada  et  al.  (1971);  Mori  et  al.  (1971); 
Richards  and  KJawe  (1972);  Harada,  Muruta  and  Miyashita 
(1973);  Harada,  Muruta  and  Furutani  (1973);  Yasutake  et  al. 
(1973);  Harada  et  al.  (1974);  Ueyanagi  et  al.  (1974);  and  Mayo 
(1973). 

Most  larvae  can  be  identified  using  a  combination  of  char- 
acters, principally  number  of  myomeres,  body  shape,  head  spi- 
nation,  and  distribution  of  melanophores.  Larvae  are  unknown 
for  only  three  genera— Gasterochisma.  Orcynopsis  and  Cybio- 
sarda.  The  present  state  of  knowledge  of  larval  scombrids  is 
shown  in  Table  158.  Morphological  characters  common  to  lar- 
vae of  this  family  are:  (1)  large  head,  large  mouth  opening  and 
large  eye;  (2)  development  of  head  spination;  (3)  posterior  mi- 
gration of  anus  (anus  located  in  anterior  region  of  body  in  early 
larval  stage;  it  migrates  posteriorly  toward  anal  fin  during  de- 
velopment). 

The  following  accounts  follow  the  order  of  presentation  in 
Table  158. 

Scomber  and  Rastrelliger  (Fig.  326).  — These  two  genera  are 
thought  to  be  the  most  primitive  and  lack  some  of  the  larval 
specialization  seen  in  the  other  genera  (Rastrelliger  is  not  illus- 
trated). The  first  dorsal  fin  forms  after  the  second  dorsal  whereas 
in  other  genera  the  first  dorsal  develops  before  the  second  dorsal. 


The  head  is  not  large  (less  than  'A  SL)  in  comparison  with  other 
genera.  The  dorsal  profile  of  the  head  is  gently  arched  from 
above  the  eye  to  the  tip  of  the  snout  which  is  rounded.  Head 
spination  is  not  developed.  The  typical  pigmentation  is  the  pres- 
ence of  melanophores  on  the  mid- ventral  side  of  the  trunk  and 
tail  in  both  genera.  Myomeres  number  3 1 .  The  species  of  Scom- 
ber can  be  separated  except  it  is  difficult  to  distinguish  S.  ja- 
ponicus  from  5.  australasicus.  Head  proportion  and  pre-anal 
length  may  be  useful  as  diagnostic  characters. 

Grammatorcynus  (Fig.  326).  — Grammatorcynus  bilineatus  lar- 
vae resemble  Scomber  and  Rastrelliger  in  dorsal  profile  of  head 
but  have  a  pointed  snout.  Head  spination  is  not  developed  but 
preopercular  spines  are  present.  Typical  pigmentation  is  the 
presence  of  a  lateral  pigmented  blotch  above  the  anal  fin  and 
the  development  of  saddle-shaped  pigment  blotches  on  the  body 
and  a  pigment  patch  on  the  caudal  fin  base  in  larger  larvae.  Also 
characteristic  are  two  lateral  lines  which  are  discernible  at  57 
mm  SL  in  juveniles.  Myomeres  number  31.  The  larvae  were 
recently  re-described  by  Nishikawa  (1979),  but  larvae  of  the 
second  species,  G.  bicarinatus.  recently  recognized  by  Collette 
(1983)  are  unknown. 

Scomberomorus  (Fig.  327).— This  speciose  genus  is  character- 
ized by  having  a  supraoccipital  protuberance  (Euthynnus  has  a 
slightly  discemable  protuberance).  The  head  is  large  with  an 
elongate  snout  and  large  mouth.  Preopercular  spines  are  well 
developed,  and  in  at  least  one  species,  S.  cavalla,  are  the  longest 
in  the  family.  A  spiny  supraorbital  crest  is  well  developed.  Me- 
lanophores appear  on  the  mid-dorsal  and  mid-ventral  side  of 
the  trunk  and  tail.  Adequate  descriptions  have  been  published 
for  S.  cavalla  and  S.  maculatus  and  recently  (Jenkins  et  al., 
1984)  for  5.  commerson.  S.  queenslandicus  and  5.  semifascia- 
tus. 

Acanthocybium  (Fig.  327).— This  single  species  has  been  well 
described  and  is  very  easy  to  recognize  (Wollam,  1969;  Mat- 
sumoto,  1968).  It  is  characterized  by  a  large  number  of  myo- 
meres (62-64),  elongate  gut,  elongate  snout,  and  melanophores 
on  the  bases  of  the  second  dorsal  and  anal  fins  (on  larvae  >6 
mm  SL).  It  is  the  only  species  which  does  not  exhibit  posterior 
migration  of  the  anus. 

Sarda  (Fig.  327).— The  snout  is  moderately  elongate  and  the 
head  spination,  consisting  of  supraorbital  crests,  preopercular 
spines  and  pterotic  spines,  are  well  developed.  Dentition  on 
both  jaws  is  well-developed.  Melanophores  occurring  on  the 
ventral  midline  appear  to  migrate  dorsally  along  myosepta  with 
growth  in  a  posterior  to  anterior  direction.  In  postflexion  larvae 
the  pelvic  and  first  dorsal  fin  are  heavily  pigmented.  Good, 
thorough  descriptions  are  lacking  for  all  of  the  species. 

Gymnosarda  (Fig.  327).— The  larvae  of  this  monotypic  genus 
are  unique  in  the  remarkable  development  of  the  head,  espe- 
cially elongation  of  the  snout,  wide  mouth  with  fang-like  den- 
tition, and  spinous  preopercles,  supraorbital  crests,  and  pterotic 
spines.  The  extremes  of  the  body  proportions  are:  ca.  60%  for 
head  in  SL,  ca.  60%  for  snout  in  head,  and  ca.  85%  for  upper 
jaw  in  head.  Melanophores  are  absent  from  the  tail  region  and 
the  branchiostegal  rays  are  heavily  pigmented.  The  larvae  were 
described  by  Okiyama  and  Ueyanagi  (1977). 


COLLETTE  ET  AL.:  SCOMBROIDEI 


619 


Allothunmis  (Fig.  328).— This  and  the  other  genera  of  Thunnini 
are  very  similar  in  appearance  and  are  separated  on  the  basis 
of  pigment  patterns.  All  five  genera  have  similar  myomere  counts, 
preopercular  spines  present  and  spiny  supraorbital  crests  absent. 
Allothunrms  fallai  has  39  myomeres  and  unique  melanophore 
patterns  are  present  on  the  mid-ventral  surface  of  the  lower  jaw 
along  the  base  of  the  second  dorsal  fin. 

Auxis  (Fig.  329).— There  appear  to  be  two  world-wide  species 
with  39  myomeres  but  there  is  some  variation  in  pigment  pat- 
tern. The  genus  is  characterized  by  having  melanophores  deeply 
embedded  behind  the  midbrain,  cleithral  symphysis,  along  the 
ventral  margin  of  the  tail  and  melanophores  absent  from  the 
forebrain.  The  first  dorsal  fin  is  weakly  developed  and  mela- 
nophores occur  along  the  lateral  midline  of  the  tail  and  on  the 
dorsal  margin  of  the  caudal  peduncle  in  some  specimens.  The 
profile  of  the  head  is  blunt  and  the  jaws  are  short  giving  the 
larvae  a  characteristic  ".'iMA/i-look"  which  is  different  from  the 
next  three  genera. 

Euthynnus  (Fig.  329).— Two  species  have  39  myomeres  and  a 
third,  E.  lineatus,  has  37.  These  larvae  have  slightly  longer 
snouts  than  other  Thunnini  and  a  slight  supraoccipital  protu- 
berance. The  unique  pigment  pattern  is  characterized  by  me- 
lanophores occurring  on  the  forebrain,  midbrain,  cleithral  sym- 
physis, and  ventrally,  laterally  and  dorsally  on  the  tail.  The  first 
dorsal  fin  is  strongly  developed  and  heavily  pigmented. 

Katsuwonus  (Vig.  329).— The  single  species,  K.  pelamis.  has  41 
myomeres  and  a  reduction  in  melanophores  as  they  occur  only 
on  the  forebrain,  midbrain,  one  to  three  distinct  melanophores 
on  the  ventral  margin  of  the  tail  and  rarely  one  or  two  on  the 
dorsal  margin  of  the  caudal  peduncle. 

Thunnus  (Fig.  329).- All  7  species  (Gibbs  and  Collette.  1967) 
have  39  myomeres  and  show  the  greatest  reduction  in  mela- 
nophores in  the  family.  Most  species  can  be  separated  on  the 
basis  of  melanophores.  Thunnus  thynmis  and  T.  inaccoyiihavt 
melanophores  on  the  ventral  margin  of  the  tail  and  the  dorsal 
margin  of  the  trunk  and  tail.  Thunnus  ohesus  and  T.  atlanticus 
have  melanophores  only  on  the  ventral  margin  of  the  tail.  Thun- 
nus alhacares  and  T.  alalunga  lack  tail  melanophores.  Thunnus 
tongol  is  unidentified.  Geographic  distribution,  time  of  spawn- 
ing and  internal  characters  must  be  used  to  identify  larvae  of 
this  genus.  We  recommend  that  the  following  publications  be 
carefully  consulted  before  attempting  specific  identifications: 
Matsumoto  et  al.  (1 972),  Richards  and  Potthofr(  1 974),  Potthoff 
(1974,  1975)  and  Kohno  et  al.  (1982). 

Relationships 

Okiyamaand  Ueyanagi  (1978)  compared  a  classification  based 
on  larval  characters  of  12  genera  of  Scombrinae  with  the  clas- 
sification of  Collette  and  Chao  (1975).  They  selected  13  pre- 
sumed phylogenetically  important  larval  characters  (Okiyama 
and  Ueyanagi,  1978:  table  2)  and  then  coded  the  character  states 
(Table  159).  Their  dendrogram  (Fig.  330)  shows  four  groups. 
Group  A,  Scomber  and  Rastrelliger.  corresponds  to  the  tribe 
Scombrini  (Fig.  324).  Group  B  consists  only  of  Gratnmator- 
cynus.  Group  C  equals  the  Thunnini  (Fig.  324)  plus  Alloihunnus. 


This  interpretation  is  reasonable  on  cladistic  grounds  as  dis- 
cussed in  the  family  section.  Group  D  is  a  mixture  of  the  Scom- 
beromorini  and  Sardini.  Okiyama  and  Ueyanagi  admitted  that 
this  group  is  a  "heterogeneous  assemblage." 

The  question  of  whether  or  not  the  billfishes  should  be  con- 
sidered scombroids  has  been  addressed  by  Potthoff  et  al.  (ms). 
They  studied  osteological  developmental  features  as  shown  in 
Tables  160  and  161  and  Figs.  331  and  332.  Although  their  re- 
search is  still  preliminary  because  of  lack  of  adequate  devel- 
opmental series  for  many  genera,  they  conclude  that  the  Istio- 
phoridae  and  Xiphiidae  should  not  be  placed  within  the 
Scombroidei  because  of  three  developmental  characters  which 
are  not  shared  by  any  other  scombroids.  First,  all  scombroids, 
except  the  Istiophoridae  and  Xiphiidae,  have  distinctive  saddle- 
shaped  ossifications  on  the  vertebrae  before  the  centra  are  fully 
formed.  Second,  development  of  the  cartilaginous  neural  and 
haemal  spines  also  is  similar  in  all  scombroids,  except  istio- 
phorids  and  xiphiids.  Third,  scombroids  except  istiophorids  and 
xiphiids  share  a  primitive  and  an  advanced  development  of  the 
first  and  second  dorsal  and  anal  fins  and  their  supporting  pter- 
ygiophores.  In  the  primitive  development,  which  is  shared  by 
Scombrolabra.x  and  Scombrini  (and  which  is  the  basic  devel- 
opmental pattern  of  percoids),  the  second  dorsal  fin,  anal  fin 
and  pterygiophores  develop  first  from  a  center  anteriorly  and 
posteriorly  and  the  first  dorsal  fin  and  pterygiophores  develop 
second,  also  from  a  center  anteriorly  and  posteriorly.  In  the 
advanced  development,  which  is  shared  by  the  Gempylidae, 
Trichiuridae  and  Thunnini,  the  first  dorsal  fin  and  pterygio- 
phores develop  first  from  the  anteriormost  element  in  a  posterior 
direction,  and  the  second  dorsal  fin,  anal  fin  and  pterygiophores 
develop  second  from  a  center  anteriorly  and  posteriorly.  In  the 
Istiophoridae,  the  first  dorsal  fin  and  pterygiophores  develop 
first  from  a  center  anteriorly  and  posteriorly.  When  the  posterior 
portion  of  the  first  dorsal  fin  development  reaches  above  the 
anterior  portion  of  the  anal  fin,  a  few  anal  rays  and  pterygio- 
phores develop  anteriorly  but  most  are  added  posteriorly.  The 
second  dorsal  fin  develops  only  in  a  posterior  direction  consec- 
utive to  the  first  dorsal  fin.  In  Xiphias  the  second  dorsal  and 
anal  fins  and  pterygiophores  develop  first  from  a  center  ante- 
riorly and  posteriorly.  Development  of  the  first  dorsal  fin  and 
pterygiophores  then  is  continuous  with  the  second  dorsal  fin 
and  in  an  anterior  direction  only. 

In  addition  to  their  work,  one  can  see  the  striking  differences 
between  billfish  larvae  and  other  scombroids  simply  by  review- 
ing the  illustrations  of  larvae  in  this  report.  However,  these 
synapomorphies  of  istiophorids  and  xiphiids  are  not  shared  with 
any  other  group  of  fishes  and  so  cannot  be  used  as  an  argument 
to  relate  the  billfishes  to  any  other  taxa.  Billfishes  have  another 
unique  synapomorphy:  a  specialized  organ  for  heat  production 
located  beneath  the  brain  and  adjacent  to  the  eyes  (Block,  1983). 
The  Scombridae,  Istiophoridae  and  Xiphiidae  have  a  stay  on 
the  4th  phary ngobranchial  that  is  absent  in  other  perciforms  (G. 
D.  Johnson,  pers.  comm.).  Until  further  work  is  completed  and 
other  characters  thoroughly  studied,  the  billfishes  are  retained 
in  the  Scombroidei.  The  larval  evidence  presented  indicates  a 
close  relationship  among  the  families  Scombrolabracidae,  Gem- 
pylidae, Trichiuridae  and  Scombridae  and  much  more  distant, 
if  any,  relationship  to  the  Istiophoridae  and  Xiphiidae. 


620 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Appendix 

Characters  and  character  states  used  for  cladogram  of  scombroid  fishes  (Fig.  312).  1.  Epibranchials.  Tip  of  2nd  epibranchial  short 
(0,  plesiomorphous);  2nd  epibranchial  extends  over  top  of  3rd  infraphar^ngobranchial  to  connect  with  3rd  epibranchial  ( 1 ,  apomorphous). 
2.  Predorsal  bones.  Absent  (1);  present  (0).  3.  Pharyngeal  tooth  plate  stay.  Stay  absent  (0);  stay  present  on  3rd  pharyngeal  tooth  plate 
where  it  contacts  4th  pharyngeal  tooth  plate  (1).  4.  Dorsal  fin  developmental  sequence.  Second  dorsal  develops  before  first  dorsal  (1); 
first  dorsal  develops  before  second  dorsal  (0).  5.  Shape  of  premaxilla  in  larvae.  Not  beak-like  (0),  beak-like  (1).  6.  Cross-connections 
of  gill  filaments.  No  cross-connections  (0);  cross-connections  present  (1).  7.  Number  of  epurals.  Three  (0);  two(l).  8.  Number  of  vertebrae 
supporting  caudal  fin.  Two  (0);  3-4  (I).  9.  Infraorbital  bones.  Not  expanded  into  large  plates  (0);  expanded  mto  large  plates  (1).  10. 
Subocular  shelf  Present  (0);  absent  (1).  11.  Mid-lateral  keel  on  caudal  peduncle.  Absent  (0);  present  (1).  12.  Pair  of  small  keels  at  base 
of  caudal  fin.  Absent  (0);  present  (1).  13.  Bill.  Absent  (0);  present  (1).  14.  Extension  of  caudal  fin  rays  over  hypural  plate.  Not  overlapping 
or  slightly  overlapping  plate  (0);  completely  covering  plate  (1).  15.  Anterior  end  of  infraorbital  bone.  Not  tubular  (0);  tubular  (1).  16. 
Bony  keels  on  caudal  peduncular  vertebrae.  Absent  (0);  present  (1).  17.  Bony  caudal  peduncle  keels.  Poorly  or  irregularly  developed 
(0);  well-developed,  forming  a  wide  plate  (1).  18.  Inner  row  of  premaxillary  teeth.  Additional  row  of  teeth  present  on  antero-medial 
end  of  premaxilla  (1);  single  row  of  premaxillary  teeth  (0).  19.  Prolrusability  of  premaxilla.  Upper  jaw  protrusible,  premaxilla  free  from 
maxilla  (0);  premaxilla  anchored  to  maxilla  (1).  20.  Number  of  ossifications  in  last  dorsal  and  anal  pterygiophores.  A  single  ossification 
(0);  two  ossifications  ( 1 ).  2 1 .  Relationship  of  second  dorsal  fin  pterygiophores  to  neural  spines.  Relationship  2: 1  (0);  1:1(1)22.  Corselet. 
Absent  (0);  present  (1).  23.  Subcutaneous  vascular  system.  Absent  (0);  present  (1).  24.  Fronto-panetal  fenestra.  Absent  (0);  present  (1). 
25.  Tooth  shape.  Conical  (0);  compressed  (1).  26.  Prootic  pits  (on  ventral  surface  of  skull).  Absent  (0);  present  (1).  27.  Vertebral  trellis 
work.  Absent  (0);  present  (1).  28.  Joint  between  first  and  second  infraorbital  bones.  Simple  contact  (0).  tightly  bound  (1).  29.  Number 
of  vertebrae.  Moderate  numbers,  30-31  (1);  few,  24-26  (0);  many,  35-170  (2).  30.  Number  of  uroneurals  in  adult.  Two  (0);  one  (1). 
31.  Fusion  of  uroneural  to  urostyle.  No  fusion  (0);  fused  (1).  32.  Fusion  of  upper  hypural  bones.  Hypurals  3,  4,  and  5  separate  (0);  3 
and  4  fused  ( 1 );  3,  4,  and  5  fused  (2).  33.  Notch  in  hypural  plate.  Large  (0);  small  ( 1 );  absent  (2).  34.  Fusion  of  upper  and  lower  hypural 
plates.  Not  fused  (0);  fused  (1).  35.  Fusion  of  lower  hypural  bones.  Hypurals  1  and  2  separate  (0);  fused  (1).  36.  Fusion  of  parahypural 
to  hypural  plate.  Separate  (0);  fused  (1).  37.  Number  of  autogenous  haemal  spines.  Two  (0);  one  (I).  38.  Tips  of  neural  and  haemal 
spines  of  preural  vertebra  4.  Tips  of  both  flattened  (2);  tip  of  one  flattened  (1);  tips  not  flattened  (0).  39.  Number  of  pectoral  fin  rays. 
17-19  (1,  plesiomorphous);  10-17  (0,  apomorphous);  and  along  another  transition  series  to  17-23  (2),  20-27  (3),  25-29  (4),  and  30- 
36  (5).  40.  Tongue  teeth.  None  fused  to  glossohyal  (0);  two  patches  fused  to  glossohyal  (1). 


(B.B,C,  AND  J.L.R.)  National  Marine  Fisheries  Service,  Na- 
tional Systematics  Laboratory,  National  Museum  of 
Natural  History,  Washington,  District  of  Columbia 
20560;  (T.P.  and  W.J.R.)  National  Marine  Fisheries  Ser- 
vice, Southeast  Fisheries  Center,  75  Virginia  Beach 


Drive,  Miami,  Florida  33149;  (S.U.)  Faculty  of  Marine 
Science  and  Technology,  Tokai  University,  3-20-1 
Orido,  Shimizu-Shi,  424,  Japan;  (Y.N.)  Far  Seas  Fishery 
Research  Laboratory,  Fisheries  Agency,  1000  Orido, 
Shimizu-Shi,  Japan, 


Stromateoidei:  Development  and  Relationships 
M.  H.  Horn 


THE  Stromateoidei  is  a  suborder  of  perciform  fishes  com- 
posed of  six  families,  16  genera  and  approximately  65 
species.  These  fishes  form  a  reasonably  well-defined  group  that, 
with  one  exception,  is  characterized  by  toothed  saccular  out- 
growths in  the  alimentary  tract  immediately  posterior  to  the  last 
gill  arch.  All  species  have  small  uniserial  teeth  in  the  jaws. 

Stromateoids  are  marine  fishes  of  temperate  and  tropical  lat- 
itudes and  range  from  inshore  coastal  waters  to  the  open  ocean 
and  from  pelagic  (primarily)  to  demersal  habitats.  Of  the  16 
stromateoid  genera,  eight  are  exclusively  oceanic,  two  are  mixed 
coastal  and  oceanic  and  six  are  exclusively  coastal  (Table  162). 
Although  several  coastal  species  are  locally  abundant  and  com- 
mercially important,  oceanic  stromateoids  tend  to  be  rare  and 
sporadic  in  occurrence.  Juveniles  commonly  associate  with  an- 
imate or  inanimate  floating  objects  in  the  surface  layers  of  the 
ocean. 

Since  the  completion  of  Haedrich's  (1967)  comprehensive 
review  of  the  stromateoid  fishes,  several  taxonomically-oriented 
studies  have  been  conducted.  These  works  include  descriptions 
of  a  new  monotypic  family  (Haedrich,  1 969)  and  four  new  species 


(Haedrich,  1970;  Horn,  1970b;  Chirichigno,  1973;  McAllister 
and  Randall,  1975),  generic  reviews  or  revisions  (Horn,  1970b, 
1973;  Butler,  1979),  regional  reviews  of  certain  centrolophid 
taxa  (Stehmann  and  Lenz,  1973;  McDowall,  1982)  and  an  ex- 
tensive account  of  the  early  life  history  stages  of  pelagic  stro- 
mateoids (Ahlstrom  et  al.,  1976).  The  present  paper,  which 
includes  a  phylogenetic  analysis  of  stromateoid  genera,  draws 
heavily  upon  information  in  Haedrich  ( 1967)  and  Ahlstrom  et 
al.  (1976). 

Development 

Eggs 

The  eggs  of  approximately  14  species  representing  six  genera 
and  four  families  of  stromateoids  have  been  described  (Table 
163).  Stromateoid  eggs  typically  are  relatively  small  (0.70-1.80 
mm  in  diameter),  pelagic,  separate  and  spherical.  They  have 
unsculptured  surfaces,  unsegmented  yolks  and  single  oil  glob- 
ules. The  few  distinctive  features  of  the  eggs  limit  their  value 
as  a  source  of  taxonomic  characters. 


■^/-._ 


Fig.  333.  Examples  of  stromateoid  larvae  and  early  juveniles.  (A)  Amarsipus  carlshergi  (Amarsipidae),  16.7  mm  postflexion  larva;  (B) 
Schedophilus  hulloni  (Centrolophidae),  25.0  mm  early  juvenile:  (C)  Icichthys  lockingtom  (Cenlrolophidae),  20.0  mm  early  juvenile;  (D)  Nomeus 
gronovii  (Nomeidae),  22.7  mm  early  juvenile;  all  from  Ahlstrom  et  al.  (1976). 


622 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  162.    Family  Affiliation.  Habitat  and  Ranges  of  Meristic  Values  for  Stromateoid  Genera. 


Amarsipidae 

Centrolophidae 

Amarsipus 

Hyperoglyphe 

Cenfrolophus 

Schedophihts 

Icithlhys 

Tuhhia 

StTiolella 

Habitat 

Coastal 

X 

X 

Oceanic 

X 

X 

X 

X 

Coastal  and  oceanic 

X 

Meristic 

First  dorsal  spines 

X-XII 

— 

_ 

— 

— 

— 

— 

Second  dorsal  spines 

and 

rays 

23-2T 

Vl-VIll,  15-26 

V, 32-37 

V-IX.  23-54 

38-46- 

47-51- 

Vll-lX,  25-39 

Anal  spines  and  rays 

29-32» 

111,  14-20 

111.21-24 

11-111.  16-35 

25-32- 

33-37- 

11-111,  19-24 

Pectoral  rays 

17-19 

18-23 

19-23 

19-22 

16-21 

18-21 

19-23 

Precaudal  vertebrae 

16-18 

10 

10 

10-12 

23-25 

7 

10-11 

Caudal  vertebrae 

29-31 

14-15 

15 

15-20 

34-37 

9 

14-16 

Total  vertebrae 

46-48 

24-25 

25 

25-32 

49-61 

43-45 

25-26 

'  Total  fin  elements. 


Larvae  and  Juveniles 

The  larvae  and/orearly  juveniles  of  approximately  28  species 
representing  1 3  genera  and  all  six  stromateoid  families  have 
been  described  (Table  163;  see  Figs.  333  and  334). 

Ahlstrom  et  al.  (1976)  considered  a  young  stromateoid  to  be 
a  larva  and  not  a  juvenile  if  it  (1)  had  not  completely  formed 
rays  in  all  fins  and/or  (2)  had  not  initiated  development  of  scales. 
Scale  formation  is  seldom  encountered  on  specimens  that  lack 
the  full  complement  of  fin  rays. 


Characters  of  larvae  and  early  juveniles  that  Ahlstrom  et  al. 
(1976)  found  useful  in  distinguishing  among  stromateoid  fishes 
include  meristics.  sequence  of  fin  formation,  morphometries, 
pigmentation  and  skeletal  features.  Of  the  strictly  ontogenetic 
characters,  sequence  of  fin  formation  and  pigmentation  pattern 
are  the  most  important.  The  overlap  in  character  states  and  the 
lack  of  data  for  certain  genera  strongly  limit  the  number  of 
unambiguous  characters  useful  in  developing  a  stromateoid 
phylogeny. 


Table  163.    Stromateoid  Species  for  which  Eggs  and/or  Larvae  Have  Been  Described. 


Species 


Eggs 


Amarsipidae 
Centrolophidae 


Nomeidae 


Tetragonuridae 

Ariommidae 
Stromateidae 


Amarsipus  carlsbergi 

Schedophilus  ovalis 
Centrolophus  niger 
Senolella  brama 
Seriolella  caerulea 
Seriolella  punctata 
Icichthys  lockingtoni 

Cubiceps  baxteri 
Cubiceps  caeruleus 
Cubiceps  capensis 
Cubiceps  gracilis 
Cubiceps  paradoxus 
Cubiceps  pauciradiatus 
Cubiceps  squamiceps 
Nomeus  gronovii 
Psenes  arafurensis 
Psenes  cyanophrys 
Psenes  maculatus 
Psenes  pellucidus 
Psenes  sio 

Telragonurus  atlanticus 
Telragonurus  cuvieri 
Tetragonurus  pacificus 

Ariomma  regutus 

Siromaleus  fiatola 
Peprilus  burli 
Peprilus  paru 
Peprilus  simtltimus 
Peprilus  triacanthus 

Pampus  chinensis 


X? 


X? 
X 

X 
X 


Ahlstrom  et  al.  (1976) 

Padoa (1956b) 

Sanzo  (1932b)  (as  C  pompilus) 
Grimes  and  Robertson  (1981) 
Grimes  and  Robertson  (1981) 
Grimes  and  Robertson  (1981) 
Ahlstrom  et  al.  (1976) 

Ahlstrom  et  al.  (1976)  (as  C.  caeruleus.  see  Butler,  1979) 

Ahlstrom  et  al.  (1976)  (as  C.  capensis,  see  Butler.  1979) 

Ahlstrom  et  al.  (1976)  (as  C.  sp.  A.  see  Butler.  1979) 

Sparta  (1946) 

Ahlstrom  et  al.  (1976)  (as  C  sp.  B,  see  Butler.  1979) 

Ahlstrom  et  al.  (1976) 

Nellen  (1973b)  (as  Psenes  whiteleggii.  see  Ahlstrom  et  al., 

Ahlstrom  et  al.  (1976) 

Ahlstrom  et  al.  (1976) 

Ahlstrom  et  al.  (1976) 

Ahlstrom  et  al.  (1976) 

Ahlstrom  et  al.  (1976) 

Ahlstrom  et  al.  (1976) 

Grey  (1955b),  Ahlstrom  et  al.  (1976) 
Grey  (1955b),  Ahlstrom  et  al.  (1976) 
Grey  (1955b),  Ahlstrom  et  al.  (1976) 

McKenney  (1961)  (as  Psenes  regulus) 

Padoa (1956b) 

Ditty  and  Truesdale  (1983) 

Martin  and  Drewry  (1978),  Ditty  and  Truesdale  (1983) 

D" Vincent  et  al.  (1980) 

Colton  and  Honey  (1963),  Martin  and  Drewry  (1978), 

Ditty  and  Truesdale  (1983) 
Pali  (1979) 


1976) 


'  Early  juveniles 


HORN:  STROMATEOIDEI 

Table  162.    Extended. 


623 


Cenlrolophidae 

Nomeidae 

Anommidae 

Tetragonundae 

Stromateidae 

Psenopsis 

Cubtceps 

Notneus 

Psenes 

Ariomma 

Telragonurus 

Siromaieus 

Pepnius 

Pampus 

X 


X 


X 


— 

X-XII 

Xl-XII 

IX-XII 

XI-XII 

XI-XVIII 

_ 

_ 



V-VII,  26- 

-32 

I,  15-27 

25-27" 

I,  19-30 

I,  13-17 

10-13 

42-56' 

II- 

-IV,  38- 

■49 

0-X,  33-50 

II-IV,  22- 

•27 

II 

-III,  14-23 

II,  24-26 

II- 

-III,  19-30 

II 

-III,  13-16 

I-II,  8-10 

II- 

-III,  31- 

45 

II- 

-IV,  35- 

■47 

0-VII,  39-47 

16-23 

17-23 

21-23 

16-22 

20-25 

14-21 

18-25 

17-24 

24-27 

10 

12-13 

14 

12-13 

13-14 

18-27 

18-19 

12-15 

14-16 

15 

18-21 

27 

19-29 

17 

20-28 

24-26 

16-22 

19-26 

25 

30-34 

41 

31-42 

30-31 

39-54 

41-49 

29-36 

33^1 

Behavioral  and  morphological  features  of  young  stromateoids 
are  potentially  informative  as  taxonomic  characters.  Certain 
ones  of  these  traits  appear  to  be  related  to  the  widespread  as- 
sociation of  these  fishes  with  a  variety  of  floating  objects  in  the 
ocean.  In  general,  loss  of  the  swimbladder  accompanies  allo- 
metric  growth  in  pectoral  fin  length  and  changes  in  pigmentation 
pattern  as  part  of  the  transition  from  the  juvenile  to  the  adult 
stage  (Horn,  1975).  Stromateoid  fishes  associated  with  floating 
objects  usually  have  conspicuous  blotches  or  bands  of  pigment 
on  their  bodies  as  juveniles  then  become  more  uniformly  pig- 
mented as  deeper-living,  presumably  independent  and  contin- 
uously swimming  adult  fish.  Haedrich  (1967)  proposed  that 
banding  is  protective  coloration  for  the  fishes  during  the  period 
when  they  live  in  the  shifting  shadows  beneath  jelly  fishes.  There 
are  exceptions  to  this  apparent  relationship  between  pigmen- 
tation and  behavior.  For  example,  juveniles  of  Ariomma  are 
banded  yet  appear  to  seldom  associate  with  floating  objects 
whereas  young  Telragonurus  are  uniformly  pigmented  but,  as 
Janssen  and  Harbison  (1981)  observed,  associate  intimately  with 
salps  and  pyrosomes.  The  re?ra^o«Mn/5-salp/pyrosome  asso- 
ciation, however,  is  different  in  that  the  fish  are  inside  rather 
than  beneath  the  floating  objects.  Pigmentation  pattern  and  type 
of  association  are  the  two  ontogenetic  characters  used  in  the 
phylogenetic  analysis  (see  below). 

Fin  characters.  —  Meristic  characters  (Table  1 62)  have  been  used 
widely  to  distinguish  stromateoid  taxa  especially  at  the  species 
level  (e.g.,  Haedrich,  1967;  Haedrich  and  Horn,  1972;  Horn, 
1970b,  1973;  Horn  and  Haedrich,  1973;  Ahlstrom  et  al.,  1976; 
Butler,  1979;  McDowall,  1982).  As  in  most  other  perciform 
fishes,  the  pelvic  fin  (I,  5)  and  caudal  fin  (17  principal  rays,  15 
branched)  of  stromateoids  have  stabilized  counts  (the  pelvic  fin, 
however,  is  absent  in  three  stromateoid  genera).  The  number 
of  secondary  caudal  rays,  although  exhibiting  intraspecific  vari- 
ation, can  be  an  important  taxonomic  character  among  species 
within  a  genus  (Ahlstrom  et  al.,  1976).  The  dorsal  fin  of  stro- 
mateoids may  be  continuous  or  divided  into  two  fins.  This  trait 
is  used  as  a  generic  character  in  the  present  paper  (Tables  164, 
1 65).  It  is  not  always  possible  to  distinguish  between  spines  and 
rays  in  those  species  with  a  continuous  dorsal  fin  (see  Table 
162).  The  complement  of  anal  fin  rays  in  stromateoids  is  pre- 
ceded by  0  to  7  anal  spines  with  most  species  having  2  or  3 
spines.  The  number  of  pectoral  fin  rays  varies  from  14  to  27 


among  stromateoids,  but  the  overlap  among  species  limits  its 
use  as  a  taxonomic  character. 

Two  different  sequences  of  fin  formation  occur  in  oceanic 
stromateoids  depending  primarily  on  whether  the  pelvic  fins 
form  early  (before  the  other  fins)  or  whether  they  form  late. 
Ahlstrom  et  al.  (1976)  found  that  the  pelvics  are  first  to  form 
in  Amarsipus,  Psenes  and  probably  also  Nomeus  whereas  they 
are  last  to  form  in  Cubiceps,  Icichthys  and  Telragonurus.  Fahay 
(1983)  reported  that  the  pelvics  are  also  last  to  form  in  Centro- 
lophus.  These  ontogenetic  patterns  have  potential  significance 
as  taxonomic  characters;  however,  they  must  be  described  for 
other  genera  before  they  can  contribute  to  an  understanding  of 
stromateoid  relationships. 

Morphometries.  — SXrom?i\eo\ds  vary  substantially  in  their  mor- 
phologies, especially  body  depth,  but  show  no  abrupt  meta- 
morphic  changes  in  the  transition  from  the  larval  to  the  juvenile 
to  the  adult  stage.  Allometric  growth  is  common  in  these  fishes 
and  complicates  the  use  of  morphometries  as  taxonomic  char- 
acters. Taxa  at  similar  stages  of  development  must  be  compared 
if  morphometric  characters  are  to  have  validity.  Ahlstrom  et  al. 
(1976)  used  morphometries  in  distinguishing  among  species  in 
genera  such  as  Schedophilus  and  Psenes.  Because  of  allometry 
and  the  less  than  complete  information  on  different  develop- 
ment stages  of  several  genera,  morphometric  characters  were 
not  used  in  the  phylogenetic  analysis  of  stromateoid  genera  (see 
below). 

Skeletal  characters.  —  M\\s,Xrom  et  al.  (1976)  in  their  study  of 
the  early  life  history  stages  of  oceanic  stromateoids  found  the 
following  skeletal  characters  to  be  of  particular  relevance:  (1) 
total  number  of  vertebrae,  (2)  co-occurrence  of  a  pair  of  pleural 
ribs  and  a  haemal  spine  on  each  of  one  or  more  caudal  vertebrae, 

(3)  separation  of  vertebrae  into  precaudal  and  caudal  groups, 

(4)  position  of  anal  fin  pterygiophores  in  relation  to  haemal 
spines,  (5)  number  and  position  of  dorsal  fin  pterygiophores  and 
predorsal  bones  in  relation  to  neural  spines,  and,  (6)  the  number 
of  supporting  bones  of  the  caudal  fin.  While  not  strictly  onto- 
genetic in  nature,  these  characters  are  most  readily  discerned 
from  examiniation  of  cleared  and  stained  larvae  and  early  ju- 
veniles. 

Of  the  above  characters,  only  the  number  of  predorsal  bones 
and  the  number  of  hypurals  were  used  in  the  phylogenetic  anal- 


^<<Z2:^?^'^-^^  '^'"'^y^.s 


B 


Fig.  334.  Examples  of  stromateoid  larvae  and  early  juveniles.  (A)  Cuhiceps  pauciradiatus  (Nomeidae),  17.5  mm  early  juvenile;  (B)  Psenes 
cyanophn's  (Nomeidae),  19.1  mm  early  juvenile;  (C)  Tetragonurus  atlanticus  (Tetragonuridae),  17.2  mm  postfiexion  larva;  (D)  Ariomma  sp. 
(Ariommidae),  14.4  mm  early  juvenile.  Gulf  of  Mexico;  (E)  Pepnius  similtimus  (Stromaleidae),  10.8  mm  postflexion  larva.  A-C  from  Ahlstrom 
et  al.  (1976),  D  drawn  by  Betsy  Washington.  E  from  D'Vincent  et  al.  (1980). 


HORN:  STROMATEOIDEI 


625 


Table  164.    Characters  and  Character  States  Used  in  the  Phylogenetic  Analysis  of  Stromateoid  Genera. 


Character  stale  codes 

Character 

0 

1 

2 

3 

1 .  Number  of  rows  of  premaxillary  teeth 

1 

2 

2.  Number  of  rows  of  dentary  teeth 

1 

2 

3.  Pharyngeal  sac 

absent 

present 

4.  Shape  of  pharyngeal  sac 

height  >  length 

height  =  length 

height  <  length 

5.  Arrangement  of  papillae  in 

10-20  bands 

5-7  bands 

not  in  bands 

pharyngeal  sac 

6.   Papillae  on  upper  portion  of 

present 

absent 

pharyngeal  sac 

7.   Position  of  papillae  in  pharyngeal 

not  on  stalks 

short  stalks,  teeth 

long  stalks,  teeth 

sac 

on  end 

along  stalk 

8.  Shape  of  papillae  base  in  pharyngeal 
sac 

9.  Condition  of  maxilla 

round 

stellate 

mobile 

fixed 

10.  Supramaxillary  bone 

present 

absent 

1 1.   Lacrimal  bone 

prominent 

reduced 

highly  reduced 

12.  Relationship  of  gills  to  isthmus 

free 

united 

13.  Pseudobranch 

present 

absent 

14.  Scale  type 

cycloid 

ctenoid 

15.  Opercular  scalation 

present 

absent 

16.  Preopercular  scalation 

present 

absent 

17.  Prominent  preopercular  spines 

absent 

present 

18.  Number  of  branchiostegal  rays 

7 

6 

5 

19.   Pelvic  bone  and  fin 

fin  present 

fin  absent,  bone 

fin  absent. 

fin  absent,  highly 

with  spine 

reduced  bone 

reduced  bone 

20.  Number  of  predorsal  bones 

3 

2-3- 

0 

7-12 

21.  Number  of  dorsal  fins 

1 

2 

22.  Keels  on  caudal  peduncle 

absent 

present 

23.  Number  of  hypurals 

6 

3-5 

2 

24.  Procurrent  spur 

present 

reduced 

absent 

25.   Ray  base  preceding  procurrent  spur 

shortened 

slightly  shortened 

not  shortened 

26.  Juvenile  pigmentation 

uniform 

patterned 

27.  Primary  juvenile  association 

independent 

floating  objects 

inanimate  floating 
objects 

animate  floating 
objects 

■  This  character  slate  overlaps  adjacent  stale  but  occurs  in  only  one  taxon  (Girellidae). 


ysis  of  stromateoid  genera  (Tables  1 64,  165;  Fig.  335).  The  other 
characters  were  not  used  because  they  are  not  known  for  all 
genera  or,  if  known,  their  values  overlap  and,  therefore,  cannot 
be  coded  without  ambiguity. 

The  pharyngeal  sac  as  a  skeletal  feature  was  a  rich  source  of 
characters  in  developing  the  phylogenetic  hypothesis  for  stro- 
mateoid genera.  Five  characters  were  used  ranging  from  the 
shape  of  the  sac  to  the  arrangement,  location  and  position  of 
papillae  within  the  sac  (Tables  164,  165;  Fig.  335). 

Pigmentation.  —  Differences  in  pigmentation  are  mainly  of  value 
in  distinguishing  species  within  genera  for  which  the  larval  and 
juvenile  stages  are  relatively  well  known.  Ahlstrom  et  al.  ( 1 976) 
used  pigmentation  patterns  to  demonstrate  differences  among 
species  of  oceanic  stromateoid  genera  (see  Figs.  333  and  334). 
Adults  tend  to  be  more  uniform  in  pigmentation  and,  hence, 
offer  fewer  apparent  taxonomic  characters. 

Stromateoids  vary  both  in  the  density  and  in  the  pattern  of 
their  pigmentation.  As  larvae  and  early  juveniles,  some  species, 
e.g.,  Amarsipus  carlshergi  (Fig.  333A)  are  sparsely  pigmented 
whereas  others  are  more  heavily  pigmented,  e.g.,  Icichthys  lock- 
inglom  (Fig.  333C).  Certain  larvae  and  juveniles  are  rather  uni- 
formly pigmented,  e.g.,  Ciibiceps paiuiradiatus {Fig.  334A),  7et- 
ragonunis  atlanticiis  (Fig.  334C)  and  Pepriliis  simillimus  (Fig. 
334E)  while  others  have  their  pigment  concentrated  into  bands 
or  blotches,  e.g.,  Schcdophilus  huttoni  (Fig.  333B).  Numeus 


gronovii  (¥{%.  333D).  Psenes  cyanophrys  (Fig.  334B)and /lr/o«;- 
ma  sp.  (Fig.  334D).  Ahlstrom  et  al.  (1976)  used  various  detailed 
patterns  to  distinguish  the  larvae,  especially,  and  early  juveniles 
of  species  within  certain  stromateoid  genera.  In  the  present  study, 
uniform  vs  patterned  pigmentation  was  the  only  pigmentation 
character  available  that  could  be  coded  unambiguously  for  all 
stromateoid  genera  (Tables  164,  165). 

Relationships 
Relationships  within  the  Stromateoidei 
Haedrich's  (1967)  analysis  continues  to  be  the  major  system- 
atic work  on  stromateoid  fishes.  He  recognized  five  families  and 
two  main  lineages  in  the  stromateoids.  One  lineage  is  composed 
of  the  Centrolophidae  and  its  derivative,  the  Stromateidae.  The 
other,  a  less  compact  assemblage,  is  comprised  of  the  Nomeidae 
and  its  two  derivatives,  the  Ariommidae  and  the  Tetragonuri- 
dae.  The  Centrolophidae  and  the  Nomeidae  contain  the  basal 
stocks  with  the  centrolophids  having  the  more  primitive  mem- 
bers. Haedrich  (1967)  considered  members  of  the  centrolophid 
genus  Hyperoglyphe  to  be  the  most  generalized  fishes  in  the 
suborder  and  probably  not  unlike  the  ancestral  form.  He  viewed 
the  Stromateidae  as  the  current  zenith  of  stromateoid  evolution 
with  Pampus  as  the  most  advanced  stromateid  genus.  In  his 
interpretation  of  stromateoid  relationships.  Haedrich  (1967) 
recognized  trends  in  the  evolution  of  several  characters  includ- 


626 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  165.     Matrix  of  Character  State  Codes  (see  Table  164  and  Fig.  335  for  the  stromateoid  genera  and  for  the  three  perciform  families 
used  as  out-groups  in  the  phylogenetic  analysis).  Dashes  indicate  characters  that  are  inapplicable.  Question  marks  indicate  unknown  character 

states  and  were  designated  as  "missing  observations"  in  the  analysis. 


1 .  Number  of  rows  of  premaxillary  teeth 

2.  Number  of  rows  of  dentary  teeth 

3.  Pharyngeal  sac 

4.  Shape  of  pharyngeal  sac 

5.  Arrangement  of  papillae  in  pharyngeal  sac 

6.  Papillae  on  upper  portion  of  pharyngeal  sac 

7.  Position  of  papillae  in  pharyngeal  sac 

8.  Shape  of  papillae  base  in  pharyngeal  sac 

9.  Condition  of  maxilla 
10.  Supramaxillary  bone 
1  1.   Lacrimal  bone 

12.  Relationship  of  gills  to  isthmus 

13.  Pseudobranch 

14.  Scale  type 

15.  Opercular  scalation 

16.  Preopercular  scalation 

17.  Prominent  preopercular  spines 

1 8.  Number  of  branchiostegal  rays 

19.  Pelvic  bone  and  fin 

20.  Number  of  predorsal  bones 

21.  Number  of  dorsal  fins 

22.  Keels  on  caudal  peduncle 

23.  Number  of  hypurals 

24.  Procurrent  spur 

25.  Ray  base  preceding  procurrent  spur 

26.  Juvenile  pigmentation 

27.  Primary  juvenile  association 


Siromateoid  genera 

Amarsipus 

Hyperoglyphe 

Centrolophus 

Schedophtlus 

Icicfilhys 

Tuhbm 

SerioleUa 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

1 

1 

1 

1 

1 

— 

0 

0 

0 

0 

0 

0 

— 

0 

0 

0 

0 

0 

0 

— 

0 

0 

0 

0 

0 

0 

— 

0 

0 

0 

0 

0 

0 

— 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

0 

0 

0 

0 

0 

0 

1 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

0 

0 

0 

0 

0 

0 

1 

1 

1 

1 

0 

0 

1 

0 

0 

0 

1 

0 

0 

0 

1 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

3 

0 

0 

1 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

? 

0 

2 

? 

0 

0 

0 

7 

0 

2 

? 

2 

0 

1 

1 

1 

1 

1 

1 

0 

2 

3 

3 

3 

3 

3 

ing  body  size  and  shape,  fin  pattern,  presence  or  absence  of 
palatal  dentition,  shape  of  the  papillae  in  the  phai7ngeal  sacs 
and  the  number  of  branchiostegal  rays,  vertebrae  and  epural 
plus  hypural  elements  in  the  caudal  skeleton. 

Haedrich  ( 1 969)  in  describing  the  Amarsipidae  added  a  sixth 
family  to  the  stromateoid  suborder.  This  family  exhibits  a  mix- 
ture of  primitive  and  derived  characters.  It  lacks  pharyngeal 
sacs,  but  the  pharyngeal  teeth  are  extraordinarily  developed  and 
may  perform  a  shredding  function  analogous  to  the  sacs  of  other 
stromateoids.  Haedrich  (1967)  argued  that  possession  of  a  per- 
ciform caudal  skeleton,  uniserial  jaw  teeth,  an  expanded  lacri- 
mal bone,  an  inflated  and  protruding  top  of  the  head,  an  exten- 
sive subdermal  canal  system  and  a  bony  bridge  over  the  anterior 
vertical  canal  of  the  ear  provides  the  basis  for  placement  of 
Amarsipus  in  the  suborder  Stromateoidei.  He  considered  the 
new  family  to  be  distantly  allied  with  the  Nomeidae. 

In  the  present  study,  a  phylogeny  of  the  1 6  stromateoid  genera 
was  constructed  using  a  set  of  27  characters  (Tables  164,  165) 
that  could  be  coded  for  all  genera  with  little  or  no  overlap  and 
ambiguity.  Initially,  a  larger  number  of  prospective  characters 
(~65)  were  identified  and  evaluated.  Comparison  of  stroma- 
teoids with  presumed  out-groups  helped  to  generate  characters 
and  to  establish  polarity  in  the  transformation  series.  Charac- 
ters, however,  were  omitted  if  they  could  not  be  quantified  or 
if  insufficient  information  was  available  to  characterize  every 
taxon.  In  a  few  cases,  character  states  were  coded  as  "missing 
observations"  if  three  or  fewer  genera  required  this  coding  and 
the  characters  were  judged  important  in  resolving  relationships 
between  the  other  genera. 

Three  closely  related  perciform   families— Girellidae,   Ky- 


phosidae  and  Scorpididae— were  used  as  potential  out-groups 
in  the  analysis  (see  below  for  rationale).  Although  these  three 
taxa  are  frequently  classified  as  subfamilies  of  the  Kyphosidae 
(e.g..  Nelson,  1976),  G.  D.  Johnson  (pers.  comm.)  considers 
them  to  be  distinct  families. 

The  analysis  was  performed  using  the  PHYSYS  package  which 
differentiates  taxa  based  on  the  presence  of  shared  derived  char- 
acters (synapomorphies).  Several  phylogenetic  trees  were  gen- 
erated from  the  genus-character  data  matrix  using  the  Wagner 
distance  algorithm  (see  Farris,  1970;  Wiley,  1981).  These  trees 
were  diagnosed  to  identify  the  origin  of  each  apomorphy  and 
to  examine  character  reversals  and  convergences.  Transfor- 
mation Series  Analysis  was  performed  on  the  data  to  verify 
polarities  developed  through  out-group  comparison  (see  Mick- 
evich,  1982)  and  to  resolve  nonlinear  series.  Further  optimizing 
produced  the  most  parsimonious  tree  from  the  data  matrix  (Fig. 
335). 

This  phylogenetic  tree  (Fig.  335)  was  basically  similar,  with 
certain  exceptions,  to  that  proposed  by  Haedrich  (1967).  //i'- 
peroglyphe  emerged  as  the  most  plesiomorphic  stromateoid  tax- 
on  possessing  a  pharyngeal  sac  and  Pampus  as  the  genus  with 
the  greatest  number  of  apomorphies.  Arioinma  also  ranked  as 
a  highly  derived  genus  in  the  suborder.  Despite  its  several  syn- 
apomorphies with  advanced  stromateoid  genera,  Amarsipus 
emerged  as  the  sister  taxon  of  all  other  stromateoid  genera  pri- 
marily because  it  lacks  a  pharyngeal  sac.  The  major  differences 
between  the  present  analysis  and  Haedrich's  interpretation  lie 
with  the  relationships  of  Arioinma  and  Tctragonurus  to  other 
stromateoids  and  with  the  family  limits  of  the  suborder.  Based 
on  the  cladogram,  Tctragonurus  and  Ariomma  are  more  closely 


HORN:  STROMATEOIDEI 

Table  165.    Extended. 


627 


Slromateoid  genera 

Perciform  families 

Psenopsts 

Cubiceps 

Nomeiis 

Psenes 

Anomma 

Tetragonurus 

Stromateiis 

Pepnliis 

Pampus 

Kyphosidae 

Scorpididae 

Girellidae 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

1 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

1 

1 

1 

1 

1 

1 

1 

1 

1 

1 

0 

0 

0 

0 

0 

0 

0 

2 

2 

1 

1 

1 







0 

1 

1 

1 

2 

2 

2 

2 

2 







0 

0 

0 

0 

1 

0 

0 

0 

0 







0 

1 

1 

1 

2 

1 

2 

2 

2 







0 

1 

1 

1 

0 

0 

1 

1 

1 







0 

1 

0 

1 

0 

1 

0 

1 

0 
1 

0 

1 

0 

1 

0 
1 

1 
1 

0 
0 

0 
0 

0 
0 

0 

0 

0 

0 

0 

0 

1 

1 

2 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

0 

0 

0 

0 

0 

0 

1 

0 

1 

0 

0 

0 

1 

1 

1 

I 

0 

0 

0 

1 

0 

0 

0 

0 

0 

0 

0 

1 

0 

0 

0 

1 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

1 

1 

1 

1 

1 

1 

2 

0 

0 

1 

0 

0 

0 

0 

0 

0 

3 

1 

2 

0 

0 

0 

0 

0 

0 

0 

0 

2 

0 

0 

0 

0 

0 

1 

0 

1 

1 

1 

1 

1 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

1 

1 

0 

0 

0 

0 

0 

0 

0 

1 

1 

1 

2 

I 

1 

1 

1 

1 

1 

1 

? 

1 

0 

0 

2 

2 

0 

0 

0 

0 

0 

0 

7 

1 

1 

2 

2 

2 

2 

2 

2 

0 

0 

2 

0 

0 

1 

1 

1 

0 

1 

0 

0 

1 

0 

0 

3 

3 

3 

1 

0 

3 

3 

3 

3 

1 

0 

1 

related  to  the  stromateids  than  to  the  nomeids  as  proposed  by 
Haedrich  (1967).  Of  the  six  families  recognized  by  Haedrich 
and  Horn  (1972),  only  the  Amarsipidae,  Nomeidae  and  Stro- 
mateidae  appear  to  be  monophyletic  based  on  the  generic  re- 
lationships expressed  in  the  cladogram.  The  Centrolophidae,  on 
the  other  hand,  lacks  a  synapomorphy  and  there  is  no  indication 
that  Tetragonuridae  and  Ariommidae  should  be  considered  dis- 
tinct families.  Additional  characters,  however,  should  be  ex- 
amined before  a  change  in  classification  is  proposed. 

The  Scorpididae  and  Girellidae  were  part  of  a  trichotomy  at 
the  base  of  the  tree  and  together  formed  the  plesiomorphic  out- 
group  cluster  in  the  analysis. 

Character  diagnosis  showed  that  reversals  in  character  trans- 
formation occurred  most  frequently  (>  3  taxa  or  stems/character 
state)  for  scale  type  (no.  14)  and  juvenile  pigmentation  (no.  26). 
In  the  same  diagnosis,  character  convergences  occurred  most 
often  for  the  supramaxillary  bone  (no.  10),  opercular  scalation 
(no.  15),  number  of  branchiostegal  rays  (no.  18),  number  of 
dorsal  fins  (no.  21)andjuvenileassociations(no.  27).  Atnarsipus 
was  involved  in  all  five  of  these  cases  of  apparent  convergence, 
an  indication  of  its  uncertain  phylogenetic  position. 

The  tree  remains  incompletely  resolved  with  three  polychot- 
omies  (Fig.  335).  In  addition  to  the  trichotomy  at  the  base  of 
the  tree,  the  other  two  nodes  with  multiple  branches  involve 
centrolophid  genera.  Lack  of  full  resolution  in  this  region  of  the 
tree  indicates  that  further  work  is  needed  to  clarify  the  inter- 
generic  relationships  of  the  Centrolophidae.  Extending  the  anal- 
ysis to  the  species  level  would  provide  greater  resolution. 

Strictly  larval  or  juvenile  characters  have  contributed  little  to 


the  broad  understanding  of  stromaleoid  intergeneric  relation- 
ships as  perceived  by  Haedrich  (1967)  or  as  analyzed  in  the 
present  study.  Elimination  of  the  two  juvenile  characters  (pig- 
mentation and  associations)  from  the  present  analysis  resulted 
in  a  tree  virtually  identical  to  that  with  them  included  (Fig.  335). 
The  study  of  the  early  life  history  stages  of  pelagic  stromateoids 
by  Ahlstrom  et  al.  (1976),  however,  is  a  major  contribution  to 
stromateoid  systematics  especially  in  developing  an  approach 
that  can  potentially  expand  to  all  taxa  in  the  suborder.  Their 
use  of  ontogenetic  characters  was  important  at  the  species  level 
and  particularly  valuable  in  distinguishing  the  species  and  gen- 
era of  nomeids. 

Characters  employed  by  Ahlstrom  et  al.  (1976)  that  hold 
promise  for  resolving  relationships  among  stromateoids  in  gen- 
eral include  ( 1 )  sequence  of  fin  formation,  (2)  arrangement  of 
anal  fin  pterygiophores  in  relation  to  haemal  spines,  (3)  head 
armature  and  (4)  pigmentation  patterns.  The  caudal  fin  complex, 
while  not  representing  a  strictly  ontogenetic  suite  of  features, 
also  appears  likely  to  provide  characters  if  a  full  spectrum  of 
cleared  and  stained  larvae  are  carefully  examined.  Finally,  the 
various  types  of  associations  juvenile  stromateoids  hold  with 
floating  objects  may  be  more  specific  than  generally  thought  and 
could  become  a  rich  source  of  characters. 


Relationships  of  the  Stromateoidei  to 
other  groups 

Haedrich  ( 1 967)  in  his  review  of  stromateoid  systematics  pro- 
posed that  the  group  arose  from  within  a  relatively  undiffer- 
entiated assemblage  of  perciform  families  including  the  Arri- 


628 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Out-Groups       Amarsipidae 


C  e  n  t  r  o  I  o  p  h  i  d  a  e 


Ariommidae 
Tetragonuridae 


^® 


»         .  .o^        •;;?>'      -c*' 


.<f 


/  /  /  /  /  /  /  /  r  / 


Nomeidae 


S  t  r  0  m  a  t  e  i  d  a  e 


^  r  /'  J"' ./  ./  y  y  /  '^ 


#  /  /"  /"  ^i--  v^"  ,.'•  /■  ,»"' 


-27 

-  18 

_ 

-10 

_ 

7 

-  10 

-20 

^24      - 

-  14 

-  14       - 

-     6 

-  26 

H 

-  IS 

- 

-   20 

- 

-   11 

-   15 

-  24 

-  26      - 

-27       - 

-  20      - 

-     7 

-  25 

- 

-  15 

-  18 

-  21 

-  26 

-  27 

-    26 

-  26      - 

-  15 

-  16 

-  23 

-  27 

- 

-  19 

- 

-25 

-  4 

-  22 

21 
24 

-     7 

8 

=  "l9 

--8 

_ 

-    4 

--21 

-  5 

4- 

-     5 

7 

'z': 

:'°i8 

-  23 

- 

-  16 

S  t  romal  e 

-  25 

Tetragonuridae 

Ariommidae 



■  27 

\      Nomeidae                                                             / 

-  3 

-  27 

Amarsipidae                      / 

14 

--16 
-  -  23 

Centrolophidae  / 

/         / 

_ 

_      1 

\                j  ^^ 

I 

-  2 

-  26 

'--.  / 

H 

aec 

rich' 

s    (1967) 

Tree 

Fig.  335.  Phylogenetic  hypothesis  of  relationships  among  the  genera  of  stromateoid  fishes.  The  perciform  famihes  Scorpididae,  Girellidae  and 
Kyphosidae  are  out-groups.  Each  number  represents  a  character  and  each  honzontal  bar  represents  a  character  state  indicating  a  synapomorphy 
or  autapomorphy.  Character  transformation  series  1-27  are  shown  in  Table  165.  Relationships  of  the  stromateoid  families  as  perceived  by 
Haedrich  (1967)  are  shown,  with  addition  of  the  Amarsipidae,  in  the  lower  right.  Limits  of  the  six  stromateoid  families  as  recognized  by  Haedrich 
and  Horn  (1972)  are  shown  at  the  top  of  the  cladogram. 


pididae,  Girellidae,  Kuhliidae,  Kyphosidae,  Pomatomidae, 
ScoiTJididae  and  Theraponidae.  All  are  Pattern- 10  teleosts  in 
terms  of  the  ramus  lateralis  accessorius  (a  facial  nerve  complex) 
and  have  a  bony  bridge  over  the  anterior  vertical  canal  of  the 
ear.  Of  these  families,  the  Kyphosidae  bear  a  strong  resemblance 
to  the  Centrolophidae,  the  most  primitive  stromateoids.  Hae- 
drich (1967)  implicitly  favored  the  hypothesis  that  the  kyphos- 
ids  are  the  closest  perciform  relatives  of  the  most  primitive 
stromateoids  pointing  out  that  both  have  10+15  vertebrae,  a 
caudal  skeleton  with  six  hypural  and  three  epural  elements  (ac- 
tually five  hypurals  in  kyphosids),  a  perforate  ceratohyal,  seven 
branchiostegals,  an  expanded  lacrimal  bone  and  scaled  fin  bases. 
The  present  analysis  (Fig.  335)  supports  this  hypothesis. 
Ontogenetic  characters  have  not  been  used  previously  in  ana- 


lyzing the  relationships  of  stromateoids  to  other  perciform  groups. 
Use  of  the  two  juvenile  characters  in  the  present  analysis  did 
not  alter  the  phylogeny  based  on  adult  characters.  Although  they 
were  not  examined  in  the  present  study,  kyphosid  larvae  arc 
more  generalized  than  girellid  or  scorpidid  larvae  and,  therefore, 
more  similar  to  stromateoid  larvae  (G.  D.  Johnson,  pers.  com- 
mun.).  The  generalized  nature  of  stromateoid  larvae  suggests 
that  their  characters  will  continue  to  be  most  useful  in  distin- 
guishing species  (e.g..  Ditty  and  Truesdale,  1983)  and  less  valu- 
able at  higher  levels  of  classification. 


Department  of  Biological  Science,  California  State 
University,  Fullerton,  California  92634. 


Gobiesociformes:  Development  and  Relationships 
L.  G.  Allen 


THE  Gobiesociformes  includes  three  families,  the  Gobiesoci- 
dae,  Callionymidae  and  Draconettidae,  according  to  Gos- 
line  (1970)  and  Nelson  (1976).  Members  of  this  order  are  pri- 
marily marine  bottom -dwellers  in  shallow-waters  and  occur 
worldwide  in  tropical  and  temperate  seas.  Distinguishing  char- 
acteristics of  the  order  include:  a  scaleless  head  and  body;  5-7 
branchiostegal  rays;  no  circumorbital  bones  behind  the  lacrimal; 
articular  processes  of  the  premaxillae  either  fused  with  ascending 
process  or  absent;  pelvic  fin  base  well  in  advance  of  pectoral 
fin;  no  swim  bladder  (in  adults)  (Nelson,  1976).  The  order  con- 
tains about  54  genera  with  246  species  in  the  three  families 
(Bnggs,  1955;  Nakabo,  1982a.  b). 

Briggs'  (1955)  review  of  the  Gobiesocidae  remains  as  the  most 
thorough  treatment  of  this  family  to  date.  Revisions  of  both  the 
Callionymidae  and  the  Draconettidae  have  recently  been  pub- 


lished by  Nakabo  (1982a,  b).  Hypotheses  of  systematic  rela- 
tionships within  the  families  are  based  entirely  on  adult  char- 
acteristics. 

The  Callionymidae  (dragonets)  is  a  large  and  diverse  group 
within  the  Gobiesociformes.  The  ontogeny  and  systematics  of 
the  Callionymidae  is  presented  in  this  volume  by  E.  D.  Houde. 

The  smallest  family  of  the  order,  the  Draconettidae,  consists 
of  two  genera  and  seven  species  (Nakabo,  1982a).  Draconettids 
are  small  demersal  fishes  inhabiting  sand-mud  bottoms  along 
the  edge  of  the  continental  shelf  or  on  seamounts.  They  occur 
widely  in  tropical  and  temperate  waters  of  the  world  except  the 
eastern  Pacific.  Adult  draconettids  resemble  callionymids  which 
lead  one  author  (Davis,  1 966)  to  include  the  draconettids  within 
the  Callionymidae.  Gosline  (1970)  and  Nakabo  (1982a)  dis- 
agreed with  this  inclusion. 


Table  166.    Egg  Characteristics  of  18  Species  of  Gobiesocids  for  which  Larvae  Are  Known  Organized  into  Subfamilies  after  Briggs 

(1955). 


Eggc 

laraclenstics 

Size  {mm) 

Number  eggs 

Oil 

Species/  Reference 

SublamiK 

Shape 

(long  axis) 

in  mass 

globules 

Color 

Where  laid 

Conidens  lalicephatus 

Trachelochisminae 

flat,  ellipsoid 

1.28-1.38 

77-109 

+ 

green- 

under  rock 

(Shiogaki  and  Dolsu.  197 Id) 

orange 

Trachclochismus  melobesia 

Trachelochisminae 

flat,  oval 

.?=  1.65 

3—300 

10-100 

red/pink 

under  rock 

(Ruck.  1971) 

Trachclochismus  pinnutalus 

Trachelochisminae 

oval 

.?=  1.81 

198-1,500 

1-6 

yellow- 

under  rock 

(Ruck,  1973b) 

red 

Lepadogaster  lepadogaster 

Lepadogastrinae 

flat,  oval 

1.8-1.9 

200-250 

1 

yellow/ 

under  rock 

(Guitel,  1888;  Russell,  1976) 

amber 

Lepadogaster  candolei 

Lepadogastrinae 

flat,  oval 

1.2 

— 

1 

yellow 

under  rock 

(Guitel,  1888;  Russell,  1976) 

Apletodon  rnicrocephalus 

Lepadogastrinae 

— 

— 

— 

1  to 

— 

kelp  stems 

(Guitel,  1888;  Russell,  1976) 

several 

Diptecogaster  biinaculata 

Lepadogastrinae 

flat,  oval 

1.37-1.54 

— 

1 

— 

empty  shells 

(Guitel,  1888;  Russell,  1976) 

, 

Diplocrcpis  pumccus 

Diplocrepinae 

spherical 

1.80 

<  2,400 

20-30 

purple 

under  rock 

(Ruck,  1973b) 

Gastroscvphus  hectoris 

Diplocrepinae 

— 

— 

— 

— 

— 

— 

(Ruck,  1976) 

Gastrocvalhus  gracilis 

Diplocrepinae 

— 

— 

— 

— 

— 

— 

(Ruck,  1976) 

Acrvlops  bervllmus 

Gobiesocinae 

oval 

1.1 

2-40 

2.5 

green- 

Thallasia 

(Gould,  1965) 

yellow 

blades 

Gobiesox  maeandricus 

Gobiesocinae 

oval 

1.68-1.92 

— 

— 

— 

under  rock 

(Allen  and  llg.  1983) 

Gobiesox  rhessodon 

Gobiesocinae 

oval 

— 

1 50-200 

— 

orange 

under  rock 

(Allen,  1979) 

Gobiesox  slrumosus 

Gobiesocinae 

elongate  oval 

0.75-0.94 

650-2,500 

70-80 

— 

empty  shells 

(Runyan.  1961;  Dovel,  1963) 

Rimicola  muscarum 

Gobiesocinae 

— 

— 

— 

— 

— 

kelp  blades 

(Allen,  1979) 

Lepadichlhys  frenatus 

Diademichthyinae 

flat,  ellipsoid 

1.31-1.36 

240-301 

1-6 

— 

shell 

(Shiogaki  and  Dotsu,  1971b,  c) 

Aspasma  minima 

Aspasminae 

ellipsoid 

1.25-1.35 

140-619 

20 

yellow 

under  rock 

(Shiogaki  and  Dotsu,  1971a) 

Aspasmichlhys  ciconiae 

Aspasminae 

— 

— 

— 

— 

— 

— 

(Shiogaki  and  Dotsu,  1972d) 

629 


630 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  167.     Larval  Characterlstics  of  18  Species  of  Gobiesocids  for  which  Larvae  are  Known  Organized  into  Slibfamilies  after  Briocss 
(1955).  See  Fig.  338  for  abbreviations  for  regions  of  pigmentation  (coding  of  pigment  patterns  depended  on  illustrations  in  most  species.  *  denotes 

counts  on  older  postflexion  larvae). 


Pigmentation  (left  side) 


Species/Reference 


Subfamily 

DH 

DT 

LT 

1X3 

Trachelochisminae 

5 

6-10 

3-7 

11-15 

Trachelochisminae 

0 

0 

0 

26-35 

Trachelochisminae 

0 

0 

0 

14 

Lepadogastrinae 

8 

23-24 

43 

11 

Lepadogastrinae 

4 

0 

5-6 

0 

Lepadogastrinae 

4 

15 

36 

13 

Lepadogastrinae 

0-4 

13-25 

32-62 

0 

Diplocrepinae 

0 

0 

30-50 

19-23 

Diplocrepmae 

0 

0 

2-3 

19-23 

Diplocrepinae 

0 

0-2 

7-10 

15-25 

Gobiesocinae 

? 

+ 

+ 

+ 

Gobiesocinae 

0 

0 

0 

9-15 

Gobiesocinae 

0-5 

0 

8-17 

13-15 

Gobiesocinae 

0-10 

10-15 

0-24' 

9-12 

Gobiesocinae 

7 

0 

10-50 

21 

Diademichthyinae 

5-12 

15-21 

70-105 

12-19 

Aspasminae 

0 

0*-13 

10-13 

7-17 

Aspasminae 

5 

16-17 

24-31 

6-7 

Conidens  lalicephalus  (Shiogaki  and  Dotsu,  197  Id) 

Trachelocfusmus  inelobesia  (Ruck.  1971) 

Trachelochismus  pinnulatus  (Ruck,  1973b) 

Lepadogasler  lepadogasier  (G\i\Xc\.  1888;  Russell,  1976) 

Lepadogasler  candolei  {Guitel,  1888;  Russell,  1976) 

Apletodon  microcephalus  (G\i\xe\.  1888;  Russell,  1976) 

Diplecogasler  bimaculala  (Guileh  1888;  Russell,  1976) 

Diplocrepis  pumceus  (Ruck,  1973b) 

Gaslroscyphus  hecloris  (Ruck,  1976) 

Gastrocyalhus  gracilis  (Ruck,  1976) 

Acrytops  beryllinus  (Gould,  1965) 

Gobiesox  maeandricus  (Allen  and  llg,  1983) 

Gobiesox  rhessodon  (Allen,  1979) 

Gobiesox  slrumosus  (Kuny&n,  1961;  Dovel,  1963) 

Rimicola  muscarum  (Allen,  1979) 

Lepadichthys  frenatus  (Shiogaki  and  Dolsu,  1971b,  c) 

Aspasma  minima  (Shiogaki  and  Dotsu,  1971a) 

Aspasmichlhys  ciconiae  (Shiogaki  and  Dotsu,  1972d) 


The  early  life  history  stages  of  draconettids  are  unknown  at 
this  time.  Therefore,  the  Draconettidae  will  receive  little  further 
attention  in  this  paper. 

The  Gobiesocidae  (clingfishes)  is  a  diverse  group  of  primarily 
shallow  water  or  intertidal  marine  (although  a  few  species  are 
freshwater)  fishes  consisting  of  about  33  genera  and  100  species. 
Clingfishes  occur  along  tropical  and  temperate  shores  in  the 
Atlantic,  Indian  and  Pacific  Oceans.  Distinguishing  character- 
istics of  gobiesocids  include:  pelvic  fins  modified  into  a  thoracic 
suction  disc;  pelvic  fin  with  one  small  modified  spine  and  four 
or  five  soft  rays;  single  dorsal  fin  without  spines;  no  basibran- 
chials;  vertebrae  25-54  (or  78  if  the  genus  Alabes  is  included, 
see  below);  lateral  line  confined  to  head;  two  postcleithra;  hy- 
purals  fused  into  a  single  plate.  Most  species  are  small  (normally 
<70  mm),  but  a  few  attain  relatively  large  size  (up  to  300  mm) 
(Nelson,  1976).  Eight  subfamilies  are  recognized  (Briggs,  1955). 
Springer  and  Fraser  (1976)  synonymized  the  family  Cheilo- 
branchidae  (=Alabetidae)  with  the  Gobiesocidae  based  on  shared 
specializations  particularly  of  the  structure  of  the  joint  between 
the  supracleithrum  and  cleithrum.  If  valid,  this  synonymy  adds 
one  more  genus  (Alabes)  and  four  species  to  the  Gobiesocidae. 


Development 

Eggs 

Spawning  occurs  in  rocky  intertidal  or  subtidal  areas.  Eggs 
are  demersal  and  are  attached  to  the  underside  of  rocks  or  shells 
or  on  kelp  blades.  The  adults  (usually  the  male)  guard  the  eggs 
during  development.  Eggs  are  ovate  to  ellipsoidal  in  shape  and 
range  from  about  0.7  to  1.9  mm  in  greatest  dimension  (Table 
166).  The  monolayered  egg  masses  usually  contain  between  100 
and  600  eggs  with  some  reports  of  up  to  2,500  eggs  in  a  patch 
(Gobiesox  strumosus)  (Table  166).  The  initial  coloration  of  the 
eggs  ranges  between  purple  and  green  (with  pink,  yellow  and 
orange  predominating).  Eggs  contain  anywhere  from  one  to  100 


oil  globules  depending  on  species  and  stage  of  development.  As 
development  proceeds,  oil  globules  tend  to  coalesce  ultimately 
into  one. 

Larvae 

Larvae  of  18  species  of  gobiesocids  have  been  described  in 
varying  detail  (Table  167).  Larval  series  are  available  for  10  of 
these  species.  Larvae  are  well  developed  at  hatching  and  possess 
functional  jaws,  fully  pigmented  eyes,  body  pigment  similar  to 
that  of  later  larval  stages  and  a  small  (sometimes  bilobed)  yolk 
sac  in  most  species  which  is  probably  absorbed  soon  after  hatch- 
ing. Size  at  hatching  ranges  from  2.4  mm  in  Gobiesox  strumosus 
to  6.8  mm  in  G.  maeandricus  and  appears  to  be  related  to  egg 
size  and  maximum  size  of  the  adults  (Table  168).  Larvae  are 
cylindrical  and  somewhat  laterally  compressed  becoming  more 
robust  with  growth.  All  clingfish  larvae  have  long,  underslung 
guts  which  usually  extend  beyond  the  midline  of  the  body  (preanal 
length  50-70%  SL)  in  both  pre-  and  postflexion  larvae.  Size  at 
notochord  flexion  was  difficult  to  determine  from  most  of  the 
larval  descriptions,  but  generally  ranged  between  5.0  and  8.0 
mm  depending  on  the  species  (Table  167).  Gobiesocid  larvae 
have  well-developed  swimbladders  in  the  early  stages  of  devel- 
opment which  are  located  in  the  dorso-anterior  portion  of  the 
peritoneal  cavity.  In  several  species  the  swimbladder  is  hidden 
by  the  heavy  pigmentation  on  the  dorsum  of  the  gut.  No  sudden 
change  occurs  at  settlement;  rather,  larvae  gradually  attain  ju- 
venile characteristics.  Size  at  settlement  is,  therefore,  difficult 
to  determine.  Juvenile  characteristics  are  attained  at  a  wide 
range  of  sizes  (6.3-13.0  mm)  with  a  trend  toward  larger  species 
"settling"  at  larger  sizes  (Tables  167  and  168).  Presumably,  the 
loss  of  the  swimbladder  occurs  during  settlement. 

Most  gobiesocid  larvae  are  heavily  pigmented.  Furthermore, 
the  numbers  and  patterns  of  the  large,  stellate  melanophores  on 
the  body  are  species  specific,  and  are  invaluable  in  the  identi- 
fication of  species  (Figs.  336  and  337).  Melanophores  occur 
primarily  in  the  seven  regions  designated  in  Fig.  338.  Larvae  of 


ALLEN:  GOBIESOCIFORMES 
Table  167.    Extended. 


631 


Pigmentation  (left  side) 

Myomere 
counl 

Size  (mtn)  at 

_           Preanal 

lenglh/SL  (%) 

Hatching 

Notochord 
flexion 

Onset  of  Pj 
develop 

Settling 

LG 

VG 

PV 

Preflexion 

Postflexion 

0 

11-14 

4-5 

26-31 

3.4 

4.5-5.5 

4.5 

6.3 

61.4 

72.4 

0 

0 

5-9 

31-36 

4.8-5.5 

6.3-7.0 

6.5 

7.8? 

62.9 

64.9 

0 

0 

0 

31 

5.3-6.1 

_ 

— 



59.6 



16 

13 

12 

— 

5.1 

— 

_ 

_ 

62.1 

_ 

12 

1 

9 

— 

5.0-6.0 

— 

6.5? 

— 

57.2 

— 

20 

0 

2 

— 

— 

_ 

_ 

7.0? 

59.9 



15-20 

19 

0 

— 

3.0-4.3 

4.5-6.5 

8.0 

_ 

63.5 

72.4 

9-26 

0 

5-6 

— 

5.0-6.0 

6.2-7.7 

7.7 



65.7 

65.0 

0 

3 

12-16 

— 

— 

— 

_ 

_ 

61.8 

60.2 

4-7 

4 

0-1 

— 

— 

5.5-6.9 

— 

— 

60.7 

72.3 

+ 

+ 

9 

_ 

2.6 











0 

0 

9-16 

32-35 

5.6-6.8 

6.7-7.0 

7.3 

13.0 

50.0 

63.6 

0 

0 

4-7 

24-29 

3.9-4.1 

5.5-6.9 

5.5 

-10.0 

58.0 

68.0 

0 

0-1 

4-8 

28-29 

2.4-3.4 

4.7-6.5 

6.6 

-12.0 

61.4 

72.9 

0 

0 

0 

— 

4.0 

— 

— 



55.3 



0-4 

0 

9-14 

36-37 

4.2 

6.0 

6.0 

9.9 

58.8 

79.0 

0-12 

2-4 

6-12 

35 

3.6-4.0 

5.5-6.8 

4.5 

6.7-7.4 

67.3 

76.7 

6-21 

14-15 

0 

— 

- 

- 

- 

- 

62.7 

75.5 

each  species  exhibit  a  unique  distribution  of  melanophores  with- 
in and  among  these  regions  (Table  167).  The  distribution  of 
melanophores  within  regions  can  be  coded.  For  example,  Con- 
idcns  laticephalus  has  a  pigment  pattern  which  can  be  designated 
as  the  followmg:  DH5,  DT6-10,  LT3-7,  DGl  1-15.  LGO,  VGl  1- 
14,  PV4-5.  Trachelochismus  melohesia  by  the  same  process  is 
designated  as:  DHO,  DTO,  LTO,  DG26-35,  LGO,  VGO,  DV5- 
9.  If  adopted,  this  system  of  coding  pigment  patterns  will  serve 
two  purposes.  It  will  greatly  aid  identification  of  clingfish  larvae 
and  will  also  lead  to  more  comparable  descriptions  of  gobiesocid 
larvae  in  the  future. 

Pigment  patterns  do  not  appear  to  be  related  to  phylogenetic 
hypotheses  based  on  adult  characteristics.  In  virtually  all  known 
cases  closely  related  species  (subfamilial  levels)  tend  to  have 
noticeably  different  patterns  and  often  range  from  heavily  to 
lightly  pigmented  (Table  167).  Within  the  Trachelochisminae, 
Conidcns  laticephalus  is  heavily  pigmented  while  both  Tra- 
chelochismus melohesia  and  T.  pinnulala  are  lightly  pigmented. 
The  same  pattern  is  exhibited  in  all  other  subfamilies  (Table 
167)  especially  the  Lepadogastrinae  and  Gobiesocinae  (partic- 
ularly in  the  genus  Gobiesox).  Members  of  subfamilies  often 
overlap  in  their  distributions  (Briggs,  1955).  Diverse  pigment 
patterns  among  closely  related,  sympatric  clingfish  larvae  may 
well  represent  ecotypic  variation.  Heavily  pigmented  larvae  often 
live  in  surface  waters  where  the  pigmentation  may  protect  them 
against  solar  radiation  or  serve  as  protective  coloration  (Moser, 
1 98 1 ).  Less  pigmentation  may  indicate  that  the  larvae  normally 
occur  deeper  in  the  water  column  where  irradiance  does  not 
present  problems  for  development. 

Only  a  few  published  descriptions  included  myomere  counts. 
Those  accounts  which  did  revealed  a  range  from  24  up  to  37 
(Table  167).  The  number  of  myomeres  appears  to  have  great 
diagnostic  value  in  some  cases  when  used  in  conjunction  with 
pigment,  i.e.,  among  the  species  of  Gobiesox  (Table  167).  The 
lack  of  myomere  count  data  among  the  described  gobiesocid 
larvae  may,  in  part,  be  due  to  the  difficulty  in  countmg  caused 


by  heavy  trunk  pigmentation.  Nonetheless,  it  is  unfortunate  that 
this  important  character  has  not  received  greater  attention  es- 
pecially since  vertebral  counts  are  not  available  for  many  species. 
Adult  characteristics  which  are  valuable  for  identifying  older 
larvae  are  also  included  in  Table  168. 

The  most  distinctive  characteristic  of  clingfishes  is  the  suction 
disc  which  is  supported  by  the  pelvic  fins  and  distal  postcleithra 
of  the  pectoral  girdle.  The  two  types  of  discs  are  found  in  go- 
biesocids.  The  "double"  disc  has  a  small,  posterior  disc  with  a 
free  anterior  margin  separating  it  from  an  anterior  disc.  In  the 
"single"  disc  the  anterior  and  posterior  portions  are  coalesced 
into  one  continuous  structure  (Briggs,  1955).  The  onset  of  suc- 
tion disc  development  occurs  fairly  early  in  larval  development 
(ranges  from  4.5  to  8.0  mm  SL)  and  appears  to  be  closely  allied 
to  time  of  notochord  flexion  in  most  species  (Table  167).  Disc 
development  does  not  appear  to  differ  appreciably  between  sin- 
gle and  double  disc  types  except  that  in  the  single  type  a  con- 
sistent connection  remains  between  the  anterior  and  posterior 
elements  throughout  development  (Fig.  339).  The  completion 
of  the  suction  disc  is  undoubtedly  critical  in  late  larval  stages. 
Settlement  seems  unlikely  to  occur  without  a  functional  disc. 

Specialized  glandular  tissues  appear  on  the  body  surface  and 
out  onto  the  finfolds  in  several  species  of  gobiesocids  (Shiogaki 
and  Dotsu,  1971b;  Allen  and  Ilg,  1983).  Although  these  struc- 
tures are  not  specifically  mentioned  in  other  descriptions,  illus- 
trations of  larvae  from  some  of  these  studies  include  structures 
in  the  finfolds  which  may  be  these  same  glandular  tissues.  Fur- 
ther studies  are  needed  to  ascertain  the  extent  of  this  special- 
ization within  the  Gobiesocidae  and  the  possible  function  of 
these  tissues. 

Relationships 

The  systematic  relationships  among  the  Gobiesocidae  were 
addressed,  as  previously  mentioned,  by  Briggs  (1955).  His  eight 
subfamilies  reflected  both  morphological  similarities  and  zoo- 
geographic  distributions  (subfamilies  occupy  fairly  distinct  re- 


B 


ALLEN:  GOBIESOCIFORMES  633 

Table  168.     Adult  Characteristics  for  18  Species  of  Gobiesocids  for  which  Larvae  are  Known  Arranged  by  Subfamily  (Briggs,  1955). 


No. 

No. 

No 

No. 

Max. 

dorsal 

anal 

pectoral 

caudal 

No. 

length 

Species/Reference 

Subfamily 

rays 

rays 

rays 

rays 

vertebrae 

(mm  SL) 

General  distribution 

Comdens  taticephalus 

Trachelochisminae 

7-9 

5-7 

19-20 

11-13 

28 

33.0 

Southern 

(Shiogaki  and  Dotsu,  197 Id) 

Japan 

Tracheloch is mus  melobesia 

Trachelochisminae 

9-11 

7-8 

22-24 

12 

— 

30.0 

New  Zealand 

(Ruck,  1971) 

Trachetochismus  pmnulalus 

Trachelochisminae 

7-9 

5-7 

24-26 

11-12 

_ 

71.2 

New  Zealand 

(Ruck,  1973b) 

Lepadogaster  lepadogaster 

Lepadogastrinae 

16-19 

9-11 

20-23 

12-13 

— 

65.0 

NE  Atlantic/ 

(Guild,  1888;  Russell,  1976) 

Mediterranean 

Lepadogaster  candolei 

Lepadogastrinae 

13-16 

9-11 

26-29 

10-13 

— 

75.0 

NE  Atlantic/ 

(Guitel,  1888;  Russell.  1976) 

Mediterranean 

Aptetodon  microcephalus 

Lepadogastrinae 

5-6 

5-7 

21-24 

10-11 

— 

41.6 

NE  Atlantic/ 

(Guitel,  1888;  Russell,  1976) 

Mediterranean 

Dtplecogaster  himaculata 

Lepadogastrinae 

5-7 

4-6 

21-25 

9-10 

— 

41.0 

NE  Atlantic/ 

(Guitel,  1888;  Russell,  1976) 

Mediterranean 

Diplocrepis  pumceus 

Diplocrepinae 

10-11 

4-5 

23-24 

10 

— 

100.0 

New  Zealand 

(Ruck,  1973b) 

Gaslroscyphus  hectons 

Diplocrepinae 

6-8 

6-7 

20-22 

8 

— 

43.6 

New  Zealand 

(Ruck,  1976) 

Gastrocvathus  gracilis 

Diplocrepinae 

5-6 

5-7 

18-19 

8 

_ 

31.0? 

New  Zealand 

(Ruck,  1976) 

Acrylops  hervtlinus 

Gobiesocinae 

5-7 

5-7 

19-23 

10 

— 

20.0 

NW  Atlantic 

(Gould,  1965) 

Gobiesox  maeandncus 

Gobiesocinae 

14-16 

13-15 

21-23 

11-13 

32-34 

114.0 

NE  Pacific 

(Allen  and  llg,  1983) 

Gobiesox  rhessodon 

Gobiesocinae 

12-14 

11-12 

18-21 

11-12 

28-29 

39.3 

NE  Pacific 

(Allen,  1979) 

Gobiesox  slrumosus 

Gobiesocinae 

10-13 

9-11 

22-26 

11-13 

25-27 

69.3 

NW  Atlantic 

(Runyan,  1961;  Dovel,  1963) 

Rimicola  muscarum 

Gobiesocinae 

6-8 

6-8 

14-16 

8 

35-36 

53.2 

NE  Pacific 

(Allen,  1979) 

Lepadulilhys  frenatus 

Diademichthyinae 

15-17 

12-15 

25-31 

11 

— 

52.5 

W  Pacific 

(Shiogaki  and  Dotsu,  1971b,  c) 

Aspasma  minima 

Aspasminae 

7-9 

6-9 

21-24 

8-9 

— 

52.3 

NW  Pacific 

(Shiogaki  and  Dotsu,  1971a) 

Aspasmichthys  cicomae 

Aspasminae 

11-13 

8-9 

23 

10-11 

— 

56.0 

NW  Pacific 

(Shiogaki  and  Dotsu,  I972d) 

gions  of  the  world).  The  relationships  among  the  subfamilies 
were  based  pnmaiily  on  four  characters:  the  number  of  gill 
arches;  gill  membrane  state;  type  of  suction  disc  and  dentition 
type. 

In  my  opinion,  the  evolutionary  scheme  presented  in  Briggs 
(1955)  is  in  drastic  need  of  revision  from  a  cladistic  viewpoint. 
The  independent  derivation  of  the  single  suction  disc  and  mixed 
derived  character  states  in  several  divergent  evolutionary  lines, 
plus  the  "evolution"  of  one  subfamily  from  another  through 
primitive  and  derived  genera  are  particularly  troubling  aspects 
of  his  analysis. 

Gosline  (1970)  was  first  to  include  the  Callionymidae,  Dra- 
conettidae  and  Gobiesocidae  in  the  order  Gobiesociformes.  Ac- 
cording to  Gosline  (1970)  the  three  families  share  a  number  of 
characteristics  including  a  scaleless  head  and  body,  no  circum- 
orbital  bones  behind  lacrimal,  articular  processes  of  the  pre- 


maxillae,  as  well  as  others  (see  Gosline,  1970:  365  and  377). 
These  similarities  coupled  with  evidence  that,  in  Gosline's  words, 
"the  Gobiesocidae  has  evolved  from  the  notothenoid  section  of 
the  perciform  suborder  Blennioidei  and  in  small  part  at  least 
over  the  same  route  as  the  draconettids  and  callionymids"  form 
the  basis  for  including  all  three  families  in  the  order  Gobiesoc- 
iformes. 

Greenwood  et  al,  (1966)  placed  the  Gobiesociformes  which 
included  only  the  Gobiesocidae  mto  the  superorder  Paracan- 
thopterygii  in  their  provisional  classification  of  teleostean  fishes. 
Apparently  this  placement  was  based  on  a  relationship  between 
batrachoidids  and  gobiesocids  proposed  by  Briggs  (1955)  and 
McAllister  (1968),  although  Briggs  did  note  some  resemblance 
between  the  Gobiesocidae  and  the  Callionymoidea.  Gosline 
(1970)  believed  that  characteristics  held  in  common  by  gobie- 
socoid  and  batrachoid  fishes  (e.g.,  the  usually  scaleless  body. 


Fig.  336.  Representative  larvae  of  seven  genera  within  the  Gobiesocidae:  (A)  Comdens  taticephalus.  5.5  mm  (from  Shiogaki  and  Dotsu, 
1971d);  (B)  Trachelochismus  melobesia.  7.8  mm  (after  Ruck,  1971);  (C)  Lepadogaster  lepadogaster.  6.0  mm  (after  Russell,  1976);  (D)  Apletodon 
microcephalus.  4.5  mm  (after  Russell,  1976);  (E)  Dtplecogaster  bimaculata.  6.5  mm  (after  Russell,  1976);  (F)  Diplocrepis  puniceus.  7.7  mm  (after 
Ruck,  1973b);  and  (G)  Gastrocyathus  gracilis.  6.9  mm  (after  Ruck,  1976). 


634 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


B 


ALLEN:  GOBIESOCIFORMES 


635 


Dorsal   Trunk   (DT) 


Dorsal    Head  (DH) 


Lateral  Trunk   (LT) 


Postanal  Ventral  (PV) 

Dorsal  Gut  (DG) 
Lateral   Gut  (LG) 

Fig.  338.     Hypothetical  clingfish  larva  showing  regions  which  form  the  basis  for  coding  patterns  of  melanophores. 


Ventral  Gut   (VG) 


flattened  head,  anterior  pel  vies,  incomplete  circumorbital  series) 
are  the  result  of  convergence.  Gosline  (1970)  then  cited  five 
morphological  and  osteological  features  which  differ  between 
the  Gobiesociformes  (including  the  three  families)  and  the  Ba- 
trachoidiformes.  These  features  include  differences  in  pelvic  fin 
structure  and  orientation,  structure  of  the  upper  hypurals,  as- 
cending process  of  premaxilla,  ossification  of  the  median  eth- 
moid and  presence  (Batrachoidiformes)  or  absence  (Gobiesoc- 
iformes) of  a  swimbladder.  On  the  other  hand,  he  upheld  that 
gobiesociform  (three  families)  fishes  have  almost  all  of  the  di- 
agnostic characteristics  of  the  superfamily  Notothenioidca  of 
the  perciform  suborder  Blennioidea  (see  Gosline,  1968).  He 
further  pointed  out  structural  similarities  between  members  of 
the  Gobiesociformes  and  certain  genera  of  notothenioid  fishes 
as  evidence  supporting  this  proposed  relationship.  Based  on  this 
work  on  adults,  gobiesociform  fishes  are  currently  considered 
perciform  derivatives  in  the  superorder  Acanthopterygii.  How- 
ever, the  issue  remains  far  from  resolved  and  future  investi- 
gations into  both  the  ordinal  and  superordinal  relationships  are 
still  very  much  in  order.  In  fact,  William  Eschmeyer  (California 
Academy  of  Sciences)  is  currently  investigating  possible  rela- 
tionships between  gobiesociform  (particularly  gobiesocids)  and 
scorpaeniform  fishes  (pers.  comm.). 


The  early  life  history  stages  of  gobiesocids  and  callionymids 
(see  Houde,  this  volume)  lend  little  support  to  Gosline's  clas- 
sification. Gobiesocid  and  callionymid  larvae  are  usually  pig- 
mented heavily,  but  there  are  very  few  additional  similarities 
at  the  current  level  of  examination.  Gobiesocid  and  callionymid 
early  life  history  stages  differ  in:  egg  type  (demersal  versus  pe- 
lagic eggs,  respectively),  preanal  length  (>50%  versus  <50%  of 
standard  length),  general  body  shape  (relatively  large  cylindrical 
versus  small,  laterally  compressed  larvae),  myomere/vertebral 
counts  (24  to  37  versus  19  to  23),  and  shape  of  the  notochord 
tip  (no  extension  versus  a  long  extension  beyond  the  hypural 
plate).  These  basic  differences  may,  in  part,  represent  divergence 
due  to  dissimilar  reproductive  strategies.  A  more  thorough,  de- 
tailed comparison  of  the  early  life  history  stages  (larvae  in  par- 
ticular) will  be  necessary  before  any  solid  conclusions  can  be 
drawn.  Unfortunately,  the  eggs  and  larvae  of  draconettids  (pre- 
sumably the  most  primitive  members  of  the  order)  are  unknown 
and  cannot  help  clarify  the  situation. 

The  use  of  larval  characteristics  to  assess  higher  level  rela- 
tionships between  the  Gobiesociformes  and  the  Batrachoidi- 
formes or  Notothenioidea  is  limited  since  batrachoids  have  di- 
rect development  (no  larval  form)  and  the  larvae  of  notothenioids 
bear  little,  general  resemblance  to  gobiesocid  and  callionymid 


Fig.  337,  Representative  larvae  of  six  genera  within  the  Gobiesocidae:  (A)  Gastroscyphus  hectoris.  5.4  mm  (after  Ruck.  1976);  (B)  Gobiesox 
rhessodnn.  6.2  mm  (from  Allen,  1979);  (C)  Rimuola  miiscarum.  4.0  mm  (from  Allen,  1979);  (D)  Lepadichthys frenalus,  7.3  mm  (from  Shiogaki 
and  Dotsu,  in  prep.);  (E)  Aspasma  minima,  6.8  mm  (from  Shiogaki  and  Dotsu,  1971a);  and  (F)  Aspasmichlhys  ciconiae.  6.9  mm  (from  Shiogaki 
and  Dotsu,  1972d). 


636 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Aspasma  minima 
(double  disc) 


CO 


.8mm 


;  V    6.9mm 


Lepadichthvs   frenatus 
(single  disc) 


7.3mm 


larvae.  Detailed  studies  with  the  larval  (or  embryonic)  forms  of 
the  above  mentioned  groups  should  be  fruitful  in  leading,  I 
believe,  to  a  clearer  understanding  of  their  relationships. 

Future  investigation  into  gobiesociform  systematics  should 
first  concentrate  on  whether  the  current  gobiesociformes  rep- 
resents a  monophyletic  grouping.  Only  when  this  question  is 
answered  satisfactorily  can  the  higher  order  relationships  be 
addressed. 

Department   of   Biology,    California    State    University, 

NORTHRIDGE,   NORTHRIDGE,  CALIFORNIA  91330. 


8.3mm 


9. 1  mm 


8.2mm 


¥\  10. Or 


10.0mm 


Fig.  339.     Development  sequences  of  the  pelvic  suction  disc  in  larval 
gobiesocids  (double  and  single  types). 


Callionymidae:  Development  and  Relationships 
E.  D.  HouDE 


THE  Callionymidae  are  one  of  three  families  in  the  order 
Gobiesociformes  (Gosline,  1970;  Nelson,  1976).  System- 
atics,  ontogeny  and  relationships  of  Callionymidae  and  the  other 
families,  Gobiesocidae  and  Draconettidae,  have  been  reviewed 
and  summarized  for  this  symposium  by  Allen  (this  volume). 
The  callionymids  are  small  demersal  fishes  found  in  all  warm 
seas.  Most  species  are  less  than  100  mm  in  length.  Maximum 
length  is  about  200  mm  (Nelson,  1976;  Nakabo,  1 982b).  Nelson 
( 1976)  indicated  that  there  are  eight  genera  with  about  40  species 
in  the  family.  Fncke  (1980,  1981a,  1981b)  believed  the  family 
to  be  more  diverse  with  perhaps  130  species  worldwide,  75  in 
the  genus  Callionymus.  and  Nakabo  (1982b)  recently  has  pro- 
posed 19  genera  and  139  species.  Callionymids  are  most  abun- 
dant and  diverse  in  shallow  marine  waters  of  the  Indo-Pacific 
(Smith,  1963;  Fricke,  1980,  1981b).  They  also  are  common  in 
the  Atlantic  (Davis,  1966).  Although  usually  found  in  depths 
less  than  1 00  m,  some  species  occur  to  depths  of  >  600  m  (Davis, 


1 966).  Larvae  sometimes  are  abundant  in  coastal  ichthyoplank- 
ton  surveys.  For  example,  Callionymus  pauciradiatus  was  the 
second  most  common  species  of  larva  in  Biscayne  Bay,  Florida 
(Houde  and  Alpem  Lovdal,  in  press)  while  Callionymidae  were 
the  sixth  most  abundant  family  of  larvae  in  Persian  Gulf  fish 
larvae  collections.' 

Nakabo  (1982b)  has  extensively  revised  the  Callionymidae, 
establishing  7  new  genera  and  redefining  12  previously  recog- 
nized genera/subgenera.  Genera  are  defined  based  on  cephalic 
lateral  lines,  lateral  lines  on  the  body,  morphology,  secondary 


'  Houde,  E.  D.,  J.  C.  Leak,  S.  Al-Matar  and  C.  E.  Dowd.  1981. 
Ichthyoplankton  abundance  and  diversity  in  the  western  Arabian  Gulf. 
Kuwait  Institute  for  Scientific  Research.  Mariculture  and  Fishenes  De- 
partment. Final  Report,  Project  MB- 16,  3  volumes.  (This  report  was 
not  available  for  distnbution  at  the  time  the  present  paper  was  written.) 


Table  169. 

Meristics 

OF  Calliony 

MID  Genera 

Recognized 

>  BY  Nakabo  (198: 

2b).  His  new  gene 

ra  are  designated  n.g. 

Genus 

Dorsal 

Anal 

Pectoral 

Pelvic 

Caudal 

Type  species  by  onginal  designation 

Callionymus 
Balhycallionymus  n.g. 

in 

-IV, 
IV, 

6- 
9 

•10 

9 
9 

i 
11 

+  16-20 
+  17-19 

I, 
I, 

5 
5 

i  +  7 
i  +  3 

+  ii 

+  ii  +  2  +  ii 

Callionymus  lyra 
Callionymus  kaianus 

Foetorepus 
Eocallionymus  n.g. 
Paracallionyinus 
Neosyncluropus  n.g. 

IV, 
IV, 

IV, 
IV, 

8 

7 
9- 

8 

■10 

7 
6 
9 

7 

i 
i 
i 
i 

+  18-21 
+  18-19 
+  18-19 

+  17-21 

I, 
1, 
I, 
1, 

5 
5 
5 
5 

i  + 
i  + 

i  + 
i  + 
1  + 

3  +  i  +  3  +  ii 
7  -1-  ii 
7  +  ii 
7  +  11 
7  +  11 

Callionymus  calauropomus 
Callionymus  papitio 
Callionymus  coslatus 
Callionymus  ocellalus 

Pterosynchiropus  n.g. 
Minysynchiropus  n.g. 

IV, 

IV, 

8 
9 

7 
8 

ii 

iii 
ii 

+  17-19 

or 
+  18 
+  29-30 
+  16 

or 
+  14-15 

or 
+  13-14 
+  15-20 

I, 
I, 

5 
5 

i  + 
1  + 

8 

7 

+  i 
+  ii 

Callionymus  splendidus 
Synchiropus  laddi 

Paradiplngrammus  n.g. 

IV. 

8- 

-9 

7-8 

iii 

iv 
i 

I, 

5 

i  + 

7 

+  ii 

Callionymus  enneaciis 

Diplogrammus 

IV, 

6- 

■8 

4-7 

ii 
i 

+  14-15 
+  16-18 

or 
+  15-17 
+  18 
+  17-18 
+  17 
+  16-19 
+  17-18 

I, 

5 

i  + 

7 

+  ii 

Callionymus  goramensis 

Synchiropus 
Orhonymus 
Dactylopus 
Calhunchlhys 
Pseudocalhunchlhys  n.g. 

IV, 
IV, 
IV, 
IV, 
IV, 

8 
8 
8 
9 
8 

7 
7 
7 
8 
7 

ii 
i 
i 

ii 

II 

1 

I, 
1, 
1. 
1. 
I, 

5 

5 

1-4 

5 

5 

i  + 
i  + 
i  + 
i  + 
i  + 

7 
7 
7 

7 
7 

+  ii 

+  ii 
+  ii 
+  ii 
+  ii 

Callionymus  lateralis 
Callionymus  rameus 
Callionymus  daclylopus 
Callionymus  japonicus 
Callionymus  variegatus 

Repomucenus 

Spinicapilichlhys 

Anaora 

Eleulhcrnchir 

III 
1 

-IV,  9 
IV,  8 
IV,  8 

absent 
or 

-IV.  9- 

■13 

8-9 

8 

7 
9-13 

ii 

i 
i 

1 

+   14-17 
+  16-21 
+   18-20 
21-25 
+  16-23 

I, 
I. 
I, 
I. 

5 
5 
5 
5 

i  -t-  7  +  ii 
i  -f  7  +  ii 
ii  +  6  +  ii 
i  +  7  +  ii 

or  li  +  6  +  ii 

Callionymus  calcaralus 
Callionymus  spiniceps 
Anaora  lenlaculata 
Callionymus  opercularoides 

637 


638 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


HOUDE:  CALLIONYMIDAE 


639 


sex  characteristics  and  body  size.  Meristics  also  vary  among 
species  (Table  169).  In  Nakabo's  classification  the  genus  Cal- 
lionvimis  includes  only  five  species,  all  of  which  are  fiaund  in 
the  northeast  Atlantic,  Mediterranean  or  Black  seas.  Nakabo 
(1982b)  assigned  39  species  to  the  Indo-Pacific  genus  Repo- 
mucemts.  making  it  the  most  species-nch  genus  of  Callionym- 
idae,  followed  by  the  Indo-Pacific  genus  Calliurichthys  with  16 
species.  Neither  eggs  nor  larvae  were  described  or  discussed  by 
Nakabo  m  his  systematic  account  of  the  Callionymidae. 

Description 

The  Callionymidae  are  characterized  by  having  a  small,  pore- 
like gill  opening,  greatly  reduced  in  size  compared  to  that  of 
Draconettidae,  their  closest  relatives  in  the  Gobiesociformes. 
The  preoperculum  has  a  strong,  often  serrate,  spine,  useful  for 
specific  identifications;  the  operculum  and  suboperculum  are 
spineless.  Eyes  are  dorsal  and  adjacent.  Hypurals  are  fused  into 
a  single  plate.  Vertebrae  number  21  to  23.  Dorsal  fin  spines 
usually  four;  soft  rays  6-13;  anal  fin  with  4-13  soft  rays.  Pelvic 
fins  are  inserted  in  advance  of  the  pectoral  base,  the  two  fins 
often  connected  at  their  bases  by  a  membrane.  The  sexes  usually 
are  dimorphic,  males  having  longer  and  broader  median  fins, 
sometimes  with  filamentous  rays  in  the  dorsal  and  caudal  fins. 

Development 
Size  at  maturity  varies  among  species  but  generally  is  less 
than  100  mm.  Some  species  may  mature  at  <  15  mm  in  length 
(Davis,  1966).  Callionymid  male-female  pairs  exhibit  well-de- 
fined courtship  and  spawning  behavior  (Wilson,  1978;  Takita 
and  Okamoto,  1979)  m  which  male  display  plays  a  prominent 
role.  Individual  females  may  spawn  on  successive  days.  Judging 
from  larval  occurrences,  spawning  seasons  are  protracted,  last- 
ing 6  months  or  more  for  temperate  species  such  as  C.  lyra.  C. 
maculatus  and  C.  reticulatus  (Demir,  1972;  Russell,  1976). 
Spawning  may  occur  year-round  in  subtropical  species  such  as 
C.  pauciradiatus  (Houde  and  Alpem  Lovdal,  in  press)  and  Par- 
acallionymus  coslatus  (Brownell,  1979)  or  tropical  species  such 
as  C.  decoratus  (Watson  and  Leis,  1974). 

Eggs 

Eggs  are  colorless,  pelagic  and  spherical,  reported  diameters 
ranging  from  0.55  to  0.97  mm  (Mito,  1962a;  Watson  and  Leis, 
1974;  Russell,  1976;  Brownell,  1979;  Miller  et  al.,  1979;  Takai 
and  Yoshioka.  1979;  Takita,  1980,  1983).  A  polygonal  (usually 
hexagonal)  sculpturing,  sometimes  with  fine  cilia-like  processes, 
usually  is  associated  with  the  chorion,  but  in  some  species  (e.g., 
P.  costatus)  the  chorion  apparently  is  unsculptured  (Brownell, 
1979).  Buoyant,  adhesive  egg  masses  have  been  described  for 
C.  calliste.  which  break  up  into  individual  pelagic  eggs  prior 
to  hatching  (Takita,  1983).  The  yolk  is  segmented  peripherally. 
The  perivitelline  space  is  narrow.  There  are  no  oil  globules. 
Takai  and  Yoshioka  (1979)  and  Takita  (1980)  have  provided 
good  illustrations  and  photographs  of  typical  callionymid  eggs. 


Larvae 

At  hatching,  pelagic  larvae  of  callionymids  range  from  ap- 
proximately 1 .0  to  2. 1  mm  in  length.  Most  species  are  less  than 
1.5  mm  at  hatching,  making  them  among  the  smallest  of  larval 
fishes.  Reported  myomere  numbers  range  from  19-22.  Callio- 
nymid larvae  are  distinctive  and  easy  to  recognize.  Larvae  of 
several  species  (referred  to  as  Callionymus)  have  been  described 
(e.g..  Page,  1918;  Mito,  1962a;  Demir,  1972,  1976;  Miller  et  al., 
1979;  Takai  and  Yoshioka,  1979;  Takita,  1980,  1983).  Brownell 

(1979)  has  illustrated  larvae  of  Paracallionymus  coslatus.  All 
larvae  described  to  date  are  similar,  differing  in  pigmentation 
patterns,  meristic  characters  and  sizes  at  which  fin  development 
and  metamorphosis  are  completed. 

Yolk-sac  larvae  are  short  and  deep-bodied  with  a  large,  bul- 
bous yolk  sac  (Mito,  1962a;  Brownell,  1979;  Takita,  1980,  1983). 
The  yolk  is  segmented  peripherally.  Dendritic  or  stellate  me- 
lanophores  may  develop  in  the  finfold  (Fig.  340B)  within  one 
day  after  hatching  (Mito,  1962a;  Brownell,  1979;  Takai  and 
Yoshioka,  1979;  Takita,  1980,  1983).  The  snout-to-anus  length 
of  newly-hatched  larvae  is  >50%  of  notochord  length,  but  it 
declines  to  <50%  within  several  hours  after  hatching. 

Preflexion  larvae  are  moderately  deep-bodied  and  laterally 
compressed  both  preanally  and  postanally.  All  species  described 
to  date  have  a  broken  line  of  melanophores  along  the  lateral 
midline,  particulariy  on  the  tail  (Fig.  340).  The  larvae  are  mod- 
erately to  heavily  pigmented  and  often  are  first  recognized  in 
samples  because  of  their  relatively  dark  color.  A  swimbladder 
which  develops  at  this  stage  subsequently  is  lost  during  meta- 
morphosis. Curious  processes,  termed  "spine-like"  by  Takita 
( 1 980,  1 983)  or  called  "serrations"  by  Mito  ( 1 962a)  develop  at 
the  margins  of  the  dorsal  and  ventral  finfolds  (Fig.  340A),  which 
apparently  vary  in  number  among  individual  larvae.  Takita 

( 1 980)  described  and  illustrated  a  "vacuole"  in  the  dorsal  finfold 
of  small,  preflexion  larvae  of  C.  flagris.  C.  richardsoni  and  C. 
ornalipinnts.  Multiple  vacuoles  were  reported  in  the  finfolds  of 
C.  ca/fa/f  (Takita,  1983). 

Postflexion  larvae  are  heavily  pigmented  and  robust  (Fig. 
340C).  They  have  a  prominent  and  highly  visible,  upturned 
notochord  tip  (urostyle).  Caudal,  pelvic,  second  dorsal  and  anal 
fin  ray  counts  may  be  complete  in  some  species  at  3-4  mm  SL 
(Miller  et  al.,  1979;  Takai  and  Yoshioka,  1979).  The  head  be- 
comes flatter  and  broader  as  development  progresses  and  the 
eyes  gradually  assume  their  dorsal,  adjacent  position.  The  pre- 
opercular  spine  first  appears  in  the  length  range  3.5  to  5.0  mm 
SL.  For  most  species,  size  at  metamorphosis  is  approximately 
10  mm  SL. 

Relationships 

Callionymid  eggs  and  larvae  offer  little  clue  to  systematic 
relationships  among  gobiesociform  fishes.  Like  the  gobiesocids, 
callionymid  larvae  are  heavily  pigmented  (Allen,  this  volume) 
but  there  are  few  additional  similarities.  Callionymid  larvae 
hatch  from  pelagic  eggs;  gobiesocids  have  demersal  eggs.  From 


Fig.  340.  Larvae  of  Callionymidae:  (A)  1.7  mm  larva  of  Callionymus  (Paradiplogrammus)  calliste  (from  Takita,  1983:  fig.  21,  p.  443);  (B)  4.7 
mm  larva  of  Callwnvmus  reticulatus  (from  Demir.  1972:  fig.  2.  p.  998);  (C)  4.1  mm  lar\'a  of  Callionymus  (Repomucenus)  beniteguri  {from  Takai 
and  Yoshioka,  1979:  fig.  2-4,  p.  150);  (D)  2.9  mm  larva  of  Callionymus  (Calliurichthys)  decoratus  (from  Miller  et  al.,  1979:  fig.  96,  p.  96);  and 
(E)  2.3  mm  larva  of  Paracallionymus  costatus  (from  Brownell,  1979:  fig.  69,  p.  50). 


640 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


several  hours  after  hatching  and  during  preflexion  stages,  most 
species  of  caliionymid  larvae  have  snout-to-vent  lengths  <50% 
of  standard  length,  while  gobiesocids  have  snout-to- vent  lengths 
>50%.  Both  callionymids  and  gobiesocid  larvae  have  swim- 
bladders  which  are  lost  during  metamorphosis,  a  characteristic 
common  to  many  teleost  families.  Caliionymid  preflexion  larvae 
are  small  and  laterally  compressed.  They  have  low  myomere/ 
vertebral  numbers  (19  to  23).  Gobiesocid  larvae  are  relatively 
large,  basically  cylindrical  in  shape,  with  high  myomere/verte- 
bral  counts  (24  to  37)  (Allen,  this  volume).  The  various  species 
of  caliionymid  larvae  differ  little  from  each  other  but  they  can 
be  identified  by  distinctive  pigment  patterns  and  median  fin  ray 
counts  (Demir,  1972;  Miller  et  al.,  1979:  Takai  and  Yoshioka, 
1979;  Takita,  1980,  1983;  Olney  and  Sedberry,  in  press).  There 


has  been  no  attempt  yet  to  relate  larval  characters  or  differences 
among  larvae  to  the  generic  characters  proposed  by  Nakabo 
(1982b).  There  are  no  descriptions  of  eggs  or  larvae  of  Draco- 
nettidae,  adults  of  which  bear  close  resemblance  to  Calliony- 
midae  (Davis,  1966).  The  discovery  and  description  of  larval 
draconettids  might  resolve  the  systematic  uncertainties  among 
gobiesociform  fishes.  A  careful,  comparative  analysis  of  cal- 
iionymid larval  development  may  clarify  the  generic  relation- 
ships among  species  within  Callionymidae. 

University  of  Maryland,  Center  for  Environmental  and 
EsTUARiNE  Studies,  Chesapeake  Biological  Laboratory, 
Solomons,  Maryland  20688. 


Pleuronectiformes:  Development 
E.  H.  Ahlstrom,  K.  Amaoka,  D.  A.  Hensley, 

H.  G.  MOSER  AND  B.  Y.  SUMIDA 


PLEURONECTIFORM  fishes  have  both  eyes  on  one  side  of 
the  head  in  juveniles  and  adults.  The  eyes  are  symmetrical 
in  larvae,  and  migration  of  either  the  left  or  right  eye  occurs 
during  metamorphosis.  In  some  flatfish  groups  the  eyes  are  on 
the  left  side  (sinistral)  while  in  others  they  are  on  the  right  side 
(dextral);  relatively  few  species  are  indiscriminate.  In  some  flat- 
fishes the  ocular  nerve  of  the  migrating  eye  usually  lies  dorsal 
to  the  other  nerve  in  the  optic  chiasma;  in  other  groups  the 
nerve  of  the  migrating  eye  is  dorsal  or  ventral  in  the  chiasma 
with  about  equal  frequency.  In  most  groups  the  nasal  organ  of 
the  blind  side  also  migrates  to  a  position  near  the  dorsal  midline. 
Features  of  the  dentition  and  cranial  osteology  may  also  show 
asymmetry.  Flatfishes  are  highly  compressed  with  the  underside 
of  the  body  usually  unpigmented.  The  lateral  line  may  be  lacking 
on  the  blind  side;  the  pectoral  fin  is  often  shorter  on  the  blind 
side  and  has  fewer  rays;  the  pelvic  fin  on  the  blind  side  is  often 
shorter,  smaller  and  diflierently  placed  with  respect  to  the  ventral 
midline  compared  with  the  pelvic  fin  on  the  ocular  side;  squa- 
mation  may  be  different  on  the  two  sides  of  the  body.  The  dorsal 
and  anal  fins  are  long-based;  the  dorsal  extends  anteriad  to  at 
least  the  eye  in  all  flatfishes  except  Psettodes  and  the  anal  fin 
extends  well  forward  of  the  first  haemal  spine.  The  caudal  fin 
is  typically  rounded  or  truncate  with  few  or  no  secondary  rays. 
Pleuronectiforms  are  benthic  carnivores,  occurring  worldwide, 
primarily  in  shallow  to  moderate  depths,  with  some  represen- 
tatives in  brackish  and  fresh  water  habitats.  Nelson  (1976)  notes 
a  total  of  520  species. 


The  classification  presented  below  is  based  on  the  works  of 
Regan  (1910,  1929)  and  Norman  (1934,  1966)  with  modifica- 
tions by  Hubbs  (1945),  Amaoka  (1969),  Hensley  (1977),  and 
Futch  ( 1 977).  Our  removal  of  Perissias  from  the  Paralichthyidae 
and  placement  in  the  Bothidae  are  based  on  previously  unpub- 
lished information.  Those  genera  marked  with  an  asterisk  are 
misplaced  in  this  classification  and  are  discussed  in  this  paper 
and  in  Hensley  and  Ahlstrom  (this  volume). 

Order  Pleuronectiformes 
Suborder  Psettodoidei 

Family  Psettodidae  (Indo-Pacific,  West  Africa) 
Pseltodes 
Suborder  Pleuronectoidei 
Family  Citharidae 
Subfamily  Brachypleurinae  (Indo-Pacific) 

Brachyplcura*  Lepidohlepharon 
Subfamily  Citharinae  (Indo-Pacific,   Mediterranean, 
West  Africa) 
Citharoides,  Euatharus 
Family  Scophthalmidae  (North  Atlantic,  Mediterranean, 
Black  Sea) 

Lepidorhombus.  Phrynorhombus,  Scophthalmus, 
Zeugoplerus 
Family  Paralichthyidae  (Western  and  Eastern  Atlantic, 
Eastern  Pacific,  Indo-Pacific) 
Ancylopsetta,  Cephalopsctta,  Cithanchthys,  Cyclop- 


Fig.  341.  Eggs  of  Pleuronectiformes.  Captions  in  each  illustration  indicate  the  species  and  diameter  of  the  egg  in  mm.  Scophthalmus  maeoticus 
maeoticus.  from  Dekhnik,  1973;  Paralichthys  oltvaceus.  from  Mito,  1963;  Bothidae,  from  Mito,  1963;  Limanda  aspera.  from  Pertseva-Ostroumova, 
1954;  Hippoglossotdes  duhius,  from  Pertseva-Ostroumova,  1961;  Microstomus  pacijicus.  onginal,  CalCOFI;  Pleuronichlhys  cornutus.  from  Mito. 
1963;  PehlretisJIavilalus.  from  Robertson.  1975a;  Pellorhamphus  novaezeelandiae.  from  Robertson.  1975a;  Tnnecles  maculalus.  from  Hildebrand 
and  Cable,  1938;  Pegusa  lascans  nasula.  from  Dekhnik,  1973;  Cynoglossus  robuslus.  from  Fujita  and  Uchida,  1957. 


AHLSTROM  ET  AL.:  PLEURONECTIFORMES 


641 


1.10-  1.33 


Scophthalmus  maeotlcus 
maeoticus 


0.92 


Paralichthys  olivaceus 


0.64 


Bothidae 


0.76  -  0.85 


2.10  -  2.94 


2.05  -  2.57 


Limanda  aspera 


Hippoglossoides  dubius 


Microstomus  pacificus 


1.22 


0.62  -  0.68 


0.62  -  0.68 


Pleuronichthys  cornutus 


Pelotretis  flavilatus 


Peltorhamphus 
novaezeelandiae 


0.67  -  0.86 


1.09-  1.35 


0.85  -  0.90 


Trinectes  maculatus 


Pegusa  lascaris  nasuta 


Cynoglossus  robustus 


642 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  1 70.    Characters  of  Eggs  of  Pleuronectinae  Species  which  Lack  Oil  Globules. 


Type  of  egg 

(pelagic  or 

Egg  size 

Species 

Region' 

demersal) 

(mm) 

Chonon 

References 

Cteisthenes  herzensteini 

WNP 

P 

0.84-1.03 



Pertseva-Ostroumova,  1 96 1 

Embassichlhys  balhybius 

ENP 

P 

3.0 

smooth 

Richardson,  1981b 

Eopsella  grigorjewi 

WNP 

P 

1.10-1.20 

striations 

Yusa,  1961;  Fujita,  1965 

E.  jordani 

ENP 

P 

1.21-1.25 

smooth 

Alderdice  and  Forrester,  1971 

Glyptocephalus  cynoglossus 

NA 

P 

1.07-1.25 

striations 

Cunningham,  1887;  Ehrenbaum, 
1905-1909 

G.  slelleri 

WNP 

P 

1.20-1.61 

thick,  reticulate 

Pertseva-Ostroumova,  1961; 
Dekhnik,  1959 

G.  zachirus 

ENP 

P 

1.9-2.15 

striations 

Original 

Hippogtossoides  dubius 

WNP 

P 

2.10-2.94 

smooth 

Pertseva-Ostroumova,  1961 

H.  elassodon 

NP 

P 

2.45-3.75 

smooth 

Thompson  and  Van  Cleve,  1936; 
Pertseva-Ostroumova,  1961 

H.  platessoides 

WNA 

P 

1.38-2.64 

smooth 

Russell,  1976  (summary) 

H.  robuslus 

WNP 

P 

2.04-2.69 

smooth,  thin 

Periseva-Ostroumova,  1961 

Hippoglossus  hippoglossus 

NA 

P 

3.0-3.8 

smooth,  thick 

Tuning,  1936;  Pertseva-Ostroumova, 

1961 
Thompson  and  Van  Cleve,  1936 

H.  stenolepis 

NP 

P 

2.9-3.8 

minute  honey- 

comb structure 

Isopsetta  isolepis 

ENP 

P 

0.90-1.10 

smooth 

Richardson  et  al.,  1980 

Kareius  bicoloratus 

WNP 

P 

1.00-1.15 

reticulate 

Pertseva-Ostroumova,  1961 

Lepidopsetta  bilineala 

NP 

D 

1.02-1.09 

sticky-orange 

Pertseva-Ostroumova,  1961 

L.  inochigarei 

WNP 

D 

0.87-0.95 

thick,  gluey 

Pertseva-Ostroumova,  1961 

Limanda  aspera 

NP 

P 

0.76-0.85 

smooth 

Pertseva-Ostroumova,  1954 

L.  ferruginea 

WNA 

P 

0.79-1.01 

striations 

Miller,  1958;  Colton  and  Marak, 
1969 

L.  limanda 

ENA 

P 

0.66-1.20 

— 

Russell,  1976  (summary) 

L.  punctatissima 

WNP 

P 

0.66-0.87 

smooth 

Pertseva-Ostroumova,  1961 

L.  proboscidea 

WNP 

P 

0.72-0.87 

smooth 

Pertseva-Ostroumova,  1961 

L.  schrenki 

WNP 

D 

0.73-0.83 

adhesive 

Yusa,  1960a 

L.  schrenki  (as  Pseudopleuronectes 

WNP 

D 

0.74-0.83 

adhesive 

Pertseva-Ostroumova,  1961 

yokohamae) 

L.  yokohamae 

WNP 

D 

0.81-0.84 

adhesive 

Yusa,  1960a,  b 

Liopsetta  glacialis 

NP 

P 

1.20-1.60 

thin 

Pertseva-Ostroumova,  1961 

L.  obsciira 

WNP 

D 

0.78-0.94 

thick,  sticky 

Pertseva-Ostroumova,  1961 

L.  pinnifasaata 

WNP 

P 

1.43-1.66 

thin,  folds 

Pertseva-Ostroumova,  1961 

Lyopsella  exihs 

ENP 

P 

1.47-1.68 

smooth 

Original;  Ahlstrom  and  Moser,  1975 

Microslomus  kilt 

WNA 

P 

1.13-1.45 

striations 

Russell,  1976;  Dekhnik,  1959 

M.  pacificus 

ENP 

P 

2.05-2.57 

smooth 

Original;  Ahlstrom  and  Moser,  1975 

Parophrys  vetulus 

ENP 

P 

0.89-0.93 

striations 

Budd,  1940;  Original 

Platichthys  flesus 

ENA 

P 

0.80-1.13 

_ 

Russell,  1976  (summary) 

P.  f.  luscus 

B 

P 

1.05-1.35 

smooth 

Dekhnik,  1973 

p.  slellatus 

NP 

P 

0.89-1.01 

smooth,  thin 

Orcutt,  1950;  Yusa,  1957 

Pleuronecles  pallasii 

NP 

P 

1.67-2.21 

— 

Pertseva-Ostroumova,  1961 

P.  ptatessa 

ENA 

P 

1.66-2.17 

— 

Russell,  1976  (summary) 

Pleuronichlhys  coenosus 

ENP 

P 

1.20-1.56 

polygonal 
pattern 

Sumidaetal.,  1979;  Budd,  1940 

P.  decunens 

ENP 

P 

1.84-2.08 

polygonal 
pattern 

Sumida  et  al.,  1979;  Budd.  1940 

P.  verlicalis 

ENP 

P 

1.00-1.16 

polygonal 
pattern 

Sumida  et  al.,  1979;  Budd,  1940 

Psellichlhys  melanosticlus 

ENP 

P 

ca.  1.0 

— 

Hickman,  1959 

Pseudopleuronectes  amertcanus 

WNA 

D 

0.71-0.96 

adhesive 

Breder,  1923 

P.  herzensteini 

WNP 

P 

0.80-1.0 

smooth 

Pertseva-Ostroumova,  1961 

Reinhardlius  hippoglossoides 

NA/NP 

P 

4.00-4.50 

— 

Jensen.  1935 

Tanakius  kitaharai 

WNP 

P 

1.20-1.30 

striations 

Fujita,  1965 

Verasper  vanegatus 

WNP 

P 

1.60-1.64 

smooth 

Takitaetal.,  1967;  Mito,  1963 

'  B  =  Black  Sea,  ENA 
Pacific. 


=  eastern  North  Atlantic.  ENP  =  eastern  North  Pacific,  I  =  India,  NA  =  North  Atlantic.  NP  =  North  Pacific.  WNA  =  western  North  Atlantic.  WNP  =  western  North 


setta.  Etropus.  Gastropsetta.  Hippoglossina. 
Lioglossina.  Paralichthys.  Pseudorhombus, 
Syaciurn.  Tarphops,  Tephrinectes*  Thysanop- 
setta*  Verecundum.  Xystreurys 
Family  Bothidae 

Subfamily  Taeniopsettinae  (Western  Atlantic,  Eastern 
Pacific,  Indo-Pacific) 


Engyophrys.  Perissias,  Taeniopsetta,  Trichopsetta 
Subfamily  Bothinae  (Indian,  Pacific,  Atlantic,  Medi- 
terranean, SoutheiTt  oceans) 
Arnoglossus.  Asterorhoinhiis,  Bothus.  Chascanopset- 
ta,   Crossorhoinbus.  Engyprosopon.   Grainma- 
lobothus,  Japonolaeops.  Kamoharaia,  Laeops, 
Lophonectes,   Monolene,   Mancopsetta*   Neo- 


AHLSTROM  ET  AL.:  PLEURONECTIFORMES 


643 


Table  171.    Characters  of  Pleuronectiform  Eggs  with  a  Single  Oil  Globule. 


Egg  size 

Oil  globule 

Size  at 

Taxon 

Region' 

(mm) 

si/e  (mm) 

Chorion 

halching  (mm) 

References 

Pleuronectidae 

Hypsopsella  gutlulala 

ENP 

0.78-0.89 

0.12-0.14 

smooth 

1.7-2.3 

Sumidaet  al.,  1979; 
Eldridge,  1975 

PIcuromchthys  cornutus 

WNP 

1.03-1.25 

polygonal 
pattern 

2.8-3.8 

Mito,  1963;  Takita  and 
Fujita,  1964 

P.  niieri 

ENP 

0.94-1.08 

0.08-0.14 

polygonal 
pattern 

2.1 

Sumidaet  al.,  1979 

Scophthalmidae 

Lepidorhombus  whiffiagonis 

ENA 

1.02-1.22 

0.25-0.30 

striations 

ca.  4.0 

Mcintosh,  1892;  Holt, 
1893 

Phrynorhombus  norvegicus 

ENA 

0.72-0.92 

0.09-0.16 

rugose 

2.5-2.8 

Ehrenbaum,  1905-1909; 
Hefford,  1910 

P.  regius 

ENA 

0.90-0.99 

0.16-0.18 

_ 

2.4 

Holt,  1897 

Scophthalmus  aquosus 

WNA 

0.90-1.38 

0.15-0.30 

striations 

ca.  2.0 

Martin  and  Drewry,  1978 
(summary) 

S.  maeoticus 

B 

1.10-1.33 

0.17-0.23 

_ 

3.5  TL 

Dekhnik,  1973 

S.  mcLximus 

ENA 

0.90-1.20 

0.15-0.22 

rugose 

2.1-3.0 

Holt,  1892;  Jones.  1972 

S.  rhombus 

B 

1.20-1.50 

0.16-0.25 

striations 

3.8 

Jones,  1972 

Zeugoplerus  punclatus 

ENA 

0.92-1.07 

0.17-0.20 

- 

2.5-2.9 

Hefford,  1910 

Paralichthyidae 

llippoglossina  obtonga 

WNA 

0.91-1.12 

0.17 

smooth 

2.7-3.2 

Miller  and  Marak,  1962 

H.  slomala 

ENP 

1.22-1.38 

0.20-0.26 

smooth 

3.7 

Sumida  et  al.,  1979 

Paralichlhys  californicus 

ENP 

0.74-0.82 

0.10-0.19 

smooth 

ca.  2.0 

Original 

P.  dentatus 

WNA 

0.90-1.10 

0.18-0.31 

_ 

2.4-2.8 

Smith  and  Fahay,  1970 

P.  olivaceus 

WNP 

0.83-1.03 

0.13-0.21 

smooth 

2.6-2.8 

Mito,  1963 

Pseudorhombus  cinnamoneus 

WNP 

0.77-0.89 

0.12-0.14 

_ 

1.8-2.0 

Mito,  1963 

Cilhanchlhys  arclifrons 

WNA 

0.70-0.82 

— 

smooth 

ca.  2.0 

Richardson  and  Joseph, 

1973 

Bothidae 

Arnoglossus  capensis 

ESA 

0.72 

0.12 

smooth 

2.2 

Brownell,  1979 

A.  kessleri 

B 

0.59-0.70 

0.10-0.13 

smooth 

1.8-1.9 

Dekhnik.  1973 

A.  taterna 

ENA 

0.60-0.76 

0.11-0.15 

smooth 

2.6 

Russell,  1976  (summary) 

A.  scapha 

NZ 

0.78-0.88 

0.11-0.12 

smooth 

— 

Robertson,  1975a 

A.  ihon 

ENA 

0.67-0.74 

0.12 

smooth 

1.6-2.0 

Russell,  1976  (summary) 

'  ESA  =  eastem  South  Atlantic;  NZ  =  New  Zealand;  key  to  other  regions  as  in  Table  1 70. 


laeops.    Parabothus.    Pelecanichthys.    Psettina, 
Tosarhoinbus 
Family  Pleuronectidae 

Subfamily   Pleuronectinae  (Atlantic,   Mediterranean, 
Pacific,  Arctic) 
Acanthopsetta,  Atheresthes.   Cleisthenes.   Clidoder- 
ma.  Dexistes,  Embassichthys,  Eopsetta.  Glyp- 
tocephalus,  Hippoglossoides.  Hippoglossus, 
Hypsopsella.  Isopsetta.  Lepidopsetta.  Limanda. 
Liopseila.  Lyopsetla,  Mtcrostomus.  Parophrys. 
Platichlhys.  Pleuronectes,  Pleuromchthys.  Psel- 
tichthys.    Pseudopleuronectes.    Reinhardlius. 
Tanakius.  Verasper 
Subfamily  Poecilopsettinae  (Indo-Pacific,  Atlantic) 

Marleyclla.  Nematops.  Poecilopselta 
Subfamily  Paralichthodinae  (Indian  Ocean  off  South 
Africa) 
Paralichthodes 
Subfamily  Samarinae  (Indo-Pacific) 

Samans,  Samariscus 
Subfamily  Rhombosoleinae  (New  Zealand,  Southern 
Australia,  South  America) 
Ammolretis.  Azygopus.  Colislium,  Oncopterus.  Pe- 
lotretis.  Peltorhamphus.  Psammodisciis.  Rhom- 
bosolea 


Suborder  Soleoidei 
Family  Soleidae 
Subfamily  Soleinae  (Worldwide,  tropical  to  temperate) 

Norman  (1966)  recognized  22  genera 
Subfamily  Achirinae  (American  coasts,  some  fresh 
water)  Norman  (1966)  recognized  9  genera 
Family  Cynoglossidae 

Subfamily  Symphurinae  (Tropical-Subtropical  Amer- 
ican coasts,  Mediterranean,  West  Africa,  Indo-Pa- 
cific) 
Symphurus 
Subfamily  Cynoglossinae  (Indo-Pacific,  Mediterra- 
nean, West  Africa,  Japan,  some  fresh  water) 
Cynoglossus,  Paraplagusia 

A  profuse  literature  on  the  life  history  stages  of  flatfishes  has 
accumulated  since  the  early  work  of  Cunningham  (1887,  1889, 
1890,  1891)  who  described  numerous  series  reared  from  eggs 
collected  from  running  ripe  females.  Other  European  workers 
(Holt.  1893;  Mcintosh  and  Pnnce,  1890;  Petersen,  1904,  1905, 
1906,  1909;  Schmidt,  1904;  Kyle,  1913)  identified  early  life 
history  series  of  additional  species  so  that,  by  the  time  of  Eh- 
renbaum's  (1905-1909)  summary,  ontogenetic  stages  of  a  major 
portion  of  the  eastem  North  Atlantic  flatfish  fauna  were  known. 
Padoa  (1956k)  summarized  ontogenetic  information  on  Medi- 


644 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  172.    Characters  of  Pleuronectiform  Eggs  with  Multiple  Oil  Globules. 


Egg  size 

Region' 

(mm) 

Number  of  oil  globules 

Yolk 

Chonon 

References 

Achirinae 

Achirus  lineatus 

WNA 

0.71-0.76 

12-14 

homogeneous 

smooth,  thin 

Houdeet  al.,  1970 

Trinecles  maculaius 

WNA 

0.67-0.86 

15-34 

homogeneous 

smooth 

Hildebrand  and  Cable. 
1938 

Soleinae 

Aesopia  cornula 

WNP 

1.45-1.60 

many;  scattered 

homogeneous 

polygonal  mesh 

Mito.  1963 

Auslroglossus  microlepis 

ESA 

0.88 

12-20 

homogeneous 

smooth 

Brownell,  1979 

Buglossidiuni  luieum 

ENA 

0.64-0.94 

12-21;  scattered 

peripheral 

smooth 

Holt,  1891;  Hefford, 
1910 

Dicologoglossa  cuneata 

ENA 

0.70-0.84 

60-80;  scattered 

peripheral 
segmentation 

Ugardere,  1980 

Microchirus  ocellatus 

ENA 

0.90-1.10 

30-40;  scattered 

penpheral 
segmentation 

smooth 

Palomera  and  Rubies, 

1977 

M.  variegatus 

ENA 

1.28-1.42 

to  50  +  ;  scattered 

peripheral 
segmentation 

smooth 

Cunningham,  1889 

Pegiisa  impar 

M 

1.06 

Padoa,  1956k 

P.  lascaris  lascaris 

ENA 

1.28-1.38 

to  50  + 

peripheral 
segmentation 

Holt,  1891;  Hefford, 
1910 

P.  lascaris  nasuta 

B 

1.36-1.38 

many;  clumped 

homogeneous 

polygonal  mesh 

Dekhnik.  1973 

Solea  solea 

ENA 

0.98-1.58 

many;  highly 
clumped 

peripheral 
segmentation 

smooth 

Cunningham,  1889; 
Fabre-Domergue  and 
Bietnx,  1905 

Synaptura  kleiiu 

ESA 

1.34 

many;  clumped 

homogeneous 

smooth 

Brownell,  1979 

Zebrias  japomcus 

WNP 

1.75 

many;  scattered 

homogeneous 

smooth 

Mito,  1963 

Z.  zehra 

WNP 

1.60 

many;  scattered 

homogeneous 

Mito,  1963 

Symphunnae 

Symphurus  atricauda 

ENP 

0.71-0.78 

10-23 

homogeneous 

smooth,  colored 

Onginal 

Cynoglossinae 

Cynoglossus  capensis 

ESA 

0.75 

2-16 

homogeneous 

smooth 

Brownell,  1979 

C.  robustus 

WNP 

0.85-0.90 

5-15 

homogeneous 

fine  hexagonal 
network 

Fujita  and  Uchida,  1957 

C.  {Areliscus)  tngrammus 

WNP 

1.19-1.23 

30-50 

homogeneous 

smooth 

Fujita  and  Takita,  1965 

Cynoglossidae  no.  5 

WNP 

0.71 

14 

homogeneous 

smooth 

Mito,  1963 

Cynoglossidae  sp.  A 

I 

0.84 

13-15 

homogeneous 

smooth 

Vijayaraghavan,  1957 

Cynoglossidae  sp.  B 

I 

0.82 

1 8-22;  clustered 

homogeneous 

smooth 

Vijayaraghavan,  1957 

Cynoglossus  I 

I 

0.60 

16-30 

homogeneous 

smooth 

Nair,  1952a 

Cynoglossidae  (as  Solea 

I 

0.61-0.71 

17-25 

homogeneous 

smooth 

John,  1951b 

ovaia) 

Rhombosoleinae 

Ammotrelis  rostratus 

NZ 

ca.  0.8 

8-11 

homogeneous 

smooth 

Thomson,  1906 

Colistium  gunlheri 

NZ 

1.0-1.08 

14-26 

homogeneous 

Robertson.  1975a 

C.  nudipinnis 

NZ 

ca.  1.5 

21-28 

homogeneous 

Robertson,  1975a 

Pelotretis  JIavilaliis 

NZ 

0.85-0.95 

8-18 

homogeneous 

smooth 

Robertson,  1975a 

Peltorhamphus  novaczee- 

NZ 

0.62-0.68 

2-6 

homogeneous 

smooth 

Robertson,  1975a 

landiae 

P.  tenuis 

NZ 

0.58-0.68 

2-4 

homogeneous 

smooth 

Robertson.  1975a 

Rhombosolea  Icporina 

NZ 

0.58-0.70 

2-7 

homogeneous 

smooth 

Robertson,  1975a 

R.  pleheia 

NZ 

0.58-0.72 

2-13 

homogeneous 

smooth 

Robertson,  1975a;  Rob- 
ertson and  Raj,  1971 

Bothinae 

Mancopsetta  maculala 

S 

2.45-3.00 

20  + 

homogeneous 

smooth 

Efremenko  et  al.,  1981 

anlarclica 

'  M  =  Mediterranean.  S  =  southern  oceans,  key  to  other  regions  as  in  Table  I  70, 


terranean  flatfishes  and  more  recently  Russell  (1976)  provided 
an  extensive  review  of  previous  European  contributions. 
Knowledge  of  ontogenetic  stages  of  western  Atlantic  flatfishes 
is  summarized  by  Martin  and  Drewry  ( 1 978)  and  Fahay  ( 1 983). 
Early  life  histories  of  North  Pacific  flatfishes  are  treated  com- 
prehensively by  Pertseva-Ostroumova  { 1 96 1 ).  Japanese  and  In- 
dian workers  have  provided  a  long  list  of  contnbutions  to  flatfish 
life  history  studies  and  Amaoka  (1969,  1979),  Hensley  (1977) 
and  Futch  (1977)  employed  ontogenetic  characters  in  assessing 


phylogenetic  relationships.  The  individual  contributions  to  flat- 
fish ontogeny  are  too  numerous  to  summarize  concisely  and  are 
cited  in  the  section  that  follows. 

Development 
Eggs 
Eggs  are  known   for  most   species  in   Pleuronectidae  and 
Scophthalmidae  and  for  only  a  few  to  moderate  numbers  of 


Fig.  342,     Larvae  of  Psettodidae  and  Citharidae.  (A)  Psellodes  erumei,  4.3  mm.  from  Leis  and  Rennis,  1983:  (B)  P.  eruinei.  8.7  mm.  ibid;  (C) 
Brachypleura  novaezeelandiae.  5.0  mm,  from  Pertseva-Ostroumova,  1965;  (D)  B.  novaezeetandiae.  7.5  mm,  ibid. 


646 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


.«!»-*■ 


M.    :^r..M 


^^^ 


B 


f 


D      / 

-^ 

V 

\\N                         ^ 

\\ 

V---' 

'-•^6<:<;^ 


^ 


,i-.-  s* 


A  />-    '"^   -^ 


^;;-'5*^« 


^  *  '"  i-fi':'  -."♦J 


^vl*-' 


t;. 


Fig.  343.  Larvae  of  Scophthalmidae.  (A)  Zeugopteriis  punctatus.  8.9  mm,  dorsal  view,  from  Petersen,  1909;  (B)  Z.  punclalus.  9.0  mm,  ibid; 
(C)  Lepidorhombus  boscii.  9.7  mm,  ibid;  (D)  Phrynorhombits  regius.  8.0  mm,  ibid;  (E)  Scophlhalmus  maximus.  7.4  mm,  from  Jones,  1972;  (F) 
5.  rhombus.  8.0  mm,  ibid. 


Fig.  344.  Larvae  of  Paralichthyidae.  (A)  Paralichthys  californkus.  7.0  mm.  original,  CalCOFI;  (B)  As  above,  dorsal  view;  (C)  Xvstreurvs 
liolepis.  6.7  mm,  original,  CalCOFI;  (D)  As  above,  dorsal  view;  (E)  Hippoglossina  stomata.  8.6  mm,  from  Sumida  et  al.,  1 979;  (F)  Pseudorhombus 
penlophlhalmus,  9.2  mm,  from  Okiyama,  1974a;  (G)  Tarphops  oligolepis.  9.2  mm,  ibid. 


AHLSTROM  ET  AL.:  PLEURONECTIFORMES 


647 


648 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  173.    Summary  of  Ontogenetic  Characters  of  Pleuronectiforms.  (Line  indicates  data  unavailable  or  presented  elsewhere  in  column.) 


Eggs 

Lar\'ae 

Oil  globule: 

Chonon: 

Single/ 

Yolk: 

Smooth/ 

Size  at 

Elongate  dorsal 

Egg  size 

Multiple/ 

Homogeneous/ 

Slnated/ 

hatehing 

Size  at  meta- 

rays: Present 

Taxon 

(mm) 

Absent 

Segmented 

Polygonal 

(mm) 

morphosis  (mm) 

(no.)/ Absent 

Psettodidae 

— 

— 

— 

— 

— 

9-10 

10 

Citharidae 

_ 

— 

_ 

_ 

_ 

_ 

_ 

Brachypleura 

- 

- 

- 

- 

- 

7-8 

ca.  6 

Scophthalmidae 

0.72-1.5 

s 

H 

St 

2.0-4.0 

8->20 

A 

Paralichthyidae 

<0.70-1.38 

s 

H 

s 

1.8-3.7 

— 

— 

Paralichthys  group 

— 

— 

— 

— 

— 

7.5-15 

4-5 

Pseudorhomhus  group 

— 

— 

— 

- 

— 

8.7-12.5 

5-9 

Cyclopsetta  group 

- 

- 

- 

- 

- 

9->35 

*« 

Bothidae 

0.59-0.88 

s 

H 

s 

1.8-2.6 

_ 

_ 

Taeniopsettinae 

— 

— 

— 

— 

— 

19-60 

1 

Bothinae 

- 

- 

- 

- 

- 

15-120 

1 

Pleuronectidae 















Pleuronectinae 

0.66-4.5 

A* 

H 

S,  P.  St 

1.7-16.0 

4.4-65 

A 

Poecilopsettinae 

— 

— 

— 

— 

— 

ca.  30 

A 

Samarinae 

— 

— 

— 

— 

— 

ca.  30 

1 

Rhombosoleinae 

0.58-1.5 

M 

H 

s 

ca.  1.8 

- 

- 

Soleidae 

0.64-1.75 

M 

H,S 

S.  P 

1.6-4.1 

_ 

_ 

Soleinae 

_ 

— 

— 

— 

— 

3.4-18 

A 

Achirinae 

- 

- 

- 

- 

- 

3-5.5 

*** 

Cynoglossidae 

0.60-1.23 

M 

H 

S,  P 

1.3-3.2 

— 

— 

Cynoglossinae 

— 

— 

— 

— 

— 

ca.  4-18 

2 

Symphurinae 

— 

— 

— 

— 

— 

ca.  12-32 

usually 
4-5 

•  Single  oil  globule  present  in  3  speeies- 

**  0-2  in  EtropiLi,  0.  2.  3  in  Cnhanchthys;  5-8  in  Syactum:  8-1 1  in  Cyclopsena. 
••*  Third  ray  elongated  in  Achirus. 

■•••  0.  I.  or  2  in  Cithanchthy^.  2  or  3  in  Cyclopsetta  and  Syacium:  0  or  I  in  Etropus. 
•*••*  S  Iti^tlata  develops  elongate  third  and  fourth  rays, 
******  Protruding  in  Chascanupsetta.  Petecanichthys.  and  Kamuharaia. 


species  in  other  groups,  including  Soleidae,  Cynoglossidae, 
Paralichthyidae,  and  Bothidae. 

With  a  few  exceptions,  the  eggs  of  flatfishes  are  pelagic,  round, 
have  homogeneous  yolk,  a  narrow  perivitelline  space,  and  an 
unsculptured  chorion  (Fig.  341).  The  eggs  of  all  flatfishes  are 
spawned  separately.  The  characters  of  eggs  showing  greatest 
differences  among  flatfishes  are  1)  egg  size,  and  2)  the  presence 
or  absence  of  an  oil  globule(s)  (Tables  170-172). 

Of  the  approximately  60  species  of  pleuronectine  flatfishes  of 
the  North  Pacific  and  North  Atlantic,  eggs  are  known  for  at  least 
45  (Table  170).  Six  species  are  known  to  have  demersal  eggs; 
these  are  round  or  occasionally  off-round  and  have  a  sticky, 
adhesive  chorion  that  permits  clustering  or  adhesion  to  bottom 
objects.  Egg  diameters  range  from  0.66  to  4.5  mm  within  the 
subfamily.  The  yolk  is  homogeneous  in  all  pleuronectine  eggs. 
The  perivitelhne  space  is  narrow  to  moderate,  except  for  eggs 
of  Hippoglossoides,  which  have  a  wide  perivitelline  space,  usu- 
ally 25-30%  of  the  egg  diameter  on  either  side  of  the  yolk  mass. 
The  chorion  has  the  appearance  of  being  smooth  on  eggs  of  most 
species,  but  closer  inspection  reveals  striations  or  reticulations 
on  the  chorion  of  some  kinds.  The  chorion  of  Pleuromchthys 
eggs  has  a  striking  hexagonal  pattern.  The  eggs  of  pleuronectine 
flatfishes,  except  for  three  species,  lack  an  oil  globule.  The  state 
of  embryonic  development  achieved  in  the  egg  is  related  to  egg 
size,  more  specifically  to  yolk  size.  Larvae  hatching  from  small 
eggs  lack  eye  pigment,  a  functional  mouth  and  pectoral  fins; 
those  hatching  from  larger  eggs  are  much  more  advanced,  with 
pigmented  eyes,  a  functional  mouth  and  pectorals.  Embryos  in 


middle-  and  late-stage  eggs  are  pigmented,  with  patterns  varying 
between  genera  and  species.  Among  species,  yolk  pigment  can 
range  from  unpigmented,  to  some  pigment  on  yolk  adjacent  to 
the  embryo,  to  heavily  pigmented.  Pigment  can  also  be  present 
on  finfolds  of  late-stage  eggs  of  some  flatfishes. 

Eggs  of  Scophthalmidae.  Paralichthyidae  and  Bothidae  have 
a  single,  small  to  moderate-sized  oil  globule,  are  pelagic,  round, 
have  a  narrow  to  moderate  perivitelline  space,  and  homoge- 
neous yolk  (Fig.  341,  Table  171).  In  late-stage  eggs  and  newly 
hatched  larvae  the  single  oil  globule  usually  is  in  the  rear  of  the 
yolk  mass. 

Eggs  are  known  for  8  of  the  10  species  of  scophthalmid  flat- 
fishes. They  range  in  size  from  0.72  to  1.50  mm.  The  chorion 
is  striated  or  rugose  in  six  species  and  this  may  apply  to  all. 
Embryos  develop  considerable  pigment  over  the  head  and  body 
and  often  in  finfolds;  pigment  over  the  yolk  mass  and  oil  globule 
can  range  from  none,  or  sparse,  to  intense. 

Eggs  are  known  for  only  a  few  species  in  the  family  Paralich- 
thyidae. These  range  in  size  from  0.70-1.38  mm;  chorions  are 
unsculptured.  Except  for  a  few  species  of  Arnoglossus.  eggs  of 
bothid  flatfishes  are  practically  unknown.  Mito  (1963)  lists  10 
kinds  of  bothid  eggs  oflT  Japan,  unidentified  to  genus;  8  of  these 
have  diameters  under  1.0  mm.  Eggs  of  his  Bothidae  No.  9  are 
slightly  ofl"-round  and  three  different  eggs  have  a  conspicuous 
wart-like  appendage.  Much  work  remains  to  be  done  in  iden- 
tifying eggs  of  fishes  of  these  families,  preferably  through  rearing 
eggs  from  known  parents. 

Eggs  with  multiple  oil  globules  are  typical  of  the  families 


AHLSTROM  ET  AL.:  PLEURONECTIFORMES 


649 


Table  173.     Extended. 


Larvae 

Elongate 

pelvic  rays: 

Present  Absent 

Gul: 

Normal/ 

Protruding/ 

Trailing 

Preopercular 

spines:  Present 

Absent 

Otic  region 

spines;  Present/ 

Absent 

Frontal  region 

spines:  Present/ 

Absent 

Urohyal  spines: 
Present/ Absent 

Basipter>'gial 

spines:  Present/ 

Absent 

Cleithral 

spines:  PrescnL 

Absent 

Body  spines: 
Present/Absent 

A 

N 

P 

A 

A 

A 

A 

A 

A 

_ 

















P 

N 

p 

A 

A 

A 

A 

A 

A 

A 

N 

P,  A 

P.  A 

P,A 

A 

A 

A 

A 









— 

_ 

_ 

_ 

— 

A 

N 

P 

A,P 

A 

A 

A 

A 

A 

A 

N 

P 

A,P 

A 

A 

A 

A 

A 

**** 

N 

P 

P,A 

A 

A 

A 

A 

A 















— 

— 

A 

N 

A 

P 

A 

P 

P 

P 

A 

A 

****** 

A 

A 

A 

P,A 

P,A 

P,A 

P,  A 

A 

N 

A,P 

A,P 

A,  P 

A 

A 

A 

A 

A 

N 

A 

A 

A 

A 

A 

A 

A 

A 

P 

A 

A 

A 

A 

A 

A 

A 

A 

N,  P 

A 

A 

A 

A 

A 

A 

A 

A 

N.  P 

A.  P 

A,  P 

A,P 

A 

A 

A 

P.  A 

A 

P 

A 

A 

A 

A 

A 

A 

A 

A,  *"" 

P.  T 

A 

A 

A 

A 

A 

A 

A 

Soleidae.  Cynoglossidae.  the  pleuronectid  subfamily  Rhombo- 
soleinae,  and  Mancopsetia.  previously  considered  a  bothid  (Fig. 
341.  Table  172).  Eggs  have  been  described  for  about  a  dozen 
kinds  of  soleids  ranging  in  size  from  0.64-1.75  mm.  Eggs  are 
round,  or  occasionally  slightly  off-round.  Oil  globules  are  usually 
numerous  but  vary  in  number,  size  and  distribution  within  the 
yolk.  They  can  be  highly  clumped,  as  in  Solea  solea.  or  scattered 
throughout  the  yolk,  as  in  Microchirus  variegalus.  In  eggs  of  the 
latter,  oil  globules  were  observed  to  range  in  size  from  0.015- 
0.12  mm.  whereas  they  are  much  smaller  and  more  uniform  in 
size  in  Solea  solea  or  Pegusa  lascans.  Eggs  of  the  two  achirine 
soleids  described  from  the  western  Atlantic  have  a  relatively 
low  number  of  oil  globules.  Perivitelline  space  is  narrow  to 
negligible  in  soleid  eggs.  The  yolk  is  peripherally  segmented  in 
eggs  of  the  four  species  known  from  the  eastern  North  Atlantic. 
Yolk  is  more  completely  segmented  in  the  egg  designated  as 
Synapturinae  No.  1  by  Mito  (1963).  Yolk  can  remain  unseg- 
mented,  however,  as  for  example  in  Achirus  lineatus  and  Tri- 
nectes  maculatus.  Although  the  chorion  of  soleid  eggs  is  usually 
smooth  and  unsculptured,  Mito  (1963)  found  eggs  of  Aesopia 
cormita  to  have  a  pattern  of  large  hexagonal  meshes.  0.18-0.24 
mm  wide,  covering  the  chorion,  and  Dekhnik  (1973)  shows  fine 
polygonal  sculpturing  on  the  chorion  of  P.  lascans. 

Eggs  of  the  few  cynoglossid  species  known  (Table  172)  are 
small,  have  homogeneous  yolk  without  secondary  segmentation, 
a  narrow  perivitelline  space  and  either  an  unsculptured  chonon 
or  one  with  small  polygonal  meshes.  Oil  globules  range  in  num- 
ber between  5-50,  and  can  be  variously  distributed  in  the  yolk. 

Robertson  (1975a)  descnbed  eggs  of  seven  species  of  Rhom- 
bosoleinae,  belonging  to  four  genera  (Table  1 72).  Egg  diameters 
range  from  0.58  to  1.5  mm.  Oil  globules  in  described  eggs  range 


in  number  from  2-28.  Yolk  is  homogeneous,  the  perivitelline 
space  IS  narrow,  and  the  chonon  is  smooth. 

Efremenko  et  al.  ( 1 98 1 )  described  the  ovarian  and  planktonic 
eggs  of  Mancopsetia  macidala  antarctica  and  showed  that  they 
are  large  (2.45-2.75  mm)  and  have  multiple  oil  globules  (>20). 
This  finding  provides  evidence  that  Mancopsetta  does  not  be- 
long in  the  Bothidae. 

Larvae 

In  addition  to  such  features  as  meristics,  fin  arrangement,  and 
osteology  of  the  fin  supports  and  axial  skeleton  (which  develop 
gradually  during  ontogeny  and  are  essential  for  identification  of 
flatfish  larvae)  the  larval  stage  itself  provides  many  characters 
useful  in  identification  and  systematic  analysis.  Larval  charac- 
ters are  summarized  in  Table  173  and  below. 

Psettodidae  (Fig.  342).  — Aboussouan's  (1972c)  description  of 
prcflexion  larvae  of  Psettodes  bennetti  was  based  on  five  spec- 
imens. 4.4-5.7  mm  in  length.  Leis  and  Rennis  (1983)  describe 
a  series  of  five  larval  specimens  of  Psettodes  erumei.  3.0-8.7 
mm  in  length.  The  smallest  specimen  has  a  large  yolk  sac.  the 
6.0-mm  larva  is  in  mid-flexion  and  the  largest  specimen  is 
undergoing  eye  migration.  Larvae  have:  a  deep,  relatively  thick 
body;  large  head  with  massive  jaws  that  extend  well  beyond  the 
rear  margin  of  the  eye  and  bear  large,  early-forming  cursed  teeth; 
large  eye;  small  preopercular  spines;  and  1 0  early-forming  elon- 
gate dorsal  rays.  Dorsal  and  anal  fin  rays  are  all  present  at  6.0 
mm  but  rays  do  not  appear  in  paired  fins  until  about  8.0  mm. 
Prcflexion  larvae  have  a  series  of  large  melanophores  along  the 
dorsal  midline,  large  melanophores  alternating  with  smaller  ones 


650 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


.jf(im£Mm!^y^^- 


Fig.  345.  Larvae  of  Paralichthyidae.  (A)  Citharichthys  stigmaeus.  14.8  mm,  from  Ahlstrom  and  Moser,  1975;  (B)  C.  sordtdus,  14.5  mm,  ibid; 
(C)  C.  plalophrys.  8.6  mm,  onginal,  CalCOFI;  (D)  Etropus  crossotus.  6.0  mm,  from  Tucker.  1982;  (E)  Cyclopselta  chitlendem.  13.0  mm,  from 
Evseenko,  1982a;  (F)  Syacium  ovale.  6.5  mm,  original,  CalCOFI. 


AHLSTROM  ET  AL.:  PLEURONECTIFORMES 


651 


Table  174.    Numbers  of  Rays  in  Dorsal  Crest  and  Size  at  Developmental  Events  in  Paralichthyidae. 


Number  of 

elongate 

dorsal  fin 

Si/c  al 

Si/e  al 

Size  at 

Species 

rays 

hatching  (mm) 

flexion  (mm) 

Iransfoimation  (mm) 

References 

Hippoglossina  stomata 

6 

3.7 

6.2-8.8 

9.1->11.7 

Sumitia  et  al.,  1979 

H  ohionga 

~6 

2.7-3.2 

6.3-7.7 

10-14 

Leonard.  1971 

Paralichlhys  califormcus 

5 

2.0 

6.0-7.3 

7.5-9.4 

Original,  Ahlslrom  and  Moser, 
1975 

P.  dematus 

4-8 

2.4-2.8 

7-9.5 

-9.5 

Smith  and  Fahay,  1970 

P.  olivaceus 

5-6 

2.6-2.8 

7.1-8.7 

10.2->14.2 

Okiyama,  1967 

Xystreurys  liolepis 

6 

2.0 

6.0-6.7 

7.5->8.7 

Original 

Psciidorhomhus  elevatus 

9 

_ 

5.5-6.4 

-10 

Devi,  1969 

P.  pentophlhalmus 

7 

— 

7.1-7.6 

8.7-12.2 

Okiyama,  1974a;  Minami,  1981a 

Tarphops  oligolepis 

8 

— 

— 

9.2-12.5 

Okiyama,  1974a 

Cithanchthys  arctifrons 

3 

<2.3 

-5-8 

13-15 

Richardson  and  Joseph,  1973 

C.  cornutus 

3 

<2.2 

5.8-8.9 

-18 

Tucker,  1982 

C.  gymnorhinus 

3 

— 

5.3-7.7 

-18 

Tucker,  1982 

C.  platophrys 

3 

<2.0 

5.3-6.1 

11.2-18.5 

Original 

C.  sordidus 

2 

-2.0 

10.4-11.4 

20->39 

Ahlstrom,  1965;  original; 
Ahlstrom  and  Moser,  1975 

C.  spiloplerus 

2 

— 

5.7-6.8 

9-12 

Tucker,  1982 

C.  sligmaetis 

0 

-2.0 

9.2-10.2 

24.0->35.5 

Ahlstrom,  1965;  onginal; 
Ahlstrom  and  Moser,  1975 

Eiropus  crossotus 

2 

<2.3 

4.9-9.5 

-11 

Tucker,  1982 

E.  microslomus 

0 

<2.3 

5-7 

10-12 

Richardson  and  Joseph,  1973 

Cydopsella  chillendeni 

8-9 

— 

-7.5 

>13.0 

Evseenko,  1982a 

C.  fimbriala 

~9 

-1.5 

6.9 

14.0 

Gutherz,  1970;  Evseenko,  1982a 

C.  querna 

8-11 

— 

— 

>32 

Ahlstrom,  1972a 

Syacium  guineensis 

7 

<2.1 

<6.5 

>13.9 

Aboussouan,  1968b 

S.  gunleri 

5-8 

<1.8 

— 

— 

Evseenko,  1982a 

S.  micruruin 

5-8 

<1.8 

— 

— 

Evseenko,  1982a 

S.  ovale 

5-8 

-1.6 

4.1-4.8 

-14-- 20 

Ahlstrom,  1972a;  original 

S.  papillosum 

5-8 

<2.3 

5.5-6.0 

15-13 

Futch  and  Hoff,  1971;  Evseenko, 
1982a 

along  the  ventral  midline,  and  small  melanophores  on  the  trunk, 
tail,  ventral  gut,  pectoral  fin,  brain  and  lower  jaw.  During  flexion 
the  entire  body  except  for  the  caudal  fin  base  becomes  solidly 
pigmented,  a  darker  band  forms  forward  of  the  caudal  peduncle, 
and  the  snout  becomes  heavily  pigmented. 

Cilharuiae  (Fig.  342).  — Larvae  of  this  family  are  known  from 
five  specimens  (4.0-8.0  mm)  of  Brachypleura  novaezeelandiae 
described  by  Pertseva-Ostroumova  (1965).  Notochord  flexion 
occurs  between  5.0  and  7.0  mm  and  transformation  at  about 
8.0  mm.  Larvae  have  a  moderately  deep,  thick  body  and  a  large 
head  with  large  jaws  and  eyes  and  about  10  large  preopercular 
spines;  the  sixth  dorsal  ray  is  elongate  and  the  rays  anterior  to 
it  are  assumed  to  be  elongate,  although  damaged  in  all  available 
specimens;  pelvic  fins  are  elongate,  extending  beyond  the  anus; 
pigment  consists  of  a  series  of  melanophores  along  the  dorsum, 
a  series  along  the  horizontal  septum,  and  a  postanal  series  along 
the  ventrum,  melanophores  below  the  gut,  and  on  the  pelvic 
fin. 

Scophthalmidae  (¥'\g.  343).  — Larvae  are  known  for  9  of  the  10 
species  of  this  family.  Petersen  ( 1 909)  described  5  of  the  7  species 
occurring  in  the  eastern  North  Atlantic;  Jones  (1972)  provided 
excellent  illustrations  of  the  2  species  of  eastern  Atlantic  Scoph- 
thalmus  and  Bigelow  and  Welsh  (1925)  described  larvae  of  5. 
aquosus.  the  only  western  Atlantic  representative  of  the  family. 
Newly  hatched  larvae  are  2.0-4.0  mm  in  length  (Table  171); 
size  at  notochord  flexion  for  most  species  is  6.0-8.0  mm.  Meta- 


morphosis can  begin  by  8  or  9  mm  and  be  completed  by  1 3 
mm  (S.  aquosus.  Phrynorhombus  norvegicus,  Zeugopterus 
punctaius)  or  delayed  to  over  20  mm  (S.  maximus.  S.  rhombus). 
Larvae  are  deep-  and  thick-bodied,  especially  at  the  gut,  have 
a  large  head  and  jaws  and  moderate  to  large  eyes.  Scophthalmid 
larvae  develop  extensive  head  spination.  Three  species  (Z.  punc- 
tatus,  P.  regius,  Lepidorhoinbus  whiffiagonis)  develop  paired 
otic  spines.  In  Z.  punctatus,  spines  also  develop  at  the  lateral 
aspect  of  the  midbrain  and  on  the  opercle.  Larvae  of  P.  nor- 
vegicus develop  spines  along  the  lower  jaw,  on  the  opercle  and 
preopercle,  and  at  the  shoulder  (posttemporal  region)  while  L. 
boscii  has  preopercular  spines  and  a  shoulder  cluster.  S.  ma.\- 
imus  and  S.  rhombus  have  a  supraocular  spiny  ridge,  numerous 
spines  on  the  opercle  and  preopercle  and  a  shoulder  cluster. 
Pigmentation  is  heavy  on  the  head  and  body  in  most  species. 
Z.  pimctatus  has  a  series  of  finfold  bars  and  L.  boscii  develops 
these  and  also  incomplete  bars  on  the  body.  Late  larvae  of  all 
species  develop  bars  on  the  dorsal  and  anal  fins. 

Paralichthyidae  (Figs.  344,  345).  — Three  subgroups  are  recog- 
nized in  this  family  on  the  basis  of  adult  characters:  Paralichlhys 
and  relatives  (Ancylopsella.  Gaslropsella.  Hippoglossina.  Lio- 
glosslna.  Verecundum.  Xystreurys):  Pseudorhombus  and  rela- 
tives (Cephalop.tetta.  Tarphops):  and  Cyclopsetta  and  relatives 
(Cltharlchthys.  Eiropus.  Syacium). 

In  the  first  group  larvae  are  known  for  species  oi Paralichlhys 
and  Hippoglossina  and  for  Xyslreurys  liolepis  and  in  the  second 
group  larvae  are  known  for  Pseudorhombus  and  Tarphops.  In 


652 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  175.     Meristic  and  Larval  Characters  of  Bothidae. 


Basip- 

leryg- 

2nd 

Length  al 

Larvae 

Urohyal 

lal 

Cleilhral 

Otic 

Scale 

dorsal 

transforma- 

Taxon 

described' 

Number  of  vertebrae 

spines- 

spines 

spines 

spines 

spines 

ray' 

tion  (mm) 

References 

Taeniopsettinae 

Engyophn's 

+ 

10  +  27-31  = 

=  37-41 

+  + 

+  + 

+  + 

+ 

0 

M 

-19 

Hensley,  1977 

Taeniopselta 

+ 

10  +  30-32  = 

=  40-42 

+  + 

+  + 

+  + 

+ 

0 

s 

-60 

Amaoka,  1970 

Trichopsetta 

+ 

10-11  +  30-33  = 

=  40-43 

+  + 

+  + 

+  + 

+ 

0 

s 

-28 

Futch,  1977 

Perissias 

- 

10  +  29-30  = 

=  39-40 

Bothinae 

Arnogtossus 

+ 

10-12  +  27-36  = 

=  37-48 

0 

0 

0 

0 

+ 

M,  L 

21-46 

Kyle,  1913;  Pertseva- 
Ostroumova,  1965; 
Amaoka.  1973,  1974 

Bolhus 

+ 

10  +  25-32  = 

=  35-42 

0 

0 

0 

0 

0 

M,  L 

9-42 

Kyle,  1913 

Chascanopselta 

+ 

16-18  +  37-44  = 

=  53-59 

0 

0 

0 

0 

0 

M 

78-120 

Bruun,  1937;  Nielsen, 
1963b;  Amaoka,  1971; 
Pertseva-Ostroumova, 
1971 

Crossorhomhus 

+ 

10  +  24-27  = 

=  34-37 

0 

+ 

0 

0 

+ 

M,  L 

15-20 

Ochiai  and  Amaoka,  1963 

Engyprosopon 

+ 

10  +  23-27  = 

=  33-37 

+  + 

+  + 

0.  +  + 

0 

0 

M 

16-18 

Pertseva-Ostroumova, 
1965 

Grammalobolhus 

+ 

10  +  27-28  = 

=  37-38 

+ 

+ 

0 

0 

0 

M 

-15 

Kamoharaia 

+ 

13-14  +  37-39  = 

=  50-53 

0 

0 

0 

0 

0 

M 

-91 

Nielsen.  1963c 

Laeops 

+ 

10-12  +  35-42  = 

=  46-53 

0 

0 

0 

0 

0 

L 

51-80 

Balaknshnan,  1963; 
Amaoka,  1972a;  Hubbs 
and  Chu,  1934 

Lophonectes 

+ 

10  +  32-33  = 

=  42-43 

0 

+ 

0 

0 

0 

M 

-20 

Original 

Monolene 

+ 

10  +  28-38  = 

=  38-48 

0 

0 

0 

0 

0 

L 

-30 

Futch,  1971 

Pelecanichlhys 

+ 

17  +  40  = 

=  57 

0 

0 

0 

0 

0 

M 

>90 

Struhsaker,  pers.  comm. 

Pseltina 

+ 

10  +  29-30  = 

=  35-40 

+ 

+ 

0 

0 

+ 

M 

16-20 

Pertseva-Ostroumova. 

1965;  Amaoka.  1976 


Asterorhombus 

Japonotaeops 

Mancopsetta 

Neolaeops 

Parabolhus 

Tosarhombus 


10  +  26-27  =  36-37 

11  +  41-44  =  52-53 
13-16  +  38-50  =  52-66 

13  -I-  38  =  51 
10-1-  31-36  =  41-46 
10  -I-  28-30  =  38-40 


=  yes; 


^  0  =  absent;  +  =  present;  +  +  =  strong. 


^  S  =  shon;  M  =  moderate;  L  -  long. 


these  two  subgroups  hatching,  notochord  flexion  and  meta- 
morphosis occur  at  a  small  size  (Table  1 74).  Larvae  of  these 
groups  are  noted  for  a  dorsal  crest  consisting  of  elongate  early 
forming  rays,  beginning  with  the  second  dorsal  ray  (Table  1 74). 
Larvae  of  the  Paralichthys  group  are  moderate  in  body  depth, 
with  a  deep  head  and  moderate-size  jaws.  Body  thickness  is 
moderate  except  that  Paralichthys  is  more  laterally  compressed 
than  in  other  genera  reported  (Fig.  344).  The  gut  mass  is  large. 
Preopercular  spination  consists  of  an  anterior  and  posterior  se- 
ries in  Paralichthys.  Pseudorhombus  and  Tarphops  and  an  an- 
terior series  only  in  Hippoglossina.  Larvae  of  Paralichthys  den- 
talus  have  one  to  several  minute  sphenotic  spines  (Smith  and 
Fahay.  1970)  and  P.  olivaceus  develops  a  spine  cluster  on  the 
sphenotic,  one  spine  on  the  epiotic.  and  1-2  spines  on  each  bone 
in  the  opercular  series.  Larvae  of  Pseudorhombus  pentophthal- 
inus  have  a  single  sphenotic  spine,  and  some  on  the  opercular 
bones  (Okiyama,  1 974a);  Devi  ( 1 969)  shows  two  rows  of  sphen- 
otic spines  in  P.  elevatus. 

Yolk-sac  larveofthe  Paralichthys  and  Pseudorhombus  e,Toups 
develop  moderate  to  heavy  pigmentation  with  some  on  the 
finfolds.  Later-stage  larvae  have  pigment  over  the  brain,  on  the 
lower  head  and  jaw  region  and  below  and  lateral  to  the  gut. 
Most  species  have  a  melanophore  series  along  the  dorsum  and 


ventrum.  Lateral  pigment  may  consist  of  a  series  along  the 
horizontal  septum  {Paralichthys.  Tarphops),  a  wide-spread  zone 
of  melanophores  (Xysireurys.  Hippoglossina)  or  a  posterior  bar 
(Pseudorhombus  penlophlhalmus).  Most  species  have  a  series 
of  internal  melanophores  above  the  spinal  column  and  some 
melanophores  on  the  posterior  region  of  the  finfold  and  devel- 
oping dorsal  and  anal  fins. 

Larvae  of  the  Cyclopsetta  assemblage  are  similar  morpholog- 
ically to  those  of  the  Paralichthys  and  Pseudorhombus  assem- 
blages, but  differ  in  spination  and  fin  ray  development.  The  rays 
forming  the  dorsal  crest  are  typically  longer  and  stand  out  more 
abruptly  compared  with  Paralichthys  and  associated  genera.  The 
fin  ray  complement  of  the  crest,  along  with  other  characters, 
divides  the  assemblage  into  two  generic  pairs:  Citharichthys- 
Etropus  and  Cyclopsetta-Syacium.  Species  of  the  former  group 
have  either  two  or  three  elongate  rays,  except  for  two  species 
which  lack  a  crest  altogether  (Table  1 74).  Species  of  Syacium 
have  5-8  elongate  dorsal  rays  and  8-1 1  occur  in  Cyclopsetta. 
The  left  pelvic  fin  forms  before  the  right  and  may  develop  elon- 
gate rays  in  some  species.  The  first  two  pelvic  rays  are  elongate 
in  Citharichthys  sordidus  and  C.  plalophrys.  the  second  ray  only 
is  elongate  in  C.  cornutus.  C.  gymnorhinus.  C.  spilopterus  and 
Etropus  crossotus;  C.  arctifrons.  C.  stigmaeus  and  E.  micros- 


AHLSTROM  ET  AL.;  PLEURONECTIFORMES 


653 


Fig.  346.  Larvae  of  Bothidae.  (A)  Trichopselta  vcntralis.  21.9  mm,  from  Evseenko.  1982a;  (B)  Engyophrys  senla.  12.3  mm.  from  Hensley. 
1977;  (C)  Taeniopsetla  oceltala.  59.0  mm,  from  Amaoka,  1970;  (D)  Monolene  sessiticauda.  14.3  mm,  redrawn  from  Futch,  1971;  (E)  Psettina 
hamancnsis.  4.2  mm,  from  Pertseva-Ostroumova,  1965;  (F)  P.  hamanensis.  18.1  mm.  ibid. 


loniits  lack  elongate  pelvic  rays.  The  first  three  pelvic  rays  be- 
come markedly  elongate  in  Cyclopsetta  and  the  entire  left  fin 
becomes  moderately  elongate  in  Syacium. 

Eiropus  and  Citharichthys  (except  for  C.  arctifrons)  develop 
one  or  more  rows  of  small  preopercular  spines.  According  to 
Tucker  (1982),  small  frontal-sphenotic  spines  are  present  in 
some  species  of  Cithanchlhys  and  Etropus  (6-8  spines  on  each 
side  in  C.  conmtus.  up  to  6  in  C.  gymnorhinus.  1-2  m  C.  spi- 
lopterus,  and  3-4  in  E.  crossotus).  Syacium  and  Cyclopsetta 
develop  a  series  of  large  preopercular  spines  at  the  margin  of 
the  bone  and,  in  some  species,  an  irregular  anterior  series.  The 


spine  at  the  angle  of  the  primary  series  becomes  antler-like  in 
preflexion  larvae  of  Syacium  and  in  postflexion  larvae  of  Cy- 
clopsetta. Early  preflexion  larvae  of  Syacium  develop  single 
elongate  sphenotic  spines  which  remain  prominent  during  the 
remainder  of  the  larval  period.  Sphenotic  spines  in  Cyclopsetta 
are  early-forming  but  short. 

Larvae  of  both  subgroups  of  the  Cyclopsetta  assemblage  typ- 
ically have  pigment  above  the  brain,  on  the  lower  head  region, 
below  the  gut,  lateral  to  the  posterior  region  of  the  gut,  and 
above  the  gas  bladder.  Early  preflexion  larvae  of  most  species 
have  a  series  of  small  postanal  melanophores  and  a  bar  or  a 


654 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  1 76.    Size  Data  for  Pleuronectinae  Larvae. 


Size  al 
hatching  (mm) 


Size  al  noto- 

chord  flexion 

(mm) 


Size  at 

transformation 

(mm) 


Acanthopsetta  nadeshnyi 
Athereslhes  evennanni 
A.  slomias 

Cteisthenes  herzensleini 
Embassichlhys  halhyhius 
Eopsella  grigorjewi 
E.  jordani 

Glyptocephalus  cynoglossus 
G.  stelleri 

G.  zachirus 

Hippoglossoides  dubius 
H.  elassodon 
H.  platessoides 

H.  robustus 

Hippoglossus  hippoglossus 
H.  slenolepis 

Hypsopsetta  gultutata 

Isopsella  isolepis 

Kareius  hicoloralus 

Lepidopsella  hilineata 

L.  mochigarei 

Limanda  aspera 

L.  ferrugmea 

L.  limanda 

L.  punclalissima 

L.  schrenki 

L.  schrenki  (as  Pseudopleuronectes 

yokohamae) 
L.  yokohamae 
Liopsetla  glacialis 
L.  obscura 
L.  pinnifasciata 
L.  pulnami 
Lyopsella  exilis 
Microstomiis  achne 
M.  kill 
M.  pacificus 
Parophrys  velulus 
Plalichlhys  flesus 
P.  stellatus 

P.  pallasii  (as  Platessa  quadriluberculata) 

P.  platessa 

Pleuronichthys  coenosus 

P.  cornutus 

P.  decurrens 

P.  rilien 

P.  verlicalis 

Psetlichlhys  melanostictus 
Pseudopleuronectes  americanus 
P.  herzensteini 
Reinhardtius  hippoglossoides 
Tanakius  kitaharai 
Verasper  variegatus 


<3.0 

8.4-9.9 

ca.  20-? 

<8.4 

ca.  11.5-15 

— 

— 

— 

32.9-? 

2.2-2.6 

7.0-7.9 

8.1-11.4 

ca.  9.0 

_ 

1 6.2-7 

2.5-3.0 

7.2-8.9 

11.4 

2.8 

_ 

_ 

3.5-5.6 

15-21 

25-? 

4.1-5.2 

15-17 

19-48 

ca.  6 

15.3-24.0 

49-72 

3.0-3.4  TL 

<12.4 

18.1 

5.4-6.6  TL 

9.0-10.2 

— 

4.0-6.0 

9.5-17.5 

18-30 

ca.  4.0 

ca.  I1-? 

>28.6 

6.5-7.0 

16-18 

22-34 

7.8-8.5 

13.6-17.8 

14.7-24.1 

L7-2.3 

4.0-5.2 

4.4-:>8.8 

2.7-2.9 

9.1-14.0 

15->21.9 

ca.  3.0 

ca.  4.0-9.0 

ca.  14-? 

3.4-3.8 

ca.  8.4-9.9 

>17.7 

3.95-4.48 

ca.  8.9 

_ 

2.2-2.8 

7.5-9.5 

ca.  10-? 

2.0-3.5 

5.9-ca.  10 

ca.  14 

2.7-4.0 

7-8.7 

12-20 

1.79-2.21 

— 

8.1->9.6 

ca.  2.4 

_ 

_ 

ca.  2.4 

<7.4 

1 2.0-? 

3.5-3.8 

ca.  7.0 

ca.  7.5-10.0 

3.7 

— 

— 

2.5-3.5 

-6.6 

>9.0 

3.15-3.93 

8.11-8.45 

>8.5 

3.1-3.6 

6.0-7.1 

7.3 

ca.  5.6 

9.0-10.9 

15.7-24.7 

_ 

8.8 

— 

4.84 

12-15 

18-28 

ca.  6.0 

ca.  10-15 

ca.  20->45 

2.3-2.8 

8.8-10.5 

ca.  20 

2.25 

5.9-7.1 

9-12 

1.9-2.1 

5.5-6.0 

- 

5.6 

8.9 

ca.  10.0-? 

6.0-7.5 

8.9-10.2 

10.5-14 

3.9 

6.2-8.5 

8.2->11.4 

2.65-2.8 

>3.6 

7.25-13.0 

4.9-5.5 

7.8-11.0 

10.5->21.0 

2.1 

4.3-5.6 

6.0->10.0 

2.4 

5.0-7.2 

7.3->11.0 

<3.0 

ca.  8.0 

>22.6 

2.3-3.5 

5.0-7.1 

6.8 

2.6-2.9 

ca.  6.0-8.5 

ca.  10.4-? 

10-16 

25-27 

45-65 

ca.  3.0 

— 

18.9-ca.  20 

3.8 

ca.  9-12.4 

ca.  16.4-? 

Pertseva-Ostroumova,  1961 

Pertseva-Ostroumova,  1961 

Pertseva-Ostroumova.  1961 

Okiyama  and  Takahashi,  1976;  Dekhnik,  1959 

Richardson,  1981b 

Okiyama  and  Takahashi,  1976 

Alderdice  and  Forrester,  1971 

Petersen,  1904 

Okiyama,  1963;  Dekhnik,  1959;  Pertseva- 
Ostroumova,  1961 

Original;  Ahlstrom  and  Moser,  1975 

Okiyama  and  Takahashi,  1976 

Dekhnik,  1959;  Pertseva-Ostroumova,  1961 

Petersen,  1904;  Russell,  1976;  Nichols, 
1971;  Colton  and  Marak,  1969 

Pertseva-Ostroumova,  1961 

Schmidt,  1904;  Russell,  1976  (summary) 

Thompson  and  Van  Cleve,  1936;  Pertseva- 
Ostroumova,  1961 

Sumidaet  al.,  1979 

Richardson  et  al.,  1980 

Pertseva-Ostroumova,  1961 

Pertseva-Ostroumova,  1961 

Yusa,  1958;  Okiyama  and  Takahashi,  1976 

Dekhnik,  1959;  Pertseva-Ostroumova,  1961 

Bigelow  and  Welsh,  1925;  Miller,  1958 

Russell,  1976  (summary) 

Pertseva-Ostroumova,  1961 

Hikila,  1952 

Pertseva-Ostroumova,  1961 

Yusa,  et  al.,  1971;  Minami,  1981a 

Pertseva-Ostroumova,  1961 

Pertseva-Ostroumova,  1961;  Kurata,  1956 

Pertseva-Ostroumova,  1961 

Laroche,  1981 

Original;  Ahlstrom  and  Moser,  1975 

Okiyama  and  Takahashi,  1976 

Petersen,  1904 

Original;  Ahlstrom  and  Moser,  1975 

Onginal;  Budd,  1940;  Ahlstrom  and  Moser,  1975 

Nichols,  1971;  Russell,  1976  (summary) 

Orcutt,  1950;  Yusa,  1957;  Pertseva- 
Ostroumova,  1961 

Pertseva-Ostroumova,  1961 

Nichols,  1971;  Russell.  1976  (summary) 

Sumidaet  al.,  1979;  Budd,  1940 

Takita  and  Fujita,  1964;  Minami,  1982a 

Sumida  et  al.,  1979;  Budd,  1940 

Sumida  et  al.,  1979 

Sumida  et  al.,  1979;  Ahlstrom  and  Moser,  1975; 
Budd,  1940 

Hickman,  1959 

Breder,  1923;  Laroche,  1981 

Dekhnik.  1959 

Jensen,  1935 

Okiyama  and  Takahashi,  1976 

Takita  et  al.,  1967;  Pertseva-Ostroumova, 
1961;  Uchida,  1933 


shorn  lateral  pigment  series  posteriad  on  the  tail.  In  some  species 
of  Cithanchthys  the  ventral  series  coalesces  into  a  more  sparse 
series  of  larger  spots  and  a  similar  series  develops  along  the 
dorsum  (e.g.,  C.  arctiforns.  C.  cornutus.  C.  sordidus).  In  other 
species,  series  along  the  dorsum  and  ventrum  are  abbreviated 
or  absent  and  only  the  tail  bar  may  be  present  (e.g.,  C.  gym- 


norhinus.  C.  platophrys)  or  absent  (C.  spilopterus).  Etropus  lar- 
vae have  dorsal  and  ventral  series  and  cither  a  short  lateral 
series  {E.  crossotus)  or  a  long  one  (C  inicroslomiis).  Cyclopsctta 
and  Syacnim  have  dorsal  and  ventral  series  and  a  short  lateral 
series  posteriad  on  the  tail.  Fin  pigment  is  principally  on  the 
spatulate  tips  of  the  elongate  dorsal  and  pelvic  fin  rays.  Late- 


Fig.  347.  Larvae  and  transforming  specimens  of  Bothidae.  (A)  Crossorhombus  kobensis,  16.0  mm,  from  Amaoka,  1979;  (B)  Engyprosopon 
xenandrus.  ca.  20.0  mm;  (C)  Lophonectes  gallus.  18.5  mm,  original.  K  1 38/74,  New  Zealand;  (D)  Bothits  thompsoni.  ca.  36.0  mm;  (E)  B.  mancus. 
ca.  30.0  mm.  B,  D,  and  E  from  P.  Struhsaker,  unpublished. 


656 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  348.     Larvae  of  Bothidae.  (A)  Arnoglossus  dehilis.  ca.  59.0  mm,  from  P.  Struhsaker,  unpublished;  (B)  Chascanopsetta  lugiihris.  1 20.0  mm, 
from  Amaoka,  1971;  (C)  Laeops  kilaharae.  79.0  mm.  from  Amaoka,  1972;  (D)  Pelecanichthys  sp.,  ca.  95.0  mm.  from  P.  Struhasker,  unpublished. 


AHLSTROM  ET  AL.:  PLEURONECTIFORMES 


657 


Table  177.    Size  Data  for  Larvae  of  Achirinae  and  Soleinae. 


Si7c  al 

Si/e  al 

hatching 

Size  at 

Iransformalion 

Taxon 

(mm) 

flexion  (mm) 

(mm) 

References 

Achirinae 

Achirus  linealus 

1.6 

ca.  2.5-3 

3.0-5.5 

Houdeetal.,  1970 

Trinecles  maculatiis 

1.7-1.9 

ca.  3-4 

ca.  5 

HiWebrand  and  Cable,  1938 

Soleinae 

Aesopia  cornula 

4.1 

— 

9.2-? 

Mito,  1963 

Austroglossus  nucrolepis 

1.7 

5.5-6.6 

8.5-ca.  16 

OToole.  1977;  Brownell.  1979 

Balhysolea  profundicola 

— 

— 

4.32 

Aboussouan,  1972c 

Buglossidium  lutcurn 

1.8-2.3 

ca.  6-8 

ca.  6-10 

Holt,  1891;  Ehrenbaum,  1897 

Dicologoglossa  cuneala 

1.3 

ca.  6.3-6.5 

7-7.5 

Lagardere,  1980;  Lagardere  and  Aboussouan,  1981 

Euryglossa  pan 

<2.66 

<4.6 

3.4-6.5 

Jones  and  Menon,  1951 

Heteromyctens  capensis 

1.7 

?-6.5 

ca.  6.2-? 

Brownell,  1979 

H.  japonicus 

— 

ca.  4.55 

5.0-7.0 

Minami,  1981b 

Microchirus  hoscanion 

— 

— 

7.2 

Aboussouan,  1972c 

M.  frechkopi 

— 

— 

5.68 

Aboussouan,  1972c 

M.  ocellatus 

2.0 

4.6-5.1 

6.8->8.2 

Palomera  and  Rubies,  1977 

M.  vanegalus 

2.4-2.9 

6.1->7.1 

ca.  7-12(18) 

Cunningham,  1890;  Petersen,  1909 

Pegusa  cadenali 

— 

— 

7.0 

Aboussouan,  1972c 

P.  impar 

— 

— 

8.5->12 

Padoa,  1956k 

P.  lascaris  lascaris 

<3.5 

5.3-8.1 

9.5->11.2 

Clark,  1914 

P.  lascaris  nasuta 

2.1-2.5 

_ 

_ 

Dekhnik,  1973;  Padoa,  1956k 

Solea  cuneala 

— 

— 

7.0 

Aboussouan,  1972c 

S.  heinti 

<2.2 

>2.7-3.2 

_ 

Balakrishnan  and  Devi,  1974 

S.  hexophthalma 

— 

— 

8.0 

.Aboussouan,  1972c 

S.  ovaia 

— 

ca.  3-4 

4. 5-'' 

Balaknshnan,  1963 

S.  solea 

2.5-3.8 

5.5-? 

ca.  7-14.6 

Russell,  1976  (summary) 

Synaplura  kleini 

3.0 

?-6.5 

ca.  7-9 

Brownell,  1979 

Zehrias  japonicus 

4.1  TL 

— 

— 

Mito,  1963 

Z.  zebra 

4.0  TL 

- 

- 

Mito,  1963 

stage  larvae  of  most  species  develop  chevron-shaped  bars  on 
the  epaxial  and  hypaxial  myosepta.  Metamorphosing  specimens 
of  Cyclopsetta  have  series  of  large  ocelli  on  the  dorsal  and  anal 
fins.  Dorsopsetta  norma  described  by  Nielsen  (1963b)  on  the 
basis  of  two  metamorphosing  specimens  is  apparently  a  species 
of  Cyclopsetta. 

Bolhidae  (Figs.  346-348).  — Two  bothid  subfamilies  are  recog- 
nized, Taeniopsettinae  and  Bothinae.  Bothid  larvae  are  thin- 
bodied  to  diaphanous,  sparsely  pigmented,  and  all  develop  an 
elongate  second  dorsal  ray  (Table  1 75).  Also,  spines  may  appear 
on  the  urohyal,  basipterygia,  cleithra  and  epiotics  in  a  pattern 
which  is  generally  consistent  for  subfamilies  and  genera  (Table 
175).  Bothid  larvae  reach  a  relatively  large  size  before  meta- 
morphosis. Early  larval  stages  are  often  poorly  represented  in 
collections. 

Larval  series  are  known  for  all  taeniopsettine  genera,  except 
Perissias.  Larvae  of  Tnchopseita  and  Engyophrys  are  ovate  while 
those  of  Taemopsetta  are  round  (Fig.  346).  All  have  a  complete 
complement  of  head  spines  (Table  1 75).  The  second  dorsal  fin 
ray  is  slightly  or  moderately  elongate.  Taeniopsetta  lacks  me- 
lanophores,  but  live  larvae  have  four  reddish-orange  spots  along 
the  bases  of  the  dorsal  and  anal  fins,  and  orange,  reddish  and 
yellow  blotches  and  bands  on  the  body  and  head.  Tnchopsclla 
has  three  series  of  melanistic  blotches  along  the  dorsal  and  anal 
pterygiophores  and  along  the  body  axis  (left  side).  Engyophrys 
lacks  melanophores. 

Larvae  of  Bothinae  have  an  ovate,  round,  or  elongate  shape 
(Figs.  347,  348)  and  lack  epiotic  spines.  Engyprosopon  has  nu- 
merous urohyal  and  basipterygial  spines  and  some  species  have 
numerous  spines  on  the  cleithrum.  Psellina  and  Grammato- 


hothiis  have  urohyal  and  basipterygial  spines,  and  early  larvae 
of  the  former  have  a  hook-like  projection  on  the  lower  jaw  (Fig. 
346).  Crossorhombiis  and  Lophonectes  have  basipterygial  spines 
only  and  all  other  known  bothid  larvae  lack  head  spines.  Cros- 
sorhombus  larvae  have  a  series  of  scale  spines  along  the  bases 
of  the  dorsal  and  anal  fins,  one  scale  per  ray,  and  species  of 
Pseltina  and  .-irnoglossus  also  develop  such  scale  spines.  In  the 
species  of  .•lr«(),^/aM(«  described  by  Kyle  (1913),  patches  of  scale 
spines  develop  on  the  median  and  ventral  regions  of  the  ab- 
domen. The  second  dorsal  ray  is  usually  moderately  elongate 
but  can  be  greatly  elongate  and  ornamented,  as  in  Arnoglossus. 

Pigmentation  is  sparse  in  most  bothine  larvae  and  lacking  in 
some  species.  Exceptions  are  found  in  species  of  .Arnoglossus 
and  Psettina  which  usually  have  melanophores  above  the  brain, 
ventrally  on  the  gut,  above  the  gas  bladder,  in  series  along  the 
dorsal  and  ventral  midlines,  and  along  the  horizontal  septum; 
in  some  species  a  complete  or  partial  bar  is  present  posleriad 
on  the  tail.  Preflexion  larvae  of  Bothus  have  a  melanistic  blotch 
near  the  tip  of  the  notochord;  later  larval  stages  are  unpig- 
mented,  except  that  transforming  specimens  of  some  species, 
B.  myriaster  (Amaoka,  1964)  and  B.  mancus(Fig.  347),  become 
heavily  spotted  over  the  body  and  fins,  Laeops  has  melanistic 
blotches  forming  an  irregular  pattern  over  the  body  and  median 
fins. 

Monolene  shares  some  adult  characters  with  taeniopsettines 
but  larval  characters  place  it  with  the  bothines.  Larvae  are  elon- 
gate, lack  head  spines,  have  an  elongate  ornamented  second 
dorsal  ray,  and  melanistic  pigment  above  the  gut,  on  the  right 
side  of  the  brain  and  on  the  dorsal  fin  membrane  (Fig.  346). 

Pleuronectidae   (Figs.    349-355).— Of   the    five    pleuronectid 


658 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


AHLSTROM  ET  AL.:  PLEURONECTIFORMES 


659 


subfamilies  recognized  (see  introduction),  the  Pleuronectinae  is 
the  largest  with  26  genera,  representing  %  of  the  genera  in  the 
family.  Three  contributions  that  summarize  egg  and  larval  in- 
formation for  pleuronectine  flatfishes  from  the  eastern  North 
Atlantic  and  Mediterranean  are  Ehrenbaum  (1905-1909),  Pa- 
doa  (1956k).  and  Nichols  (1971).  Bigelow  and  Welsh  (1925). 
Bigelow  and  Schroeder  (1953),  Martin  and  Drewry  (1978),  and 
Fahay  (1983)  give  information  on  eggs  and  larvae  of  western 
Atlantic  pleuronectine  flatfishes.  The  most  comprehensive  work 
dealing  with  early  life  history  stages  of  flatfishes  from  the  western 
North  Pacific  is  Pertseva-Ostroumova  (1961). 

Yolk-sac  larvae  of  pleuronectine  flatfishes  can  be  as  small  as 
1.7  mm  (Hypsopseila  giittulata)  or  as  large  as  10-16  mm  (Rein- 
hardtius  hippoglossoides)  and  size  at  hatching  is  a  primary  char- 
acter for  identifying  yolk-sac  larvae  (Table  1 76).  The  pigment 
pattern  can  be  quite  distinctive,  as  for  example  in  the  genus 
Pleuronichthys.  but  in  many  pleuronectines  the  body  pigment 
migrates  during  the  yolk-sac  stage,  and  is  variable  from  speci- 
men to  specimen  of  the  same  species.  The  yolk  sac  itself  can 
lack  pigment  (as  in  Parophrys  vetulus.  Hippoglosstis  stenolepis 
or  Eopsetta  jordani),  can  be  moderately  pigmented  (as  in  Lyop- 
setta  exilis,  Lepidopsetta  bilineata  or  Psettichthys  melanostictus) 
or  can  be  heavily  pigmented  (as  in  Pleuronichthys  decurrens  or 
I'erasper  variegaliis).  Similarly  the  finfold  can  lack  pigment  or 
be  variously  pigmented  and  useful  in  identification. 

Early  preflexion  pleuronectine  larvae  are  slender;  the  head  is 
of  moderate  size;  snout-anus  length  can  be  as  much  as  50%  NL 
(as  in  four  species  oi Pleuronichthys  larvae,  Sumida  et  al.,  1979) 
but  usually  is  shorter  (i.e.,  35-45%  NL).  The  gut  is  initially 
straight  but  develops  a  coil  soon  after  the  completion  of  yolk 
absorption.  Greatest  body  depth  after  the  gut  becomes  looped 
is  either  at  the  anus  or  slightly  anterior  to  it.  Body  shape  of 
preflexion  larvae  is  quite  similar  from  species  to  species.  There 
are  few  distinctive  characters  unique  to  the  larval  period  of 
pleuronectine  flatfishes.  Only  a  few  kinds  of  pleuronectine  larvae 
develop  head  spination.  Preopercular  spines  form  in  larvae  of 
Athercslhcs.  Glyptocephalus,  Tanakiiis  and  E opsei! a;  oiic  spines 
develop  on  larvae  of  Microstomus  (at  least  on  2  species),  Hyp- 
sopsetta,  and  Pleuronichthys  ( 1  species);  Athercsthes  has  a  spi- 
nous supraocular  crest.  Head  spination  develops  during  the  pre- 
flexion stage,  but  usually  is  best  developed  on  flexion  or  early 
postflexion  larvae. 

The  caudal  fin  begins  forming  either  slightly  before  or  together 
with  the  dorsal  and  anal  fins.  The  first  caudal  supporting  bones 
to  form  as  cartilage  are  the  hypurals.  Usually  several  caudal  rays 
(2  +  2  or  3  +  3)  are  formed  before  flexion  begins.  In  late  flexion 
and  early  postflexion  larvae,  the  end  of  the  notochord  can  project 
beyond  the  hypural  plates.  The  complete  complement  of  caudal 
rays  is  usually  laid  down  during  the  flexion  period. 

The  dorsal  and  anal  fins  form  in  the  finfold  at  some  distance 
from  the  main  part  of  the  body.  The  intervening  space  becomes 
filled  with  the  pterygiophores  that  support  the  dorsal  and  anal 
fin  rays,  causing  an  increase  in  body  depth.  In  both  dorsal  and 
anal  fins  the  rays  begin  forming  at  the  anterior  ends  of  the  fins 
and  the  differentiation  proceeds  posteriad.  The  first  few  rays  in 
both  fins  are  reduced  in  size  and  the  terminal  ray  is  often  minute. 


Pelvic  fin  buds  usually  form  during  the  flexion  stage  but  pelvic 
rays  usually  are  not  developed  until  the  postflexion  stage.  As  in 
all  flatfishes,  formation  of  pectoral  fin  rays  is  delayed  to  the  end 
of  the  transformation  stage. 

The  vertebral  processes  ossify  before  the  centra.  In  the  caudal 
group  of  vertebrae,  ossification  of  haemal  and  neural  processes 
proceeds  posteriad.  Ossification  of  abdominal  neural  processes 
can  follow  several  patterns,  but  usually  proceeds  anteriad.  The 
last  neural  and  haemal  processes  to  ossify  are  the  truncate  spines 
of  the  2  or  3  vertebrae  anterior  to  the  urostyle.  Centra  ossify 
initially  at  the  bases  of  neural  and  haemal  processes  and  ossi- 
fication proceeds  peripherally  until  a  complete  ring  is  formed. 
On  first  formation  only  the  middle  portion  of  a  vertebral  cen- 
trum is  ossified,  hence  the  space  between  adjacent  centra  may 
be  as  wide  as  the  ossified  portions  of  the  centra.  The  ural  centra 
are  the  first  to  ossify  in  some  pleuronectines  or  they  can  ossify 
at  the  same  time  as  other  centra.  The  last  centra  to  form  are 
those  of  the  2  (or  3)  vertebrae  anterior  to  the  urostyle. 

All  pleuronectine  larvae  that  have  been  described  have  body 
pigment.  The  pigment  pattern  changes  with  growth,  often  mark- 
edly. Also,  there  is  often  considerable  variation  in  pigmentation 
of  larvae  of  similar  sizes  of  the  same  species.  Notwithstanding, 
body  and  finfold  pigment  constitutes  a  primary  character  for 
identification  of  flatfish  larvae  during  the  preflexion  stage. 

To  show  the  variety  of  pigment  patterns  found  on  preflexion 
stage  pleuronectine  larvae,  preflexion  larvae  of  1 7  species  from 
the  North  Pacific  are  illustrated  (Figs.  349-351).  Heavily  pig- 
mented larvae  are  in  the  genera  Pleuronichthys.  Hypsopsetta, 
and  I'erasper  (Fig.  349).  The  posterior  portion  of  the  tail  is 
unpigmented  or  pigment  is  confined  to  marginal  spots  along  the 
notochord.  The  unpigmented  tail  area  is  more  extensive  in  some 
species  than  in  others.  Finfold  pigment  is  very  useful  in  iden- 
tifying these  larvae  to  species  taken  in  conjunction  with  larval 
size  and  extent  of  tail  pigment. 

In  the  other  14  kinds  of  larvae  representing  as  many  genera, 
tail  pigment  appears  in  a  number  of  patterns.  The  larvae  illus- 
trated in  Figs.  350  and  35  1  are  arranged  in  the  order  of  increasing 
complexity.  In  the  simplest  pattern  pigment  is  concentrated 
along  the  ventral  midline  with  only  moderate  dorsal  or  lateral 
pigment,  as  in  Hippoglossus  stenolepis  or  Reinhardtius  hippo- 
glossoides. Although  Parophrys  vetulus  and  Lyopsetta  exilis  have 
more  ventral  margin  pigment  than  dorsal,  it  is  almost  contin- 
uous on  both  margins.  Platichthys  stellatus  has  more  diffused 
pigment  over  the  tail  portion  of  the  body,  but  it  is  not  in  a 
pattern.  The  most  unusual  pigment  is  found  in  Atherestes.  There 
are  two  conspicuous  dorsal  patches  as  opposed  to  almost  no 
ventral  pigment.  Pigment  on  Eopsetta  jordani  is  limited  to  a 
mid-tail  band  and  a  terminal  notochord  patch.  A  more  common 
pattern  is  encountered  in  Isopsetta,  which  has  two  pigment  bands 
across  the  tail  together  with  the  terminal  notochord  pigment.  A 
basically  similar  pattern  is  found  in  Lepidopsetta  bilineata.  Pset- 
tichthys is  unusual  in  having  alternating  dorsal  and  ventral 
blotches.  Hippoglossoides  elassodon  has  three  tail  pigment  areas 
(i.e.,  opposing  dorsal  and  ventral  pigment  patches)  together  with 
terminal  notochord  pigment.  This  is  also  the  basic  pattern  in 
Microstomus.  Emhassichthys  increases  opposing  tail  patches  to 


Fig.  349.  Larvae  and  transforming  specimens  of  Pleuronectidae.  (A)  Pleuronichthys  coenosus.  3.7  mm,  from  Sumida  et  al.,  1979;  (B)  P. 
coenosus.  8.9  mm.  ibid;  (C)  Hypsopseila  giitndala.  2.6  mm,  ibid;  (D)  //.  giittulata.  6.6  mm,  ibid;  (E)  I'erasper  variegaliis.  5.6  mm,  from  Pertseva- 
Ostroumova.  1961  after  Uchida,  1933;  (¥)  \'.  variegatus.  12.4  mm.  ibid. 


660 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES- AHLSTROM  SYMPOSIUM 


four  plus  terminal  notochord  pigment  (Richardson,  1981b); 
G/yptocephalus  zachirus  has  pigment  bands  which  alternate  with 
ventral  patches,  plus  the  terminal  notochord  pigment. 

At  least  four  other  genera  of  pleuronectine  flatfishes  occur  in 
the  eastern  North  Pacific.  The  preflexion  stage  larvae  of  Pleu- 
ronectes  pallasii,  Liopsetta  glacialis.  and  Limanda  aspera  lack 
melanistic  bands  (Pertseva-Ostroumova,  1961).  Larvae  are  un- 
known for  the  fourth  genus,  Clidoderma. 

Larvae  are  known  for  species  representing  six  additional  gen- 
era in  the  western  North  Pacific.  According  to  Pertseva-Os- 
troumova (1961),  preflexion  larvae  o{  Acanthopsetta  nadeshnyi 
and  Kareius  bicoloratiis  lack  bands;  those  of  Cleisthenes  her- 
zensteini  (see  also  Okiyama  and  Takahashi,  1976),  Pseudopleu- 
ronectes  herzensteini  and  P.  yokohamae  (see  also  Dekhnik,  1 959; 
Yusa,  1960a,  b;  Yusa  et  al.,  1971)  have  two  tail  pigment  bands 
plus  terminal  notochord  pigment.  Preflexion  larvae  of  I  erasper 
vahegatus  (Fig.  349)  are  as  heavily  pigmented  as  those  of  Pleii- 
ronichthys  (Takita  et  al.,  1967;  Uchida,  1933).  The  pigment 
pattern  on  preflexion  larvae  of  Tanakius  kitaharai  is  very  sim- 
ilar to  that  on  larvae  of  Glyplocephalus  stelleh  (Okiyama  and 
Takahashi,  1976).  Larvae  have  not  been  described  for  the  mono- 
typic  genus  Dexistes. 

In  species  with  banded  preflexion  larvae,  the  bands  usually 
persist  into  later  larval  stages;  those  with  diffuse  or  linear  pig- 
ment patterns  generally  do  not  develop  bands  in  later  stages, 
although  pigment  may  become  associated  with  myosepta  (Figs. 
352,  353).  Virtually  all  late  postflexion  and  metamorphic  pleu- 
ronectines  develop  a  distinct  pattern  of  bars  or  blotches  on  the 
body  and  median  fins,  which  persists  into  the  juvenile  stage 
(Fig.  354). 

Of  the  four  other  pleuronectid  subfamilies,  larvae  have  not 
been  described  for  Paralichthodinae,  while  some  information  is 
available  on  the  Samarinae,  Poecilopsettinae,  and  Rhomboso- 
leinae.  Pertseva-Ostroumova  (1965)  described  two  larval  spec- 
imens (6.4,  8.7  mm)  of  Samaris  cnslatus  and  Struhsaker  (pers. 
comm.)  has  described  large  pelagic  larvae  of  Samariscus  sp.  and 
Poecilopsetta  hawaiiensis  (Fig.  355).  Larvae  of  S.  cristatus  are 
deep-bodied  in  the  gut  region,  have  a  relatively  large  head  and 
jaws  and  a  pigment  pattern  consisting  of  melanophore  patches 
along  the  dorsum  and  ventrum.  along  the  outer  margins  of  the 
pterygiophore  zones,  and  along  the  dorsal  and  anal  fins;  the 
ventral  region  of  the  gut  is  pigmented.  A  series  of  Samariscus 
triocellatus,  7.3-19.0  mm  (provided  by  Dr.  T.  A.  Clarke,  Univ. 
of  Hawaii),  is  similar  to  Samaris  cristatus  in  having  a  slender 
body  and  wide  pterygiophore  zones  but  the  gut  coil  is  elongate, 
protrudes  beyond  the  ventral  profile,  and  the  fourth  dorsal  ray 


is  elongate.  The  left  eye  has  begun  to  migrate  at  7.3  mm  and  is 
at  the  dorsal  midline  by  12.0  mm.  Larvae  of  Samariscus  cor- 
allimts  are  similar  but  attain  a  larger  size  (ca.  26  mm).  Both 
species  lack  pigment.  Late  postflexion  larvae  of  Poecilopsetta 
have  a  body  form  similar  to  samarines  (slender  body  with  wide 
pterygiophore  /ones)  but  have  a  different  gut  structure,  no  elon- 
gate dorsal  ray,  and  have  a  striking  pigment  pattern  consisting 
of  dorsal  and  ventral  myoseptal  series  and  large  blotches  over 
the  pterygiophore  zones,  dorsal  and  anal  fins,  and  gut  (Fig.  355). 
A  29-mm  late  postflexion  larva  from  the  North  Atlantic  has  a 
pigment  pattern  identical  to  Hawaiian  specimens. 

Reared  yolk-sac  and  early  preflexion  larvae  of  rhombosoleine 
species  have  been  illustrated  and  briefly  described:  Ammotretis 
rostratus  (Thomson,  1906);  Rhomhosolea  plebeia  (Anderton, 
1907);  Colistnim  glint  hen.  Pclotretis  flavilatus.  and  Peltorham- 
phus  novaezeclandiae  (Thomson  and  Anderton,  1921).  The  oil 
globules  remain  evenly  dispersed  throughout  the  yolk-sac  pe- 
riod. Heavy  melanistic  pigmentation  develops  on  the  head,  body, 
yolk  sac,  and  finfold.  Late  yolk-sac  larvae  of  C  gunthcri  deve\op 
an  unusual  lobate  projection  of  the  dorsal  finfold,  which  extends 
well  anterior  to  the  head.  A  similar  structure  appears  in  yolk- 
sac  larvae  of  the  soleid,  Pcgusa  lascaris  (Holt,  1891).  Rapson 
( 1 940)  described  and  illustrated  with  photographs  a  reared  series 
of  Pelot  ret  is  Jlavilat  us.  Flexion-stage  larvae  of  this  species  are 
deep-bodied  and  similar  in  appearance  to  paralichthyids,  al- 
though they  lack  elongate  dorsal  fin  rays  (Fig.  355).  Pigmenta- 
tion consists  of  dorsal  and  ventral  midline  series,  series  above 
and  below  the  spinal  column,  a  linear  patch  below  the  gut,  and 
embedded  melanophores  in  the  otic  region.  Postflexion  larvae 
become  mottled  with  large  blotches  on  the  body  and  fins.  Cross- 
land  (1981)  briefly  described  and  illustrated  pre-  and  postflexion 
stages  of  a  similar  larva  which  he  identified  as  Pe/torhamphus 
latus  and  stated  that  Rapson's  ( 1 940)  series  was  a  species  of 
Peltorhamphus.  Crossland's  (1982)  illustration  of  a  flexion-stage 
Pelotrelis  flavilatus  has  heavy  pigmentation,  a  protruding  gut 
mass  and  looks  very  much  like  a  soleid. 

Soleidae  (Fig.  356).— Two  subfamilies,  Soleinae  and  Achirinae, 
are  recognized  in  the  family.  In  the  Soleinae,  life  history  stages 
are  well  known  for  the  eastern  North  Atlantic  species,  Solea 
solea.  Microchirus  varicgaius.  Buglossidium  luteum  and  Pcgusa 
lascaris  (references  summarized  in  Ehrenbaum,  1905-1909  and 
Russell,  1976).  A  comprehensive  volume  on  the  development 
of  5.  solea  was  produced  by  Fabre-Domergue  and  Bietrix  (1905). 
Padoa  (1956k)  summarized  information  on  eggs  and  larvae  of 
soles  from  the  Mediterranean,  and  Aboussouan  (1972c)  briefly 


Fig.  350.  Larvae  of  Pleuronectidae.  (A)  Hippoglossus  stenolepis.  1 5.0  mm,  from  Pertseva-Ostroumova.  1 96 1 ;  (B)  Reinhardlius  hippoglossoides. 
17.0  mm,  from  Jensen,  1935;  (C)  Lyopselta  exilis.  5.9  mm  from  Ahlstrom  and  Moser,  1975;  (D)  Parophrys  vetulus.  4.3  mm,  ibid;  (E)  Ptatichthys 
slellatus.  2.6  mm,  from  Orcutt,  1950;  (F)  Atheresthes  stomias.  10.5  mm,  original;  (G)  Eopsetta  jordani.  6.2  mm,  from  Alderdice  and  Forrester, 
1971. 

Fig.  351.  Larvae  of  Pleuronectidae.  (A)  Isnpsetta  isolepis.  9.5  mm,  original,  CalCOFI  7205,  Sta.  40.38;  (B)  Lepidopsena  bihneata.  4.6  mm, 
from  Pertseva-Ostroumova,  1965;  (C)  Pseltichlhys  melanoslictus.  6.7  mm,  original,  CalCOFI  5807  Sta.  40.38;  (D)  Hippoglossnide.s  eta.ssodon. 
9.2  mm,  from  Pertseva-Ostroumova,  1961;  (E)  Microstornus  pacificus.  7.0  mm,  redrawn  from  Ahlstrom  and  Moser,  1975;  (F)  Embassichlhys 
balhybius.  18.5  mm,  original,  CalCOFI  4905,  Sta.  29.83;  (G)  Glyplocephalus  zachirus.  22.8  mm,  redrawn  from  Ahlstrom  and  Moser,  1975. 


Fig.  352.  Larvae  of  Pleuronectidae.  (A)  Lyopselta  exilis.  14.7  mm,  original.  CalCOFI  7805,  Sta.  100.29;  (B)  Parophrys  vetulus.  16.0  mm, 
redrawn  from  Ahlstrom  and  Moser,  1975;  (C)  Isopsetla  isolepis.  14.2  mm,  original,  CalCOFI  7205,  Sta.  40.38,  (D)  Eopsetta  grigorjewi,  lO.O  mm, 
from  Okiyama  and  Takahashi,  1976;  (E)  Pseltichlhys  melanoslictus.  9.4  mm,  original,  CalCOFI. 


AHLSTROM  ET  AL.:  PLEURONECTIFORMES 


661 


B 


D 


G 


662 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


B 


D 


AHLSTROM  ET  AL.:  PLEURONECTIFORMES 


663 


f^  ,--V      ,    *    . 


1 1  :^i;iiri'<<'*>ii:<>^<;^i^^  : 


■  "i^r.pr--  • 


664 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  178.    Size  Data  and  Number  of  Elongate  Dorsal  Rays  for  Larvae  of  Symphurinae  and  Cynoglossinae. 


Number  of 

Size  al 

elongate 

hatching 

size  al  flexion 

Size  al  trans- 

Taxon 

dorsal  rays 

(mm) 

(mm) 

formalion  (mm) 

References 

Symphurinae 

Svmphuriis  atricaiida 

5 

1.9 

9.4-10.8 

19-24.2 

Onginal 

S.  lacleus 

4-5 

— 

6-<10 

18 

Kyle,  1913 

S.  ligulala 

5 

— 

<10.5 

32 

Kyle,  1913 

S.  orienlalis 

5 

— 

>4.4-<9.3 

>12.3 

Pertseva-Ostroumova,  1965 

S.  ptagiusa 

4-5 

<1.3 

<6.2 

<13 

Hildebrand  and  Cable,  1930; 
Olney  and  Grant,  1976 

Cynoglossinae 

Cynogtossus  ahbrevialus 
(as  Areliscus  trigrammus 


3.2 


Fujita  and  Takila,  1965 


C.  arel  (as  C.  otigolepis) 

2 

— 

— 

— 

Pertseva-Ostroumova,  1965 

C.  btlineatus 

2 

— 

>4.0 

6-8 

Vijayaraghavan,  1957; 
Pertseva-Ostroumova,  1965 

C.  brevis 

— 

— 

— 

ca.  4.0 

Balakrishnan  and  Devi,  1974 

C.  capensis 

2-A 

1.2 

ca.  9.9 

10-15 

Brownell,  1979 

C.  cynogtossus 

2 

<1.6 

ca.  4. 1 

<4.7 

Balakrishnan  and  Devi,  1974 

C.  kopsi  (as  C.  sibogae) 

2 

— 

7.6 

— 

Pertseva-Ostroumova,  1965 

C.  lida 

2 

<2.1 

>4.6 

— 

Balakrishnan  and  Devi,  1974 

C.  lingua 

— 

_ 

— 

17.7-? 

Jones  and  Menon,  1951 

C.  macroslomus 

2 

<2.5 

4.2 

ca.  5 

Seshappa  and  Bhimachar,  1955 

(as  C.  semifasciatus) 

C.  monopus 

2 

— 

5-7 

7.0 

Balakrishnan,  1963 

C.  puncliceps 

2 

<1.4 

ca.  4.2-? 

ca.  4-5 

Balakrishnan  and  Devi,  1974 

C.  robustus 

2 

1.85 

— 

_ 

Fujitaand  Uchida,  1957 

C.  semifasciatus 

2 

<2.0 

7.2-11 

11-12.5 

Balaknshnan,  1961 

Paraplagusia  japonica 

2 

- 

<10.2 

ca.  12.2 

Minami,  1982b 

described  several  species  from  off  west  Africa.  Life  history  seires 
have  been  described  for  two  achirine  species  of  the  western 
North  Atlantic,  Trinectes  maculatus  and  Achirus  lineatus.  Eggs 
of  achirines  are  smaller  than  in  most  soleines  (Table  172)  and, 
accordingly,  size  at  hatching  is  also  smaller;  achirines  and  some 
soleines  undergo  notochord  flexion  and  transformation  at  vei^ 
small  sizes  (Table  1 77).  Achirines  are  deep-bodied,  with  a  large 
gut  that  occupies  a  major  portion  of  the  body  volume,  a  large 
deep  head  with  a  distinct  dorsal  hump;  eyes  and  jaws  are  large 
(Fig.  356).  Preflexion  larvae  of.-J.  //«ea/M5  develop  spinous  ridges 
above  the  eye  (frontal  bone),  at  the  otic  region  (parietal  and 
autopterotic  bones)  and  on  the  preopercle.  Also,  five  rows  of 
papilla-like  spines  develop  on  the  body.  Larvae  of  T.  maculatus 
develop  bony  ridges  on  the  frontal,  parietal  and  autopterotic 
bones.  A.  lineatus  larvae  are  unique  among  described  soleids  in 
having  an  elongate  third  dorsal  ray.  Early  larvae  of  .-1.  lineatus 
are  unpigmented  but  by  late  preflexion  stage  have  developed 
pigment  on  the  head,  gut,  elongate  dorsal  ray,  dorsal  and  ventral 
body  margins  and  blotches  on  the  dorsal  and  anal  fins.  Early 
larvae  of  T.  maculatus  are  heavily  pigmented  and  have  three 
large  blotches  in  the  dorsal  finfold  and  two  in  the  ventral  finfold. 
In  later  larvae  these  blotches  become  dusky  bars  that  overlie 
the  nearly  solid  background  pigment. 

Soleines  have  a  large  head  and  jaws  as  in  achirines  but  the 
eye  is  relatively  smaller  and  the  dorsal  hump  is  less  prominent 


(Fig.  356).  Also,  soleines  are  less  deep-bodied  and  the  gut  oc- 
cupies a  relatively  smaller  portion  of  the  body  mass;  in  many 
soleine  species  the  rounded  gut  mass  protrudes  well  beyond  the 
ventral  profile.  Pigmentation  is  highly  varied  ranging  from  species 
of  Aseraggodes  which  lack  pigment  to  species  such  as  Solea 
solea.  Pegusa  lascans,  Microchirus  variegatus  and  Euryglossa 
pan  which  are  solidly  covered  with  melanophores.  A  typical 
pattern  appearing  in  many  described  species  consists  of  a  series 
of  melanophores  along  the  dorsum,  ventrum  and  horizontal 
septum,  and  melanophores  on  the  head,  gas  bladder  and  fin- 
folds  (Fig.  356). 

Cynoglossidae  {Fig.  357).  — Two  subfamilies,  Symphurinae  and 
Cynoglossinae,  are  recognized  in  the  family.  The  first  larval 
descriptions  of  the  former  are  of  Symphurus  lacleus,  S.  ligulata 
and  S.  pusilla  (Kyle,  1913).  Hildebrand  and  Cable  (1930)  de- 
scribed a  series  as  5.  plagiusa.  but  Olney  and  Grant  (1976) 
described  a  different  series  as  S.  plagiusa  and  pointed  out  that 
Hildebrand  and  Cable's  descriptions  must  refer  to  another 
species.  Pertseva-Ostroumova  ( 1 965)  ascribed  a  larval  series  to 
S.  orientalis  and  we  have  identified  eggs  and  larvae  of  .S'.  atn- 
cauda.  Larval  series  or  metamorphosing  specimens  have  been 
ascribed  to  at  least  1 1  types  of  cynoglossines;  however,  most  of 
these  are  incomplete  series  and  identifications  are  tentative  (Ta- 
ble 178).  Most  cynoglossids  are  less  than  2.5  mm  at  hatching; 


Fig.  353.  Larvae  of  Pleuronectidae.  (A)  Glyptocephatus  zachinis,  48.7  mm,  redrawn  from  Ahlstrom  and  Moser,  1975;  (B)  G.  stelleri.  24.6 
mm,  from  Okiyama  and  Takahashi,  1976;  (C)  Tanakius  kitaharai.  15.8  mm,  ibid;  (D)  Microstomus  achne.  8.8  mm,  ibid;  (E)  Embassichthys 
balhybius.  16.2  mm,  from  Richardson,  1981b. 


AHLSTROM  ET  AL.:  PLEURONECTIFORMES 


665 


666 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Fig.  354.    Transforming  specimens  of  Pleuronectidae.  (A)  Hippoglossus  stenolepis.  24.0  mm,  original;  (B)  Eopseltajordam.  16.2  mm,  CalCOFl 
5104,  Sta.  70.55;  (C)  Lyopsetta  exilis.  22.0  mm,  from  Ahlstrom  and  Moser,  1975;  (D)  Pleuronichthys  ritteri,  10.0  mm,  from  Sumida  et  al.,  1979. 


i 


AHLSTROM  ET  AL.:  PLEURONECTIFORMES 


667 


Fig.  355.     Larvae  and  transforming  specimens  of  Pleuronectidae.  (A)  Samaris  crislatm,  6.4  mm,  from  Pertseva-Ostroumova,  1965;  (B)  5. 

crislatus.  8.7  mm,  ibid;  (C)  Samariscus  sp.,  ca.  24.0  mm;  (D)  Poecilopsetta  hawaiiensis.  ca.  29.0  mm;  (E)  Pelolretis  llavilalus.  4.3  mm,  redrawn 
from  Rapson,  1940.  C  and  D  from  P.  Struhsakcr.  unpublished. 


an  exception  is  C.  abbreviatus  which  has  a  relatively  large  egg. 
Notochord  flexion  and  transformation  occur  at  larger  sizes  in 
symphurines  compared  with  cynoglossines  and  some  Symphii- 
rus  have  an  extended  larval  stage  that  exceeds  30  mm  in  length 


in  having  a  large  deep  head  and  tapering  body,  but  the  jaws  are 
relatively  smaller  in  cynoglossids  and  the  body  is  more  attenuate 
(Fig.  357).  The  gut  mass  protrudes  beyond  the  ventral  profile 
and  in  some  species  it  trails  posteriad.  In  S.  laclea  a  conical 


(Table  178).  Cynoglossid  larvae  are  similar  to  those  of  soleids      structure  is  attached  to  the  trailing  gut  coil  (Kyle,  1913).  Cy- 


668 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


c 


IV^ifi^V^^*^  • -^'/^i^s. 


D 


Fig.  356.  Larvae  and  transforming  specimens  of  Soleidae.  (A)  Trinecles  maculalus.  2.0  mm.  from  Hildebrand  and  Cable,  1938;  (B)  Achirus 
linealus.  3.1  mm.  from  Houde  et  al..  1970;  (C)  Solea  solea.  7.5  mm,  from  Ehrenbaum,  1905-1909;  (D)  MUrochirus  vanegatus.  10.0  mm,  from 
Petersen,  1909;  (E)  Eun-glossa  pan.  4.6  mm,  from  Jones  and  Menon,  1951;  (F)  Solea  mala.  4.7  mm,  from  Jones  and  Pantulu,  1958;  (G) 
Microchirus  ocellatus.  5.1  mm,  from  Palomera  and  Rubies,  1977;  (H)  .iuslroglossus  microlepis.  6.6  mm,  from  O'Toole,  1977;  (I)  Heleromycteris 
japonicus.  4.9  mm.  from  Minami.  1981b;  (J)  .Aseraggodes  whitakeri.  ca.  27.0  mm,  from  P.  Struhsaker,  unpublished. 


AHLSTROM  ET  AL.:  PLEURONECTIFORMES 


669 


'^'^aa^ 


H 


~'^^.  ■■  1 


o^-      •■ 


Fig.  357.  Larvae  of  Cynoglossidae.  (A)  Cynoglossus  abbreviatus.  5.0  mm.  from  Fujita  and  Takita.  1965;  (B)  C  monopus.  7.0  mm.  from 
Balaknshnan.  1963;  (C)  C.  macrostomus.  4.5  mm,  from  Seshappa  and  Bhimachar.  1955;  (D)  Syinphurus  hgulala.  10.5  mm.  from  Kyle.  1913; 
(E)  S.  alncauda.  4.0  mm.  onginal.  CalCOFI;  (F)  S  atncauda,  6.5  mm.  onginal.  CalCOFl;  (G)  S.  alncauda.  12.8  mm.  original,  CalCOFI;  (H)  5. 
plagiusa.  6.2  mm,  redrawn  by  Fahay  (1983)  from  Olney  and  Gram.  1976;  (I)  5.  lactea.  18.0  mm.  from  Padoa,  1956k. 


noglossid  larvae  develop  a  crest  consisting  of  elongate  anterior 
dorsal  rays.  2  rays  in  Cynoglossus  and  usually  4  or  5  in  Syin- 
phurus. Pectoral  fins  are  present  during  the  larval  period,  but 
do  not  develop  rays  and  disappear  at  metamorphosis.  One 


species,  S.  ligulata.  develops  elongate  third  and  fourth  pelvic 
rays  (Kyle.  1913;  Padoa.  1956k). 

Pigmentation  in  early  larvae  of  Cynoglossus  consists  of  4-5 
opposing  blotches  along  the  dorsum  and  ventnim.  pigment  on 


670 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


the  head,  gut  and  gas  bladder.  In  some  species,  large  blotches 
in  the  finfold  distal  to  the  dorsal  and  ventral  midline  blotches 
give  the  larvae  a  barred  appearance.  In  later  stages  the  midline 
blotches  become  more  numerous  and  some  species  develop  a 
series  along  the  horizontal  septum.  Early  larvae  of  Symphurus 
have  small  melanophores  along  the  ventral  midline,  and  in  some 
species,  also  along  the  dorsal  midline.  Most  species  have  a  single 
bar  posteriad  on  the  tail  and  at  least  one,  5.  athcauda,  has  large 
blotches  at  the  finfold  margins.  The  head  (particularly  ventrally), 
gut,  gas  bladder  and  horizontal  septum  become  pigmented  and 
later-stage  larvae  have  pigment  patterns  similar  to  Cynoglossus 
species. 

Metamorphic  stages 

Pleuronectiforms  undergo  a  remarkable  metamorphosis  dur- 
ing which  one  of  the  eyes,  the  left  in  dextral  and  the  right  in 
sinistral  species,  migrates  around  or  through  the  head  to  a  po- 
sition dorsal  to  the  non-migrating  eye.  Metamorphosis  occurs 
over  a  wide  size  range  among  flatfishes,  from  about  5  mm  in 
achirine  soles  (Houde  et  al..  1970)  to  greater  than  120  mm  in 
some  bothines  (Amaoka,  1971).  Capture  of  specimens  of  the 
enormous  flatfish  larva  observed  by  Barham  ( 1 966)  from  a  div- 
ing saucer  may  double  the  maximum  size  for  flatfish  larvae. 
Most  flatfishes  metamorphose  within  the  range  of  10-25  mm 
(see  preceding  sections  and  Tables  173-178);  the  size  interval 
over  which  the  process  occurs  is  smaller  in  species  which  meta- 
morphose at  a  small  size. 

Metamorphosing  specimens  are  relatively  rare  in  plankton 
collections  because  1)  the  process  is  transitory,  2)  avoidance  is 
increased  at  larger  sizes,  and  3)  metamorphosing  individuals 
may  change  habitat.  Existing  information  indicates  a  variety  of 
mechanisms  of  eye  migration  among  flatfishes.  In  groups  where 
the  dorsal  fin  origin  in  larvae  is  at  the  posterior  margin  of  the 
eye  or  more  rearward  (psettodids,  citharids,  scophthalmids,  most 
paralichthyids,  pleuronectids),  a  depression  forms  in  the  inter- 
ocular  region  and  the  eye  migrates  over  the  dorsal  midline  an- 
terior to  the  fin  origin.  Subsequently  the  dorsal  fin  extends  for- 
ward to  its  adult  position  (except  in  psettodids).  In  larvae  of 


bothids  and  the  paralichthyid  genera  Cyclopsetta.  Syaciiun  and 
Cithanchlhys  (some  species),  the  dorsal  fin  is  attached  to  the 
skull  anterior  to  the  eye  and,  during  metamorphosis,  the  eye 
migrates  through  a  slit  which  forms  between  the  fin  base  and 
the  skull.  In  some  metamorphosing  soleids  the  dorsal  fin  projects 
forward  above  the  snout  and  the  eye  migrates  through  the  space 
between  this  protuberance  and  the  skull;  subsequently  the  fin 
projection  fuses  to  the  skull  (Houde  et  al.,  1970;  Palomera  and 
Rubies,  1977;  Minami,  1981b).  Seshappa and  Bhimachar(  1955) 
described  the  process  of  eye  migration  in  a  captive  specimen  of 
Cynoglossus  macrostomus.  Just  before  eye  migration  a  fleshy 
hook-shaped  protuberance  grew  forward  from  the  region  of  the 
head  anterior  to  the  dorsal  fin  origin.  The  right  eye  migrated 
through  the  space  between  the  protuberance  and  the  skull,  after 
which  the  fleshy  appendage  fused  to  the  dorsal  region  of  the 
skull.  The  entire  process  took  place  over  a  5-hour  period  during 
the  night.  A  similar  structure  appears  on  advanced  larvae  of  an 
unidentified  cynoglossid  illustrated  by  John  (1951b)  and  this 
mechanism  of  eye  migration  may  be  widespread  among  cyno- 
glossids. 

During  eye  migration  in  flatfishes  a  number  of  other  meta- 
morphic events  occur:  1)  larval  spines  are  lost,  2)  elongate  rays 
assume  their  j  uvenile  proportions,  3)  gut  protrusions  are  brought 
into  the  body  cavity  and  internal  organs  are  rearranged,  4)  gas 
bladder,  if  present,  is  lost,  5)  pectoral  fins  develop  rays,  except 
in  cynoglossids,  some  soleids,  some  bothids  and  Mancopsetta, 
where  (one  or  both)  fins  are  lost  altogether  during  this  period, 
6)  larval  pigment  patterns  are  replaced  by  juvenile  patterns,  7) 
ossification  of  the  vertebral  column  and  other  bony  structures 
is  completed,  8)  intermuscular  bones  appear  in  bothids.  and  9) 
scales  form. 

(H.G.M.,  B.Y.S.)  National  Marine  Fisheries  Service, 
Southwest  Fisheries  Center,  P.O.  Box  271,  La  Jolla, 
California  92038;  (K.A.)  Faculty  of  Fisheries, 
Hokkaido  University,  Hakodate,  Japan;  (D.A.H.)  De- 
partment OF  Marine  Sciences,  University  of  Puerto 
Rico,  Mayaguez,  Puerto  Rico. 


Pleuronectiformes:  Relationships 
D.  A.  Hensley  and  E.  H.  Ahlstrom 


BASICS  of  the  current  working  model  for  evolution  of  pleu- 
ronectiforms were  proposed  by  Regan  (1910,  1929)  and 
Norman  (1934).  In  his  monograph,  Norman  treated  the  floun- 
ders (Psettodidae,  Bothidae,  Pleuronectidae),  and  though  he  did 
not  publish  a  revision  of  the  remaining  pleuronectiforms,  his 
key  and  classification  of  the  soleoids  were  published  posthu- 
mously (1966).  Norman's  model  and  classification  with  the 
modifications  of  Hubbs  (1945),  Amaoka  (1969),  Futch  (1977), 
and  Hensley  (1977)  represent  the  most  recent,  detailed  hypoth- 
esis for  pleuronectiform  evolution.  We  will  refer  to  this  as  the 
Regan-Norman  model  (Fig.  358)  and  classification  (preceding 


article,  this  volume)  and  consider  it  the  working  hypothesis  to 
be  reexamined  using  adult,  larval,  and  egg  characters. 

Formation  of  the  Regan-Norman  model  involved  an  eclectic 
approach,  i.e.,  a  combination  of  phyletic  and  phenetic  methods. 
Although  some  of  the  groups  currently  recognized  appear  to  be 
based  on  synapomorphies,  many  are  clearly  based  on  symple- 
siomorphies  and  were  recognized  as  such  by  the  authors.  This 
search  for  horizontal  relationships  among  pleuronectiforms  us- 
ing eclectic  methods,  with  one  exception,  has  been  the  only 
approach  used  in  this  group.  The  exception  is  the  recent  work 
of  Lauder  and  Liem  (1983)  in  which  a  cladogram  for  flatfishes 


HENSLEY  AND  AHLSTROM:  PLEURONECTIFORMES 


671 


Psettodidae 


Scophlhalmidae 


E 


Bothidae 


Citharidae 


V   V 


Soleidae       Cynoglossidae 


Pleuronectidae 


Pleuronectoidei 


Soleoidei 


Psettodoidei 


Fig.  358. 
(1969). 


Current  hypothesis  for  interrelationships  of  pleuronectiform  fishes.  Based  on  Norman  (1934,  1966),  Hubbs  (1945),  and  Amaoka 


is  presented.  However,  these  authors  present  this  as  a  tentative 
hypothesis  and  admit  that  the  interrelationships  expressed  are 
still  problematic.  Most  of  the  character  states  they  use  are  re- 
ductive, few  characters  were  analyzed,  and  the  authors  were 
understandably  unaware  of  recent  character  surveys,  since  much 
of  this  information  is  unpublished. 

We  have  made  the  assumption  that  the  order  Pleuronecti- 
formes  is  monophyletic  and  the  sister  group  is  the  remaining 
percomorph  fishes  (sensu  Rosen  and  Patterson,  1 969  and  Rosen, 
1973).  Although  the  monophyly  and  origin  of  the  group  is  still 
open  to  question  and  hypotheses  of  multiple  origins  have  been 
proposed  (e.g.,  Kyle,  1921;  Chabanaud,  1949;  Amaoka,  1969), 
a  monophyletic  model  with  a  percomorph  sister  group  still  ap- 
pears to  be  the  most  parsimonious.  In  other  words,  with  the 
information  available,  there  appears  to  be  no  need  to  hypoth- 
esize multiple  origins  for  flatfishes;  to  do  so  demands  the  inclu- 
sion of  a  great  deal  of  convergence. 

Relationships 

The  following  discussion  of  relationships  within  the  pleuro- 
nectiforms  is  cursory  and  preliminary.  In  fact,  it  asks  more 
questions  than  it  answers  and  illustrates  that  more  work  (par- 
ticularly osteological)  is  needed  in  certain  groups  before  the 


order  can  be  subjected  to  an  in-depth  cladistic  analysis.  Until 
this  work  is  completed,  it  is  premature  to  offer  a  new  hypothesis 
of  interrelationships  for  the  entire  order. 

Adult  characters 

Several  criteria  were  used  for  selecting  characters  for  discus- 
sion: (1)  amount  of  information  available  on  the  distribution 
of  character  states;  (2)  characters  commonly  used  in  the  past  to 
define  groups  of  pleuronectiforms;  (3)  those  for  which  our 
knowledge  of  distributions  of  states  is  limited,  but  appear  to 
indicate  groupings  different  from  those  hypothesized  in  the 
working  classification  and  which  need  additional  study;  and  (4) 
characters  which  are  well  known  in  certain  groups  and  are  po- 
tentially useful  for  elucidating  relationships  within  these  groups. 
Characters  and  character  complexes  used  in  this  study  are  dis- 
cussed below.  Characters  and  states  are  presented  in  Table  179. 

Optic  chiasma.—The  relationship  between  the  optic  chiasma 
and  ocular  asymmetry  of  pleuronectiforms  has  been  investigated 
by  several  workers  beginning  mainly  with  the  work  of  Parker 
(1903).  Hubbs  (1945)  examined  this  relationship  further  and 
presented  all  data  from  previous  studies.  Parker  found  that  most 
fishes  have  a  dimorphic  optic  chiasma,  i.e.,  the  nerve  of  the  left 


672 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


or  right  eye  is  dorsal  with  about  equal  frequency  (state  referred 
to  here  as  truly  dimorphic).  Exceptions  to  this  are  species  of 
paralichthyids  (sinistral)  and  pleuronectines  (dextral)  where  the 
right  or  left  optic  nerve,  respectively,  is  always  dorsal,  even  in 
reversed  individuals,  i.e.,  the  optic  chiasma  is  monomorphic. 
The  Soleidae  and  Cynoglossidae,  however,  retain  a  truly  di- 
morphic optic  chiasma.  Subsequent  work  by  Regan  (1910)  and 
Hubbs  (1945)  showed  that  in  the  indiscriminately  dextral  or 
sinistral  Psettodes  the  optic  chiasma  is  also  truly  dimorphic.  In 
addition,  Hubbs  presented  evidence  of  a  third  state,  at  least  in 
Citharoides  (sinistral),  where  the  nerve  of  the  migrating  eye  is 
dorsal  even  in  reversed  individuals.  He  thus  interpreted  the 
Citharidae  as  having  a  basically  dimorphic  optic  chiasma  and 
predicted  the  same  for  scophthalmids,  although  apparently  no 
one  has  examined  a  reversed  scopthalmid  to  test  this  prediction. 
A  truly  dimorphic  optic  chiasma  as  found  in  Psettodes  and  the 
soleoids  has  been  interpreted  as  plesiomorphic  for  pleuronec- 
tiforms.  The  type  of  optic  chiasma  found  in  Citharoides  and 
predicted  for  scophthalmids  (i.e.,  nerve  of  the  migrating  eye 
always  dorsal)  was  interpreted  as  an  intermediate  state  between 
the  truly  dimorphic  and  the  monomorphic  chiasmata  as  found 
in  pleuronectoids.  We  agree  with  this  interpretation  of  polarity. 
However,  some  plesiomorphic  states  have  been  used  to  define 
groups:  Psettodidae,  truly  dimorphic;  Citharidae,  basically  di- 
morphic; Scophthalmidae,  predicted  to  be  basically  dimorphic; 
and  Soleoidei,  truly  dimorphic. 

Major  problems  exist  with  the  use  of  the  optic  chiasma  for 
phylogenetic  inference.  One  of  these  concerns  the  feasibility  of 
actually  determining  which  state  exists  in  a  group.  Demonstrat- 
ing the  occurrence  of  truly  dimorphic  chiasmata  is  relatively 
simple.  All  that  is  needed  is  to  show  that  either  optic  nerve  is 
dorsal  regardless  of  which  eye  has  migrated;  reversed  individuals 
are  not  necessary.  To  demonstrate  occurrence  of  the  basically 
dimorphic  state,  reversals  are  needed  and  the  nerve  of  the  mi- 
grating eye  must  always  be  dorsal.  Likewise,  reversed  individ- 
uals must  be  examined  to  show  a  monomorphic  chiasma.  Here 
the  nerve  to  the  right  eye  must  be  dorsal  in  all  individuals 
(including  reversals)  of  normally  sinistral  species  and  the  nerve 
of  the  left  eye  must  be  dorsal  in  all  individuals  of  normally 
dextral  species.  When  one  actually  examines  the  data  for  this 
character  (see  Hubbs,  1945),  states  have  been  determined  for 
very  few  pleuronectiform  groups.  The  occurrence  of  the  basically 
dimorphic  state  in  the  Citharidae  was  demonstrated  in  only  one 
species.  Of  greater  significance,  however,  is  the  fact  that  a  mono- 
morphic state  has  been  shown  for  very  few  pleuronectoid  species. 
Within  the  pleuronectoids  it  has  been  widely  assumed  that  all 
paralichthyids,  bothids,  and  pleuronectids  have  monomorphic 
optic  chiasmata,  and  that  because  of  this  they  are  monophyletic 
and  not  closely  related  to  the  soleoids  (truly  dimorphic).  It  is 
worthy  of  note  here  that  a  monomorphic  optic  chiasma  has 
never  been  demonstrated  for  four  pleuronectid  subfamilies 
(Poecilopsettinae,  Rhombosoleinae,  Samarinae,  Paralichthod- 
inae),  the  Bothidae,  or  the  paralichthyid  genus  Thysanopsetta. 

Ocular  asymmetry— T\\K  character  (sinistral,  dextral,  indis- 
criminate) is  obviously  interrelated  with  the  optic  chiasma  in 
certain  groups,  i.e.,  those  with  basically  dimorphic  and  mono- 
morphic chiasmata.  The  evolution  of  ocular  asymmetry  and  its 
relationship  to  the  optic  chiasma  is  not  well  understood,  al- 
though there  is  one  major  hypothesis  (Norman,  1934;  Hubbs, 
1945)  which  states  that  primitively,  pleuronectiforms  were  in- 
discriminate in  ocular  asymmetry  and  the  optic  chiasma  was 


truly  dimorphic.  Soleoids  became  discriminate  (soleids  dextral 
and  cynoglossids  sinistral),  but  retained  a  truly  dimorphic  chias- 
ma. Psettodids  remained  indiscriminate  and  truly  dimorphic. 
Citharids  and  presumably  scophthalmids  became  discriminate 
(scophthalmids  and  citharines  sinistral  and  brachypleurines 
dextral)  but  retained  some  ontogenetic  plasticity  in  regard  to 
the  optic  chiasma,  since  reversed  individuals  still  have  the  nerve 
of  the  migrating  eye  dorsal  (basically  dimorphic).  The  remaining 
pleuronectoids  became  discriminate  (Paralichthyidae  and  Both- 
idae sinistral  and  Plueronectidae  dextral)  and  evolved  a  mono- 
morphic chiasma.  The  only  exceptions  with  regard  to  ocular 
asymmetry  are  certain  indiscriminate  paralichthyids  and  pleu- 
ronectines. However,  most  of  these  indiscriminate  pleuronec- 
toids have  been  shown  to  have  a  monomorphic  optic  chiasma 
(a  possible  exception  is  Tephnnectes).  It  would  thus  appear  that 
indiscriminate  ocular  asymmetry  in  pleuronectoids  developed 
secondarily  from  discriminate  ancestors  (Hubbs  and  Hubbs, 
1945). 

Making  phylogenetic  interpretations  from  two  states  of  ocular 
asymmetry  is  difficult  or  impossible  without  corroborative  evi- 
dence. Thus,  a  statement  to  the  effect  that  two  or  more  dextral 
(or  sinistral)  pleuronectoid  groups  are  most  closely  related  to 
each  other  because  they  are  dextral  (or  sinistral)  without  addi- 
tional evidence  of  synapomorphies  is  circular,  and  may  lead  to 
the  recognition  of  polyphyletic  groups.  This  reasoning  was  the 
basis  for  the  proposed  close  relationship  in  the  Regan-Norman 
model  between  the  Pleuronectinae  and  the  remaining  pleuro- 
nectid subfamilies  (Poecilopsettinae,  Rhombosoleinae,  Samar- 
inae, Paralichthodinae)  and  for  treating  the  genera  Mancopsetta 
and  Thysanopsetta  as  members  of  the  Bothidae  and  Paralich- 
thyidae, respectively. 

Ribs  and  intermuscular  bones.  — In  pleuronectiforms  that  pos- 
sess ribs,  these  appear  to  be  homologous  with  the  pleural  and 
epipleural  ribs  of  other  teleosts,  and  the  presence  of  these  bones 
should  be  considered  plesiomorphic  for  the  order.  Two  groups 
lack  both  series  of  ribs,  the  Achirinae  and  apparently  the  Cyn- 
oglossidae. Chabanaud  (1940)  reports  epipleural  ribs  in  some 
cynoglossids  but  mentions  no  genera  or  species.  We  have  not 
seen  them  in  cleared-and-stained  Symphurus  species  or  in  ra- 
diographs of  several  Cynoglossus  species.  Although  it  is  still 
commonly  believed  that  all  soleoids  lack  both  series  of  ribs  (e.g.. 
Nelson,  1976;  Lauder  and  Liem,  1983),  Chabanaud  (1940,  1941) 
found  short  epipleural  ribs  in  Solea,  Microchirus,  and  Aesopta, 
and  we  have  seen  them  in  Aseraggodes. 

Chabanaud  (1940,  1950,  1969)  found  additional  rib-like 
bones  ("metaxymyostes")  in  certain  pleuronectiforms.  Some  of 
his  statements  about  these  were  in  error,  and  it  is  now  clear  he 
was  referring  to  Bothus  podas  and  Samaris  cristalus  (Hensley, 
1977).  Amaoka  (1969)  found  these  ("intermuscular")  bones  in 
all  species  of  his  Bothidae  and  presented  very  detailed  descrip- 
tions of  their  morphology.  One  of  his  primary  justifications  for 
elevating  Norman's  ( 1934)  Bothinae  to  the  family  level  was  the 
presence  of  these  bones  in  the  group  and  their  absence  in  Nor- 
man's Paralichthyinae.  Norman  considered  Engyophrys.  Tri- 
chopsetta,  Monolene.  Taeniopsetta,  and  Perissias  to  be  paralich- 
thyines.  All  of  these  genera  have  intermuscular  bones  (Amaoka, 
1969;  Futch,  1977;  Hensley,  1977;  pers.  observ.)  and  are  con- 
sidered here  to  be  bothids. 

Bothid  intermuscular  bones  are  in  five  series.  Amaoka  ( 1 969) 
called  these  series  epimerals,  epicentrals,  hypomerals,  and 
myorhabdoi  (two  series).  He  interpreted  three  of  these  (epi- 


HENSLEY  AND  AHLSTROM:  PLEURONECTIFORMES 


673 


merals,  epicentrals,  hypomerals)  as  homologous  with  those  of 
lower  teleosts  (see  Phillips,  1942).  The  presence  of  these  bones 
was  the  main  reason  both  Chabanaud  (1949a)  and  Amaoka 
(1969)  hypothesized  that  pleuronectiforms  were  polyphyletic 
and  that  at  least  the  Bothidae,  and  in  the  case  of  Chabanaud 
also  the  Samarinae,  were  derived  from  some  lower  teleostean 
group.  Hensley  (1977)  presented  arguments  for  interpreting  the 
pleuronectiforms  as  monophyletic  and  the  presence  of  inter- 
muscular bones  in  at  least  the  Bothidae  as  being  apomorphic. 

Chabanaud  (1969)  described  intermuscular  bones  in  Samaris 
as  being  in  two  series.  However,  we  recently  e.\amined  a  cleared- 
and-stained  specimen  and  found  differences  with  Chabanaud's 
description.  In  the  abdominal  region,  rib-like  or  intermuscular 
bones  are  in  three  series.  Bones  of  the  middle  series  are  un- 
branched  and  in  the  horizontal  skeletogenous  septum.  Most 
bones  of  the  dorsal  and  ventral  series  are  branched.  In  the  region 
of  the  caudal  vertebrae,  there  are  only  the  dorsal  and  ventral 
series.  There  are  none  of  the  dorsal  and  ventral  myorhabdoi  as 
found  in  the  Bothidae.  Although  the  three  series  of  bones  found 
in  Samaris  resemble  the  epimerals,  epicentrals,  and  hypomerals 
of  bothids,  a  more  detailed  comparison  is  required  before  a 
statement  about  homologies  can  be  made. 

Amaoka  (1969)  interpreted  bothids  as  lacking  pleural  and 
epipleural  ribs,  but  possessing  the  five  series  of  intermuscular 
bones.  However,  there  is  another  interpretation.  It  is  possible 
that  Amaoka's  epicentrals  (limited  to  the  horizontal  skeletog- 
enous septum  of  the  abdominal  region)  and  abdominal  hypom- 
erals are  homologous  to  epipleural  and  pleural  ribs,  respectively, 
of  other  pleuronectiforms,  and  that  the  presence  of  myorhabdoi, 
epimerals,  and  caudal  hypomerals  are  apomorphic  states. 

Postcleithra.  — The  absence  of  postcleithra  was  a  character  state, 
apparently  apomorphic,  used  by  Norman  ( 1 934)  and  subsequent 
authors  to  distinguish  the  Soleoidei  from  the  Psettodoidei  and 
Pleuronectoidei.  However,  an  adequate  survey  of  this  character 
has  never  been  made  among  the  pleuronectoids.  In  a  preliminary 
survey,  we  found  postcleithra  absent  in  certain  pleuronectoids, 
i.e.,  the  Samarinae  and  the  bothid  genera  Mancopsetta  and  Pel- 
ecantchthys.  Postcleithra  are  definitely  present  in  the  rhombo- 
soleines  Oncopterus,  Azygopus.  Ammotretis.  and  Colistium.  but 
they  may  be  absent  in  Pelotretis.  Pellorhamphus.  and  Rhom- 
bosolea  (Norman,  1934:  fig.  25c;  Chabanaud,  1949).  Although 
lack  of  postcleithra  in  pleuronectiforms  is  reductive,  their  ab- 
sence in  certain  pleuronectoids  may  indicate  a  closer  relation- 
ship between  some  of  these  groups  and  soleoids  than  hypoth- 
esized in  the  Regan-Norman  model.  The  occurrence  of  this 
specialization  in  Pelecanichthys  is  almost  certainly  an  indepen- 
dent reduction,  since  this  genus  shows  several  synapomorphies 
with  other  bothids. 

Vomerine  teeth.  — Huhhs  (1945,  1946)  interpreted  the  presence 
of  vomerine  teeth  as  a  primitive  state  for  the  order,  and  we 
concur.  However,  Hubbs  presented  this  interpretation  as  evi- 
dence that  citharids  and  scophthalmids  were  closely  related  and 
represented  an  intermediate  grade  in  pleuronectoid  evolution. 
The  presence  of  vomerine  teeth  cannot  be  used  to  infer  phy- 
logenetic  relationships  among  pleuronectiforms. 

Fin  spmes— Huhhs  ( 1 945,  1 946)  presented  the  distributions  for 
dorsal,  anal,  and  ventral-fin  spines  in  pleuronectiforms.  Psei- 
todes  is  the  only  genus  with  dorsal  and  anal  spines.  This  genus 
and  the  Citharidae  are  the  only  flatfishes  with  ventral-fin  spines. 


Hubbs  properly  interpreted  their  presence  in  these  groups  as 
plesiomorphic  for  the  order.  However,  again,  he  used  a  hori- 
zontal or  eclectic  approach  and  inferred  a  close  relationship 
between  the  citharid  genera  and  interpreted  the  group  as  an 
intermediate  grade  in  pleuronectoid  evolution.  The  presence  of 
these  spines  does  not  indicate  phylogenetic  (vertical)  relation- 
ships. 

Supramaxillae. —  SxxpvdiVmxiWac  occur  in  Psettodes  and  the  cith- 
arids Eucitharus  and  Citharoides  (Hubbs,  1945).  In  Psettodes, 
the  bones  are  well  developed  and  apparently  present  on  both 
sides.  The  two  citharid  genera  have  them  reduced  in  size,  con- 
fined to  the  blind  side,  or  sometimes  missing.  The  presence  of 
these  bones  is  plesiomorphic  for  the  order  and  should  not  be 
used  to  infer  phylogenetic  relationships. 

Ventral-fin  placements  and  base  lengths.  —  Evolution  of  ventral- 
fin  asymmetry  in  pleuronectiforms  is  not  well  understood.  Most 
of  our  knowledge  concerning  the  relationship  between  ocular 
and  ventral-fin  asymmetry  has  come  from  some  rare  examples 
of  reversals  in  forms  with  asymmetrical  ventral-fin  morphology 
(see  Norman,  1 934).  For  comparative  purposes,  i.e.,  attempting 
to  determine  homologous  states,  it  would  appear  to  be  more 
correct  to  compare  ocular  and  blind-side  ventral  fins  between 
groups  rather  than  those  of  the  right  and  left  sides  (see  Hubbs 
and  Hubbs,  1945).  At  present,  there  are  several  problems  in 
using  ventral-fin  morphology  to  elucidate  phylogenetic  rela- 
tionships. Most  work  here  has  dealt  only  with  external  mor- 
phology and  much  of  this  has  not  been  sufficiently  detailed  or 
accurate.  What  is  needed  are  thorough  comparisons  of  basip- 
terygia  as  well  as  fins.  Due  to  the  paucity  of  accurate  and  detailed 
studies  of  these  structures  in  flatfishes,  it  is  not  possible  to  ad- 
equately define  character  states  for  an  in-depth  comparison 
throughout  the  order.  Thus,  ventral-fin  characters  were  not  in- 
cluded in  Table  1 79.  What  follows  is  a  discussion  of  general 
patterns  of  ventral-fin  morphology. 

Ventral  fins  with  short  bases  and  symmetrical  placements 
have  been  correctly  considered  plesiomorphic  states  in  pleu- 
ronectiforms, and  any  type  of  asymmetry  in  placement,  size, 
shape,  o;^  meristics  as  having  been  derived  from  symmetrical 
states  (e.gir  Norman,  1934;  Hubbs,  1945;  Amaoka,  1969).  Most 
ventral-fin  characters  used  have  involved  positions  of  the  fins 
relative  to  the  midventral  line  and  relative  lengths  of  the  fin 
bases.  Unfortunately,  symmetry  (plesiomorphic  states)  in  both 
of  these  characters  has  been  used  to  define  groups.  Short-based 
fins  and  symmetry  or  near  symmetry  in  placement  and  base 
lengths  occur  in  Psettodes.  the  Paralichthyidae  (except  the  Cy- 
clopsetta  group),  the  Citharidae,  most  soleines,  most  or  all  Pleu- 
ronectinae,  and  the  Poecilopsettinae.  States  where  the  ocular 
ventral  fin  is  on  the  midventral  line  and  has  a  base  extending 
farther  anteriorly  than  that  of  the  blind  side  form  a  continuum. 
Thus,  groups  with  the  base  of  the  ocular  ventral  fin  only  slightly 
extended  anterior  to  that  of  the  blind  side  (origin  of  blind  fin 
at  transverse  level  of  about  the  second  or  third  ray  of  the  ocular 
fin)  are  the  Samarinae,  possibly  some  Soleinae,  Paralichthodes, 
the  Taeniopsettinae,  and  Monolenc,  groups  where  the  origin  of 
the  ocular  fin  is  farther  anterior  relative  to  that  of  the  blind  fin 
are  the  Rhombosoleinae,  all  Bothinae  (except  Monolene).  and 
possibly  some  Soleinae.  Two  groups,  the  Scophthalmidae  and 
Achirinae,  have  both  ventral-fin  bases  close  to  or  virtually  on 
the  midventral  line  and  the  anterior  basipterygial  processes  ex- 
tended. The  Cyciopsetta  group  has  the  ocular  fin  on  the  mid- 


674 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Table  1 79.    Characters  and  States  for  Pleuronecttform  Groups.  Where  appropriate  states  are  indicated  by  underlined  letters.  See  text  and 

Figs.  359-364  for  hypural  fusion  patterns. 


Optic 

Ocular 

chjasma 

asym- 

Post- 

(Truly 

metry 

cleithra 

Dimorphic, 

(Dc.xtral. 

Ribs 

Intermuscular  Bones 

(Present. 

Basically 

Sinis- 

(Present, Absent) 

(Present 

Absent) 

Absent 

Vomerine 

Ventral- 

tin 

Supra- 

Dimorphic, 

tral. 

Present 

teeth 

forniul-i 

Mono- 

Indis- 

Epi- 

Epi- 

Hypo- 

Epi- 

Myo- 

only  in 

(Present. 

(ocular  SI 

dc/ 

(Present, 

Taxon 

morphic) 

cnminate) 

pleural 

Pleural 

centrals 

merals 

merals 

rhabdoi 

Larvae) 

Absent) 

blind  side) 

Absent) 

Psettodidae 

TD 

I 

p 

P 

A 

A 

A 

A 

P 

P 

1,  5/1, 

5 

P 

Citharidae 

Brachypleurinae 

Brachypleura 

9 

D 

p 

P 

A 

A 

A 

A 

P 

P 

I,  5/1, 

5 

A 

Lcpidohlepharon 

9 

D 

p 

P 

A 

A 

A 

A 

P 

P 

I,  5/1. 

5 

9 

Citharinae 

Citharoides 

BD 

S 

p 

P 

A 

A 

A 

A 

P 

A 

1,  5/1, 

5 

P 

Eucitharus 

9 

S 

p 

P 

A 

A 

A 

A 

P 

P 

I,  5/1, 

5 

P 

Scophthalmidae 

9 

S 

p 

P 

A 

A 

A 

A 

P 

P.  A 

6/6 

A 

Paralichthyidae 

*Cyclopsel!a  group 

9 

S 

p 

P 

A 

A 

A 

A 

P 

A 

5-6/6 

A 

"Pseudorhomhus  group 

9 

S 

p 

P 

A 

A 

A 

A 

P 

A 

6/6 

A 

***Paratichthys  group 

M 

S.I 

p 

P 

A 

A 

A 

A 

P 

A 

6/6 

A 

Tephrinectes 

M? 

I 

p 

P 

A 

A 

A 

A 

P 

A 

6/6 

A 

Thysanopsetta 

9 

S 

p 

P 

A 

A 

A 

A 

P 

A 

6/6 

A 

Bothidae 

Taeniopsettinae 

9 

S 

p? 

P? 

P 

P 

P 

P 

P 

A 

6/6 

A 

Bothinae 

9 

S 

p? 

P? 

P 

P 

P 

P 

P.  A 

A 

6/6 

A 

****MancopseUa 

9 

s 

p 

P 

A 

A 

A 

A 

A 

A 

7/5- 

-7 

A 

Pleuronectidae 

Pleuronectinae 

M 

D,  I 

p 

P 

A 

A 

A 

A 

P 

A 

4-7/4- 

-7 

A 

Poecilopsettinae 

? 

D 

p 

P 

A 

A 

A 

A 

P 

A 

6/6 

A 

Paralichthodinae 

9 

D 

9 

9 

9 

9 

9 

9 

9 

A 

6/6 

A 

Samarinae 

9 

D 

P? 

P? 

P 

P 

P 

A 

A 

A 

5/5 

A 

Rhombosoleinae 

9 

D 

P 

P 

A 

A 

A 

A 

P,  A? 

A 

6-13/0- 

-6 

A 

Soleidae 

Soleinae 

TD 

D 

P 

A 

A 

A 

A 

A 

A 

A 

5/5 

A 

Achirinac 

TD 

D 

A 

A 

A 

A 

A 

A 

PL 

A 

3-5/2- 

-4 

A 

Cynoglossidae 

TD 

S 

A 

A 

A 

A 

A 

A 

A 

A 

0-2/4 

A 

*  Cilhahchrhys.  Cychpsella.  Etropus.  Syacmm 
•*  Cephalopseira.  Pseitdorhombw,.  Tarphops. 
***  Ancylopsetta,  Gastropsetta.  Htppvglossma.  Lioglossina.  Parahchlhys.  X'erecundttm.  Xystreuiy^ 
****  Achiropsetia  and  Neoachiropsella  are  considered  synonyms. 


ventral  line,  but  the  basipterygium  of  the  blind  fin  is  placed  in 
a  more  anterior  position  than  that  of  the  ocular  side.  Another 
unique  state  is  the  loss  of  the  blind  ventral  fin  in  some  genera 
of  the  Rhombosoleinae,  although  the  basipterygium  of  the  blind 
side  is  probably  still  present.  The  Cynoglossidae  are  the  only 
pleuronectiforms  in  which  the  blind  ventral-fin  base  is  oriented 
along  the  midventral  line  and  the  ocular  fin  is  in  a  more  dorsal 
position  or  absent.  In  cynoglossids  missing  the  ocular  fin,  at 
least  the  dorsal  process  of  the  left  basipterygium  is  still  present. 


Vertebral  transverse  apophyses.  — Regan  (1910)  used  the  pres- 
ence of  transverse  apophyses  on  caudal  vertebrae  as  a  state  to 
distinguish  his  bothid  subfamilies  Platophrinae  and  Bothinae 
from  the  Paralichthyinae  (=Paralichthyidae  with  modifica- 
tions). Norman  ( 1 934)  combined  the  Platophrinae  and  Bothinae 
into  his  Bothinae  and  Scophthalminae  and  again  used  transverse 
apophyses  on  caudal  vertebrae  to  distinguish  the  bothines  and 
scophthalmines  from  the  paralichthyines.  Amaoka  (1969)  used 
the  presence  of  these  structures  to  define  his  Bothidae  and  dis- 
tinguish them  from  the  other  sinistral  flounders  he  treated  (Para- 
lichthyidae, Citharidae,  Psettodidae).  Hensley  (1977)  and  Futch 


(1977)  found  transverse  apophyses  in  Engyophrys,  Trichopsetta, 
and  Monolene  and  suggested  this  as  a  character  state  indicating 
these  genera  were  more  closely  related  to  the  Bothidae  than  the 
Paralichthyidae.  We  have  since  found  them  in  Perissias.  As 
previously  slated,  Norman  (1934)  had  placed  these  four  genera 
in  the  Paralichthyinae.  Amaoka  (1969)  presented  the  most  de- 
tailed descriptions  of  these  structures.  Basically,  there  are  two 
pairs  of  transverse  apophyses  on  the  vertebrae,  an  anterior  and 
a  posterior  pair.  They  are  found  on  many  abdominal  and  most 
caudal  vertebrae.  Subsequently,  we  have  found  that  the  trans- 
verse apophyses  seen  by  Regan  (1910)  and  Kyle  (1921:  fig.  32) 
in  the  scophthalmids  are  very  similar  to  those  present  in  the 
Bothidae.  They  are  similar  in  shape  and  occur  in  two  pairs. 

Amaoka  (1969)  interpreted  the  occurrence  of  these  structures 
in  the  Bothidae  as  indicative  of  a  relationship  to  some  fish  group 
other  than  the  Percomorpha  and  used  this  as  evidence  that  the 
Bothidae  arose  independently  from  the  remaining  pleuronec- 
tiforms. To  support  this,  he  cites  the  occurrence  of  similar  struc- 
tures in  anguilliforms  (Trewavas,  1932;  Asano,  1962). 

Recently  in  a  preliminary  survey  of  this  character  or  complex 
in  other  flatfishes,  we  found  transverse  apophyses  on  some  ver- 
tebrae in  the  Samarinae,  Cynoglossidae,  and  Soleinae.  However, 


HENSLEY  AND  AHLSTROM:  PLEURONECTIFORMES 

Table  179.    Extended. 


675 


Position 

Infra- 

of unnary 

Position 

Articu- 

orbital 

Firsl 

papilla 

of  vent 

Haemal 

lation  of 

lateral- 

neural 

(Midven- 

(Midven- 

arch  on 

parhypural 

Haemal 

line 

Vertebral 
transverse  apophyses 

spine 

tral. 

~tral. 

parhypural 

with  termi- 

spine 

canal  on 

(Presenl. 
Absenl. 

Ocular 
side. 

Ocular 
side. 

(Presenl, 
Absent. 

nal  half 
centrum 

Hypural 
fusion 

Number 

on  PU2 

(Autog- 

Branched 

ocular 
side 

(Present. 

Absent) 

.   Absent  or 

Blind 

Blind 

Rudimen- 

(Present. 

pattern 

of  autogenous 

enous, 

caudal-lin 

(Present. 

Antenor 

Poslcnor 

Reduced) 

side) 

side) 

tary) 

Absent) 

(1-6) 

epurals 

Fused) 

Total  caudal-fin  rays 

rays 

Absent) 

A 

A 

P 

M 

M 

P 

P 

1 

1 

A 

24-25 

15 

P 

A 

A 

P 

O 

O 

R? 

A 

6 

1 

F 

21 

13 

A 

A 

A 

P 

O 

O 

P? 

P 

1 

2.3? 

A 

23 

15 

P 

A 

A 

P 

O 

O 

R? 

A 

2 

2,3? 

A 

22- 

-23  (usually  23) 

15 

P 

A 

A 

P 

O 

O 

A 

A 

4 

I 

F 

21 

14-15 

P 

P 

P 

P 

o 

M,  B 

A 

A 

6 

1,2? 

F 

16-17 

13-15 

P 

P 

P 

A 

B 

B 

A 

A 

6 

0 

F 

17 

11 

A 

A 

A 

P 

O 

B 

R? 

A 

6 

0 

F 

17- 

-18  (usually  17) 

10-13 

P 

A 

A 

P 

O 

M.  B 

A 

A 

6 

0.  1,2? 

F 

18 

13 

P 

7 

7 

P 

O 

M 

R'' 

A 

1 

2 

F 

20 

14 

P? 

7 

7 

P 

O 

B 

R? 

A 

1 

1 

F 

15-16 

0 

A? 

P 

P 

A 

O 

B 

A 

A 

6 

0 

F 

16- 

-18  (usually  17) 

10-13 

A 

P 

P 

A 

O 

B 

A 

A 

6 

0 

F 

15- 

-18  (usually  17) 

9-13 

A 

A 

A 

AR 

O 

M 

R? 

A 

1 

1 

F 

14- 

-19  (usually  16-18) 

13-14 

A 

A 

A 

P 

O 

M,  B 

A 

A 

6 

1.2? 

F 

17-24 

10-16 

P 

A? 

A? 

AR 

o 

M 

R 

A 

1 

1 

F 

20 

14-15 

7 

7 

7 

P 

9 

B 

■7 

9 

1 

0 

9 

16 

12 

7 

P 

A 

A 

o 

M 

A 

A 

5 

1? 

F 

16 

0-12 

A 

A 

A 

AR 

o 

M.  B 

A 

A 

1.4 

1 

F 

17-20 

8-15 

A 

P 

A 

A 

o 

M 

A 

A 

4 

1 

F 

16- 

-20  (usually  18) 

0-18 

7 

A 

A 

AR 

o 

B 

A 

A 

1.2?,  3 

1,2? 

A 

15- 

-18  (usually  16) 

11-17 

P 

P 

A 

AR 

M 

B 

A 

A 

4 

0.  1 

F 

8-14 

0 

A 

in  these  groups,  they  occur  only  as  one  pair  on  the  anterior  end 
of  the  vertebra.  In  addition,  the  Cyclopsetta  group  has  two  pairs 
of  very  small  lateral  protuberances  on  most  vertebrae.  How  to 
interpret  the  presence  of  vertebral  transverse  apophyses  in  pleu- 
ronectiforms  is  still  open  to  question. 

First  neural  spine.— Amaoka  {1969)  found  that  the  neural  spine 
of  the  first  vertebra  is  missing  in  the  Bothidae  and  interpreted 
this  as  a  synapomorphy  for  the  group,  since  absence  of  this  spine 
is  apparently  rare  or  unknown  in  other  teleosts.  We  have  made 
a  preliminary  survey  for  this  in  other  pieuronectiforms  not  treat- 
ed by  Amaoka.  Some  of  this  survey  was  based  on  radiographs, 
and  due  to  the  close  proximity  of  the  first  vertebra  and  neu- 
rocranium,  in  some  groups  we  are  not  sure  if  the  first  neural 
spine  is  present,  absent,  or  greatly  reduced.  The  states  in  other 
groups  are  more  certain,  since  some  cleared-and-stained  ma- 
terial was  available.  A  greatly  reduced  or  missing  first  neural 
spine  is  not  limited  to  the  Bothidae  (Table  1 79). 

Position  of  the  urinary  papilla.  — AW  flatfishes  have  a  papilla  on 
the  posteroventral  area  of  the  abdomen  near  the  anal-fin  origin. 
Schmidt  (1915,  cited  by  Norman,  1 934)  commented  on  its  po- 
sition in  flatfishes,  claiming  it  was  located  on  the  ocular  side  in 
all  species.  However,  Chabanaud  (1934),  Hubbs  (1945),  and 
Hubbs  and  Hubbs  (1945)  found  it  to  be  on  the  midventral  line 


in  Psettodes.  In  addition,  Hubbs  (1945)  and  Hubbs  and  Hubbs 
(1945)  found  the  papilla  on  the  blind  side  in  the  paralichthyid 
genera  Syacium,  Citharichthys.  and  Etropus.  We  have  found  it 
in  the  same  position  in  Cyclopsetta.  Another  exception  here 
may  be  certain  cynoglossids.  Menon  (1977:  fig.  45)  shows  the 
urinary  papilla  on  the  blind  side  in  a  species  of  Cynoglossus, 
but  claims  it  is  attached  to  the  first  anal-fin  ray  in  all  species  of 
the  family.  A  midventral  position  for  the  papilla  is  generalized 
for  teleosts  and  plesiomorphic  for  pieuronectiforms. 

Position  o/"  ve«?.  — Position  of  the  anus  in  flatfishes  has  been 
reviewed  by  Norman  (1934),  Hubbs  (1945),  and  Hubbs  and 
Hubbs  (1945).  A  midventral  position  is  plesiomorphic  for  the 
order.  In  flatfishes  where  the  vent  is  on  or  near  the  midventral 
line,  it  is  often  very  difficult  to  determine  what  state  is  repre- 
sented. It  is  on  the  blind  side  in  several  groups,  but  apparently 
on  the  ocular  side  only  in  the  Citharidae.  Hubbs  (1945)  inter- 
preted the  distribution  of  these  states  as  indicating  that  deflec- 
tion of  the  vent  to  the  blind  side  has  occurred  several  times 
within  the  order. 

Caudal-fin  complex.— The  caudal  fin  and  skeleton  of  many 
species  of  pieuronectiforms  have  been  illustrated  and  discussed 
(e.g.,  Monod,  1968;  Amaoka,  1969).  The  caudal  skeleton  of 
Psettodes  is  reported  to  be  the  most  primitive  among  living 


676 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES- AHLSTROM  SYMPOSIUM 


NS(EP?)  UN 


PHYP 


Fig.  359.  Caudal  skeleton  oi  Psetlodes  bennctti.  Hypural  pattern  1. 
EP  =  epural,  HY  1-5  =  hypurals  1-5,  NS  =  neural  spine,  PHY  =  par- 
hypural.  PHYP  =  parhypurapophysis,  PL)  2,  3  =  prcural  centrum  2,  3, 
THC  =  terminal  half  centrum.  UN  =  uroneural.  Redrawn  from  Monod 
(1968). 


flatfishes.  It  can  be  characteinzed  as  follows  (Fig.  359):  a  par- 
hypural  with  a  haemal  arch  and  parhypurapophysis;  five  au- 
togenous hypurals;  two  pairs  of  uroneurals,  i.e.,  pairs  of  stegurals 
and  splinter  bones;  two  epurals,  the  first  between  the  neural- 
arch  remnants  of  the  second  preural  centrum;  terminal  half 
centrum,  i.e.,  fusion  of  two  ural  centra  and  the  first  preural 
centrum;  haemal  spine  of  the  second  preural  centrum  autoge- 
nous; haemal  spine  of  the  third  preural  centrum  fused;  and  24- 
25  caudal  rays,  17  principal,  15  branched.  The  caudal  skeleton 
of  Psetlodes  has  been  labelled  as  basically  percoid  (e.g.,  Wu, 
1932;  Monod,  1968;  Amaoka,  1969).  It  should  be  noted  here 
that  the  neural  spine  of  the  second  preural  centrum  is  interpreted 
as  probably  a  captured  epural,  and  that  apparently  only  one  free 
epural  remains.  This  is  one  of  the  more  important  differences 
between  Psetlodes  and  all  other  pleuronectiforms,  which  have 
a  neural  spine  on  the  second  preural  centrum  and  apparently  a 
basal  number  of  two  epurals.  There  are  at  least  two  hypotheses 
which  may  explain  this  difference:  (1)  The  earliest  pleuronec- 
tiforms may  have  had  three  free  epurals,  the  anteriormost  be- 
coming wedged  in  the  neural-arch  remnant  on  the  second  preur- 
al centrum  (i.e.,  captured)  and,  thus  forming  a  secondary  neural 
spine.  In  Psetlodes  the  remainmg  epurals  were  fused  (Amaoka, 
1 969)  or  one  was  lost,  while  both  were  retained  in  the  remaining 
flatfishes,  at  least  primitively.  (2)  The  earliest  pleuronectiforms 
had  two  epurals,  the  anteriormost  being  captured  in  Psetlodes, 
leaving  one  free  epural.  In  the  remaining  flatfishes  a  neural  spine 
on  the  second  preural  centrum  was  acquired  by  fusion  of  this 
vertebra  with  an  anterior  one  bearing  a  spine.  Rosen  (1973)  has 
discussed  the  second  hypothesis  to  account  for  secondaiy  ac- 
quisition of  a  neural  spine  on  the  second  preural  centrum  and 
offered  as  evidence  the  frequent  occurrence  of  double  spines  on 
the  second  preural  centrum.  Such  anomalies  are  frequent  in 


pleuronectiforms  (see  Cole  and  Johnstone.  1902;  Barrington, 
1937;Chabanaud,  1937;Amaoka,  1969;  Okiyama,  1974;Futch, 
1977;  Fig.  360H).  However,  although  a  detailed  survey  for  these 
doubled  spines  has  never  been  done,  it  appears  that  doubled 
neural  spines  on  this  vertebra  are  just  as  frequent  as  doubled 
haemal  spines. 

In  spite  of  the  work  that  has  been  done  on  pleuronectiform 
caudal  osteology,  there  is  still  little  agreement  on  interpretation 
of  some  structures.  We  cannot  solve  these  problems  here  or 
discuss  them  in  great  detail.  Most  of  these  differences  in  inter- 
pretation concern  certain  epaxial  elements.  More  detailed  com- 
parative work  needs  to  be  done  on  these  elements  before  ho- 
mologies can  be  determined.  For  example,  there  is  one 
interpretation  that  uroneurals  occur  only  in  Psetlodes  (Ahl- 
strom).  However,  what  appear  to  be  remnants  of  a  stegural  may 
remain  in  Cilharoides,  Lepidoblepharon,  Scophlhalmus.  and 
some  achirines  (Fig.  361;  Amaoka,  1969;  Hensley,  pers.  ob- 
serv.).  Although  sufficient  comparative  work  has  not  been  done 
to  treat  these  dorsal  structures  across  all  lines  of  ffatfishes,  within 
certain  groups  we  can  be  fairly  sure  of  homologies,  due  to  certain 
consistent  patterns  of  placement  and  shape  and  to  some  larval 
work  where  fusions  have  been  observed. 

In  regard  to  neural  and  haemal  spines  of  the  second  preural 
centrum,  the  parhypural,  and  hypurals,  our  knowledge  rests  on 
firmer  ground.  Characteristics  of  these  structures  have  been 
widely  surveyed  and  there  is  much  more  agreement  on  inter- 
pretation of  homologous  states.  We  interpret  autogenous  neural 
and  haemal  spines  on  the  second  preural  centrum,  retention  of 
a  parhypurapophysis  and  haemal  arch  on  the  parhypural.  and 
articulation  of  the  parhypural  with  the  terminal  half  centrum 
as  plesiomorphic  for  the  order. 

Several  patterns  of  fusions  occur  in  regard  to  hypurals  1-4. 
Hypural  5  moves  to  an  epaxial  position  during  ontogeny  in 
flatfishes  (Figs.  360,  362),  and  its  fate  is  more  properly  discussed 
in  reference  to  fusion  (or  lack  of  it)  with  epurals.  The  most 
primitive  condition  is  where  hypurals  1-4  are  not  fused  to  the 
terminal  half  centrum  or  among  themselves  (pattern  1;  Figs. 
359,  363  upper). 

There  are  three  patterns  which  are  slightly  different  from  each 
other.  The  interpretation  of  these  is  not  so  obvious,  and  we  are 
hesitant  here  to  make  statements  concerning  homologies  be- 
tween groups.  One  of  these  (pattern  2)  is  where  hypurals  3  and 
4  are  fused  to  the  terminal  half  centrum  (Fig.  36 1 ).  This  pattern 
is  shown  by  Citharoides  and  apparently  some  Achirinae.  In 
some  achirines,  a  somewhat  different  pattern  (3)  occurs  where 
hypurals  2,  3,  and  4  are  fused  to  the  terminal  half  centrum  (Fig. 
363  middle).  A  fusion  of  hypurals  1-4  to  the  terminal  half 
centrum  (pattern  4)  is  found  in  the  Soleinae,  Cynoglossidae,  one 
cithand  (Eucilharus),  and  two  genera  of  Rhombosoleinae  (Pel- 
torhamphus.  Rhoinbosolea\  Figs.  362,  363  lower).  Caudal-fin 
development  in  a  soleine  is  illustrated  in  Fig.  362. 

Another  pattern  of  hypurals  (5)  is  unique  to  the  Samarinae 
(Fig.  364).  There  are  two  ways  to  interpret  this  pattern.  Here 
the  central  hypurals  (2  and  3  or  2-4)  are  fused  to  the  terminal 
half  centrum.   However,   unlike  the  patterns  previously  de- 


Fig.  360.  Caudal-fin  structure  of  Engyophrys  senla  larvae  (A-F).  juveniles  and  adults  (G-H).  Standard  lengths  of  specimens:  (A)  4.6  mm;  (B) 
5.5  mm;  (C)  7.0  mm;  (D)  7.6  mm;  (E)  7.7  mm;  (F)  15.3  mm;  (G)  45.7  mm;  (H)  82.4  mm.  NC  =  notochord.  other  abbreviations  as  in  Fig.  359. 
Redrawn  from  Hensley  (1977). 


HENSLEY  AND  AHLSTROM:  PLEURONECTIFORMES 


677 


EP*HY5 


HY3*4 


0  5  mm 


HY3  +  4 


HY1*2 


S HY3'4 


= HY1*2 


PHY 


0.5mm 


EP       HY5 


05  mm 


678 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


THC  •HY3t4 


Fig.  36 1 .  Caudal  skeleton  of  Citharoides  macrolepis.  Hypural  pat- 
tern 2.  HAR  =  haemal-arch  remnant,  other  abbreviations  as  in  Fig. 
359.  "V"  on  distal  end  of  fin  ray  indicates  dorsal-  and  ventralmost 
branched  ray. 

scribed,  in  the  samarines  hypural  1  does  not  articulate  with  the 
terminal  half  centrum. 

The  last  pattern  of  hypurals  (6)  is  characterized  as  follows 
(Figs.  360,  364  middle  and  lower);  hypurals  1  and  2  are  fused 
together  forming  one  element  which  articulates  with  the  pos- 
teroventral  surface  of  the  terminal  half  centrum;  and  hypurals 
3  and  4  are  fused  together  and  to  the  terminal  half  centrum. 
This  pattern  occurs  in  the  Pleuronectinae,  Paralichthyidae  (ex- 
cept Tephrinectes  and  Thysanopsetta),  Scophthalmidae,  one 
citharid  (Brachypleura),  and  the  Bothidae  (except  Mancopsella). 
We  interpret  this  pattern  as  homologous  between  these  groups, 
derived,  and  indicative  of  a  monophyletic  origin.  We  will  refer 
to  these  fishes  as  the  bothoid  group.  Caudal-fin  development  in 
a  bothid  is  illustrated  in  Fig.  360. 

Although  there  is  still  some  doubt  concerning  interpretations 
of  certain  epaxial  caudal  elements  in  flatfishes,  some  patterns 
are  apparent.  Most  of  the  information  indicates  that  at  least  in 
most  pleuronectiform  groups,  the  basal  epural  number  is  two. 
However,  there  is  a  small  third  element  that  appears  in  many 
species  (Fig.  36 1 ;  first  uroneural  of  Amaoka,  1 969).  This  element 
does  not  appear  to  be  paired  and  its  interpretation  and  fate  in 
some  groups  is  questionable.  The  two  larger  epural  elements  are 
still  present  in  some  flatfishes  (Figs.  361,  363  upper),  the  cith- 
arids  Lepidoblepharon  and  Citharoides  and  the  paralichthyid 
Tephrinectes.  The  fate  of  these  from  the  perspective  of  the  entire 
order  is  questionable.  However,  it  is  obvious  that  these  epurals 
have  been  reduced  to  one  or  zero  in  several  groups.  Which  of 
these  reductions  are  homologous  is  unknown.  Within  groups 
defined  by  other  specializations,  however,  we  are  probably  jus- 
tified in  assuming  these  epural  reductions  took  the  same  course 
and  are  homologous  states. 

Although  space  does  not  allow  a  more  detailed  discussion  of 
other  caudal-fin  characters,  some  obvious  trends  should  be  men- 


tioned; Symmetrization— There  is  a  marked  trend  among  flat- 
fishes toward  dorsoventral  symmetry  in  the  caudal  fin  and  skel- 
eton. This  has  occurred  by  various  types  of  fusions,  losses,  and 
secondary  divisions  of  elements.  These  secondary  divisions  oc- 
cur as  scissures  of  varying  depths  in  many  caudal  elements  (Figs. 
360H,  362F,  363  lower,  364  upper).  Reduction  of  total  and 
branched  caudal  rays— It  has  long  been  recognized  that  more 
primitive  flatfishes  tend  to  have  larger  numbers  of  total  and 
branched  caudal  rays.  Thus,  Psettodes  has  a  total  caudal  ray 
count  of  24-25,  15  of  which  are  branched.  In  many  groups, 
caudal  rays  have  been  reduced  to  less  than  18  and  branched 
rays  to  0-13. 

Infraorbital  lateral-line  canal  on  ocular  side.  — In  his  study  of 
sinistral  flounders  (i.e.,  Psettodidae  and  Pleuronectoidei)  of  Ja- 
pan, Amaoka  (1969)  found  ocular  infraorbital  bones  present  in 
the  Psettodidae,  two  citharid  genera  (Citharoides.  Lepidobleph- 
aron), and  the  Paralichthyidae;  they  were  absent  from  Japanese 
bothids.  We  have  since  done  some  survey  work  on  this  character 
in  other  groups  not  treated  by  Amaoka  and  found  ocular  in- 
fraorbital bones  missing  in  additional  groups  (Table  179). 

Examination  of  the  Regan-Norman  model 
using  adult  characters 

In  the  following  discussion,  the  groups  and  classification  re- 
sulting from  the  current  model  for  pleuronectiform  evolution 
will  be  reexamined.  The  limited  analysis  presented  here  sheds 
much  doubt  on  the  monophyly  of  many  of  the  currently  rec- 
ognized groups  and  their  interrelationships.  In  a  few  cases,  the 
evidence  favoring  different  interpretations  is  so  strong  that  these 
should  be  recognized  in  classifications.  However,  most  of  this 
analysis  has  produced  questions  and  alternative  suggestions  that 
need  additional  study. 

Psettodoidei,  Psettodidae. —Nearly  all  of  the  character  states  used 
to  define  this  group  (Psettodes.  two  species)  are  symplesio- 
morphies  or  have  been  interpreted  as  such.  Two  exceptions,  gill 
arches  with  groups  of  teeth  and  barbed  jaw  teeth,  are  states  that 
Hubbs  (1945)  proposed  as  synapomorphies.  Although  we  have 
no  reason  to  doubt  that  Psettodes  is  a  natural  group,  it  should 
be  redefined  using  character  states  which  have  been  shown  to 
be  synapomorphies. 

Soleoidei.— The  diflferences  between  the  Soleoidei  and  Pleuro- 
nectoidei were  noted  and  expressed  in  important  classifications 
before  the  works  of  Regan  and  Norman  (e.g.,  Jordan  and  Ev- 
ermann,  1896-1900)  and  they  are  obviously  evident  in  the  cur- 
rent model  and  classification.  In  most  previous  systematic  re- 
search on  pleuronectiforms,  the  author  has  concerned  himself 
with  one  or  the  other  group  and  assumed  that  the  two  were 
related  only  through  a  common  ancestor  near  the  early  pleu- 
ronectiform line.  The  possibility,  for  example,  that  some  so- 
leoids  may  be  most  closely  related  to  some  pleuronectoids  has 
only  rarely  been  addressed.  In  any  cladistic  analysis  of  pleuro- 
nectiform interrelationships,  character  states  used  to  unite  the 
soleoids  will  need  to  be  reinterpreted.  Some  character  states 


Fig.  362.  Caudal-fin  structure  of  Solea  solea  larvae  (A-C),  juveniles  and  adults  (D-F).  Total  lengths  of  specimens:  (A)  6.0  mm;  (B)  6.8  mm; 
(C)  8.1  mm;  (D)  1 1.5  mm;  (E)  18  mm;  (F)  470  mm.  HA  =  haemal  arch,  NA  =  neural  arch,  other  abbreviations  as  in  Figs.  359,  360.  Redrawn 
from  Fabre-Domergue  and  Bietrix  (1905). 


HENSLEY  AND  AHLSTROM:  PLEURONECTIFORMES 


679 


PUS- 


PUS 


HY5 


PHY 


0.5  mm 


0.5  mm 


HY1  HY4 


HY3 


HY4 


-HY3 


B 


0.5  mm 


0.5  mm 


HY4 

HYS/ Hys 
NS         EP 


PHY 


HY5 


THC+HY    1-4 


0.5  mm 


H 


0.5  mm 


680 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


HY5 


_THC  +  HY2-4 


HY  5 


THC  +  HY  1 


Fig.  363.  Caudal  skeleton  of  Tephnnectes  sinensis.  Hypural  pattern 
1  (upper);  caudal  skeleton  of  Tnnectes  fimbhata.  Hypural  pattern  3 
(middle);  and  caudal  skeleton  of  Rhombosolea  pleheia.  Hypural  pattern 
4  (lower).  Abbreviations  as  in  Fig.  359.  "V"  on  distal  end  of  fin  ray 
indicates  dorsal-  and  ventralmost  branched  ray. 


used  as  evidence  that  soleids  and  cynoglossids  are  most  closely 
related  are  plesiomorphic  for  the  order  (symmetrical  nasal  or- 
gans, dimorphic  optic  chiasmata),  found  in  some  pleuronectoids 
but  dismissed  as  parallelisms  [lower  jaw  not  prominent,  absence 


of  postcleithra,  several  "soleoid  characters"  found  in  rhombo- 
soleines,  (see  Norman,  1934)],  or  are  incorrect  (absence  of  all 
ribs).  Other  states  used  to  unite  the  soleoid  families  include;  (1) 
a  preopercular  margin  covered  by  skin  and  scales;  and  (2)  skin 
covering  the  dentary  and  interopercular  bones  being  continuous 
across  the  chin,  hiding  the  isthmus  and  branchioslegal  rays  (Nor- 
man, 1966).  A  covered  preopercular  margin  is  not  limited  to 
soleoids;  it  occurs  in  some  rhombosoleine  genera  (Chabanaud, 
1949;  Hensley,  pers.  observ.).  The  second  state  as  well  as  the 
absence  of  pleural  ribs  are  possible  synapomorphies  for  the 
group. 

Cynoglossidae.—lhtrt  is  little  doubt  that  the  tonguesoles  are 
monophyletic.  They  are  unique  in  having  the  ventral  fin  of  the 
blind  side  oriented  along  the  midventral  line  and  the  ocular  fin 
placed  more  dorsally  or  missing.  The  relationship  of  this  family 
to  other  groups,  however,  is  obscure  (see  Soleidae). 

Soleidae.—The  main  character  state  proposed  as  uniting  the 
two  soleid  subfamilies  (Soleinae,  Achirinae)  appears  to  be  that 
all  species  are  dextral.  This  is  still  a  poorly  known  group,  and 
we  are  not  prepared  to  make  much  of  a  contribution  here.  How- 
ever, there  are  some  marked  differences  between  these  subfam- 
ilies. In  several  characters,  the  Achirinae  are  more  primitive 
than  originally  thought.  Some  species  have  hypural  pattern  1 , 
the  most  primitive.  In  species  where  hypural  fusions  have  oc- 
curred, the  first  hypural  remains  free  and  articulates  with  the 
terminal  half  centrum  (Fig.  363  middle).  The  haemal  spine  of 
the  second  preural  centrum  is  autogenous  (i.e.,  the  plesiomorph- 
ic state  for  the  order)  in  achirines.  Uroneurals  may  still  be  pres- 
ent in  some  species.  Although  postcleithra  are  lacking  in  adult 
soleoids,  at  least  one  achirine  species  has  them  during  larval 
development  (Futch  et  al.,  1972).  Soleines  differ  from  achirines 
in  these  characters  in  that  they  show  what  appear  to  be  more 
derived  states.  The  Soleinae  have  hypurals  1-4  fused  to  the 
terminal  half  centrum  (Fig.  362F),  the  haemal  spine  of  the  sec- 
ond preural  centrum  is  attached,  there  is  no  indication  of  uro- 
neurals, and  postcleithra  have  not  been  reported  in  larvae  or 
adults.  Soleines  share  these  states  with  the  Cynoglossidae.  In 
addition,  both  groups  have  vertebral  transverse  apophyses,  which 
are  missing  in  achirines.  The  possibilities  that  the  Soleidae  are 
not  monophyletic  and  the  Soleinae  are  more  closely  related  to 
the  Cynoglossidae  should  be  more  thoroughly  explored. 

Pleuronectoidei. —  Some  of  the  character  states  used  to  define 
this  group  are  plesiomorphic  for  the  order:  (1)  preoperculum 
with  free  margin;  (2)  presence  of  postcleithra;  and  (3)  presence 
of  pleural  and  epipleural  ribs.  Some  apomorphic  states  for  the 
order  are  not  limited  to  pleuronectoids;  e.g.,  loss  of  dorsal  and 
anal  spines.  The  Regan-Norman  model  has  used  the  position 
of  the  nasal  organ  of  the  blind  side  to  separate  pleuronectoids 
from  soleoids  and  psettodids.  In  pleuronectoids,  this  nasal  organ 
follows  the  migrating  eye  during  metamorphosis.  After  meta- 
morphosis, it  remains  near  the  dorsal  edge  of  the  head.  This 
was  interpreted  as  a  specialization  of  pleuronectoids,  except  that 
this  state  does  not  occur  in  all  Rhombosoleinae  (i.e.,  nasal  organs 
remain  symmetrically  placed).  Thus,  it  is  not  a  synapomorphy 
for  the  group,  unless  it  can  be  shown  that  the  nasal-organ  sym- 
metry in  these  rhombosoleines  was  secondarily  derived  from 
the  asymmetrical  state.  We  have  not  done  a  survey  of  nasal- 
organ  symmetry,  but  incidental  observations  indicate  that  the 
supposed  differences  between  these  states  (i.e.,  symmetrical  vs 


HENSLEY  AND  AHLSTROM:  PLEURONECTIFORMES 


681 


asymmetrical  placement)  are  not  as  great  as  formerly  thought. 
Loss  of  a  truly  dimorphic  optic  chiasma  would  appear  to  be  the 
only  synapomorphy  proposed  to  date  uniting  the  pleuronec- 
toids.  However,  as  previously  discussed,  a  basically  dimorphic 
or  monomorphic  optic  chiasma  has  been  demonstrated  in  very 
few  pleuronectoid  species. 

One  might  expect  that  we  are  well  informed  about  the  inter- 
relationships among  pleuronectoids.  Unfortunately,  all  of  the 
past  work  has  used  the  eclectic  approach.  Thus,  scophthalmids 
and  citharids  have  been  related  horizontally  as  primitive  pleu- 
ronectoids, and  bothids,  paralichthyids,  and  pleuronectids  as 
higher  groups.  Again,  an  important  character  here  is  the  optic 
chiasma.  Two  states  were  recognized  in  pleuronectoids:  (1)  the 
primitive  one  (for  pleuronectoids)  where  the  nerve  of  the  mi- 
grating eye  is  always  dorsal  regardless  of  eye  position  (i.e..  the 
basically  dimorphic  state);  and  (2)  the  monomorphic  state  char- 
acteristic of  "'higher"  pleuronectoids  where  the  chiasma  is  fixed 
regardless  of  reversals.  It  has  been  assumed  that  all  bothids, 
pleuronectids.  and  paralichthyids  show  the  monomorphic  state. 
Some  evidence  from  other  characters  indicates  this  assumption 
is  not  valid. 

Due  mainly  to  the  work  of  Amaoka  (1969)  and  one  of  us 
(Ahlstrom),  we  have  a  good  survey  of  the  caudal-fin  complex 
of  pleuronectoids.  Patterns  of  hypurals  1^  are  fairly  well  known. 
The  distributions  of  these  patterns  call  into  question  much  of 
the  current  evolutionary  model  and  classification  of  the  Pleu- 
ronectoidei.  There  are  five  patterns  of  hypurals  in  this  group  as 
defined  in  the  Regan-Norman  model:  Pattern  1  (Fig.  363  up- 
per)—This  is  plesiomorphic  for  the  order.  Pattern  5  (Fig.  364 
upper)— This  pattern  is  limited  to  the  Samarinae.  We  interpret 
this  pattern  as  a  synapomorphy  uniting  the  samarines.  Pattern 
2  (Fig.  361)— Within  pleuronectoids  this  pattern  seems  to  be 
limited  to  Citharoides.  It  is  apparently  derived  from  pattern  1. 
Pattern  6  (Figs.  360,  364  middle  and  lower)— This  is  an  apo- 
morphic  pattern  which  is  very  distinctive.  We  consider  it  ho- 
mologous in  pleuronectoids  where  it  occurs  and  a  synapomor- 
phy uniting  these  groups.  Again,  we  are  calling  this  group  the 
bothoids  and  it  includes  the  Pleuronectinae.  Paralichthyidae 
(except  Tephrinectes  and  Thysanopsetta).  Scophthalmidae, 
Bothidae  (except  Mancopsetta),  and  Brachypleura.  Pattern  4 
(Fig.  363  lower)- Within  the  pleuronectoids  this  pattern  is  lim- 
ited to  certain  genera  of  Rhombosoleinae  and  Eucitharus.  Based 
on  other  characters,  the  homology  of  pattern  4  between  these 
groups  is  probably  not  true. 

Citharidae.  — Many  character  states  used  to  define  this  family 
(Hubbs,  1945.  1946)  are  plesiomorphic  for  the  order:  (1)  reten- 
tion of  pelvic  spines;  (2)  retention  of  supramaxillae  (Eucitharus 
and  Citharoides):  (3)  urinary  papilla  close  to  anus;  (4)  no  union 
of  branchiostegals;  (5)  retention  of  vomerine  teeth  {Eucitharus, 
Brachypleura.  Lepidoblepharon);  and  (6)  retention  of  short-based 
ventral  fins.  Some  are  plesiomorphic  for  the  Pleuronectoidei: 

( 1 )  basically  dimorphic  optic  chiasma  (at  least  in  Citharoides); 

(2)  gill  membranes  showing  some  degree  of  union,  but  still  fairly 
widely  separated;  and  (3)  loss  of  dorsal  and  anal-fin  spines.  The 
only  possible  character  state  proposed  to  date  that  could  be 
interpreted  as  a  synapomorphy  for  this  family  is  the  position  of 
the  anus  on  the  ocular  side.  Although  we  have  not  examined 
many  specimens  for  this  character,  it  appears  that  deflection  of 
the  anus  to  the  ocular  side  is  probably  slight.  Amaoka  (1972b) 
examined  Brachypleura  and  attempted  to  redefine  the  Cithar- 
idae. However,  he  still  showed  no  synapomorphies  for  the  group. 


EP(HY5''l 


HY5(4''I 


THC  *  HY2  -4(2+371 


HY5  *  EP 


THC*HY3  +  4 


HYl  +2 


Fig.  364.  Caudal  skeleton  of  Samanscus  iriocellatus.  Hypural  pat- 
tern 5  (upper),  caudal  skeleton  of  Cilhanchthys  macrops.  Hypural  pat- 
tern 6  (middle),  and  caudal  skeleton  of  Hippoglossina  oblonga.  Hypural 
pattern  6  (lower).  SR  =  splinter  ray.  other  abbreviations  as  in  Fig.  359. 
"V  on  distal  end  of  fin  ray  indicates  dorsal-  and  ventralmost  branched 
ray. 


682 


ONTOGENY  AND  SYSTEMATICS  OF  HSHES-AHLSTROM  SYMPOSIUM 


The  family  Citharidae  as  presently  defined  is  a  grade.  Ex- 
amination of  the  caudal  osteology  has  shown  two  derived  and 
one  plesiomorphic  pattern  of  hypurals.  Lepidoblepharon  shows 
pattern  1,  which  is  plesiomorphic  for  the  order.  Citharoides 
shows  pattern  2,  a  derived  pattern  (Fig.  361).  This  pattern  could 
represent  a  state  on  a  line  leading  toward  pattern  6,  which  is 
shown  by  Brachypleura.  Eucitharus  shows  pattern  4,  which  pos- 
sibly developed  independently  in  some  rhombosoleines.  The 
most  obvious  result  of  this  is  that  Brachypleura  belongs  to  the 
bothoid  group,  which  shares  the  derived  hypural  pattern  6.  In 
this  interpretation,  the  character  states  shown  by  Brachypleura 
that  are  primitive  for  the  order  (e.g.,  vomerine  teeth,  ventral- 
fin  spines)  are  also  primitive  for  bothoids. 

Scophthalmidae.— Based  on  ventral-fin  morphology,  the  Scoph- 
thalmidae  appear  to  be  monophyletic.  There  are  certain  simi- 
larities in  ventral-fin  morphology  between  this  family  and  the 
achirines,  but  these  are  probably  superficial.  Scophthalmids  were 
previously  thought  to  be  closely  related  to  and  derived  from  the 
Citharidae  (Hubbs,  1945).  This  hypothesis  was  based  on  certain 
symplesiomorphies,  e.g.,  the  low  degree  of  fusion  of  the  gill 
membranes  and  the  presence  of  vomerine  teeth.  The  Scoph- 
thalmidae show  hypural  pattern  6  and  are  thus  members  of  the 
bothoid  group. 

Paralichthyidae.— 'Norman  (1934)  basically  defined  the 
Paralichthyinae  (=Paralichthyidae  with  modifications)  on  ex- 
ternal pelvic-fin  morphology  and  vertebral  structure  (absence 
of  transverse  apophyses).  The  group  was  supposed  to  have  the 
ventral  fins  nearly  symmetrical  in  position  and  base  lengths,  or 
the  ocular  fin  on  the  midventral  line  and  its  base  slightly  ex- 
tended anteriorly.  Symmetries  in  ventral-fin  position  and  base 
lengths  are  plesiomorphic  for  the  order  and  bothoids.  Norman's 
paralichthyid  genera  with  an  ocular  ventral  fin  on  the  midventral 
line  and  its  base  extended  anteriorly  are  bothids  (i.e.,  Trichop- 
setta,  Engyophrys,  Taeniopsetta,  Monolene.  Perissias). 

Amaoka  (1969)  presented  a  more  thorough,  detailed  defini- 
tion of  the  family.  However,  many  or  most  of  the  character 
states  he  used  appear  to  be  plesiomorphic  for  bothoid  fishes 
(i.e.,  those  defined  by  hypural  pattern  6).  A  second  limitation 
of  Amaoka's  work  on  this  group  is  that  it  was  limited  to  three 
genera  {Paralichthys,  Pseudorhombus.  Tarphops).  An  important 
change  in  Norman's  classification  was  made  by  Amaoka  when 
he  removed  Taeniopsetta  from  the  Paralichthyidae  and  placed 
it  in  the  Bothidae.  Hensley  (1977)  and  Futch  (1977)  did  the 
same  for  Monolene,  Engyophrys,  and  Trichopsetta. 

We  have  now  examined  some  characters  in  the  remaining 
Paralichthyidae,  and  additional  changes  are  required  in  the  com- 
position of  this  group.  In  a  survey  of  caudal-fin  structure,  it  was 
found  that  Thysanopsetta  and  Tephrinectes  show  the  most  prim- 
itive type  of  hypural  pattern  (1;  Fig.  363  upper).  These  two 
genera  are  much  more  primitive  than  expressed  in  the  current 
classification  and  definitely  do  not  belong  to  the  bothoid  group. 

Within  the  remaining  Paralichthyidae  another  group  is  dis- 
cemable.  This  is  composed  of  Cyclopsetta,  Syacium,  Citharich- 
thys,  and  Etropus,  i.e.,  the  Cyclopsetta  group.  States  for  two 
complexes  of  characters,  ventral-fin  morphology  and  urinary- 
papilla  position,  are  unique  to  this  group  and  interpreted  as 
synapomorphic.  Arrangement  of  caudal-fin  rays  in  the  Cyclop- 
setta group  is  also  unique  and  probably  apomorphic  (Fig.  364 
middle).  All  species  have  1 7  caudal  rays,  none  of  which  are 


supported  by  preural,  neural  or  haemal  spines.  It  should  also 
be  noted  that  the  fifth  hypural  has  fused  with  an  epural.  This 
fusion  has  been  observed  in  larval  development  (Tucker.  1982; 
Ahlstrom,  pers.  observ.).  However,  fusion  of  the  fifth  hypural 
and  one  or  more  epurals  has  apparently  occurred  several  times 
in  pleuronectiforms,  possibly  including  the  bothoids  (e.g.,  see 
Fig.  360).  A  detailed  analysis  of  relationships  between  the  Cy- 
clopsetta group  and  other  bothoids  is  not  possible  here.  How- 
ever, some  character  states  may  indicate  a  close  relationship 
with  bothids  (absence  of  first  neural  spine,  presence  of  vertebral 
transverse  apophyses). 

Amaoka  (1 969)  and  one  of  us  (Ahlstrom)  recognized  another 
group  within  the  Paralichthyidae  composed  of  Pseudorhombus, 
Tarphops.  and  Cephalopsetla.  i.e..  the  Pseudorhombus  group. 
We  interpret  these  genera  as  more  specialized  in  certain  char- 
acters than  most  other  members  of  the  family.  Species  of  this 
group  usually  have  a  total  caudal  ray  count  of  17,  the  epural  is 
fused  to  the  fifth  hypural,  and  they  lack  a  splinter  ray  on  the 
ventralmost  caudal-fin  ray.  With  the  exceptions  noted  above 
(the  primitive  non-bothoid  genera  Tephrinectes  and  Thysan- 
opsetta and  the  Cyclopsetta  group),  the  remaining  paralichthyids 
of  the  Regan-Norman  classification  (what  we  are  calling  the 
Paralichthys  group)  have  the  apparently  plesiomorphic  states 
of  1 8  caudal  rays,  at  least  one  free  epural  (except  in  one  species 
of  Hippoglossina  (Sumida  et  al.,  1979)),  and  a  splinter  ray  on 
the  ventralmost  caudal-fin  ray  (Fig.  364  lower).  The  splinter  ray 
is  probably  a  remnant  of  a  ray  lost  through  fusion  with  an 
adjacent  ray  (Okiyama,  1974).  The  Pseudorhombus  group  may 
be  definable  by  synapomorphies  but  a  detailed  analysis  has  not 
been  done. 

After  removal  of  the  bothids  (Trichopsetta,  Engyophrys,  Tae- 
niopsetta, Monolene,  Perissias)  and  the  primitive  non-bothoid 
genera  (Tephrinectes,  Thysanopsetta),  recognition  of  the  Cy- 
clopsetta group  as  monophyletic.  and  recognition  of  the  Pseu- 
dorhombus group  as  possibly  monophyletic,  few  of  the  original 
paralichthyid  genera  remain.  We  have  been  referring  to  these 
as  the  Paralichthys  group  (Ancylopsetta,  Gastropsetta,  Hippo- 
glossina, Lioglossina,  Paralichthys,  Verecundum,  and  Xystreu- 
rys).  At  least  most  of  the  character  states  known  for  these  re- 
maining genera  are  plesiomorphic  for  the  order  (e.g.,  symmetrical 
ventral-fin  states)  or  for  bothoids  (e.g..  usual  presence  of  at  least 
one  free  epural).  The  Paralichthys  group  is  probably  not  mono- 
phyletic. 

Bothidae.— Norman  (1934)  defined  the  Bothinae  (=Bothidae 
with  modifications)  on  the  basis  of  a  high  degree  of  ventral-fin 
asymmetry  and  the  presence  of  vertebral  transverse  apophyses. 
The  ocular  ventral  fin  was  said  to  be  on  the  midventral  line  with 
its  base  extending  anteriorly  to  the  urohyal.  Norman  excluded 
Taeniopsetta,  Engyophrys,  Trichopsetta,  Monolene,  and  Peris- 
sias from  this  group  because  the  base  of  the  ocular  ventral  fin, 
although  on  the  midventral  line  and  somewhat  longer  than  that 
of  the  blind  side,  does  not  extend  to  the  urohyal. 

Amaoka  (1969)  examined  many  bothid  genera  and  redefined 
the  family  using  more  characters.  Most  of  the  characters  stressed 
by  Amaoka  have  now  been  examined  in  other  bothoids.  These 
are  discussed  below: 

Ventral-fin  asymmetry.— In  bothids  the  ocular  fin  base  is  on  the 
midventral  line,  elongated,  and  has  its  origin  anteriorly  placed 
relative  to  the  base  of  the  blind  fin.  Within  the  bothoids  this 
combination  of  states  appears  to  be  derived  and  unique. 


HENSLEY  AND  AHLSTROM:  PLEURONECTIFORMES 


683 


Preorbital  on  blind  side.— This  bone  is  absent  in  the  Bothidae. 
It  appears  to  be  present  in  all  other  bothoids  (Pleuronectinae 
not  examined  for  this  character).  Based  on  this  comparison,  we 
interpret  the  loss  of  this  bone  a  derived  state  within  the  bothoids 
defining  the  family  Bothidae. 

Infraorbital  bones  of  the  ocular  side.  — All  bothids  have  an  ocu- 
lar preorbital  bone  but  lack  the  remainder  of  the  series.  The 
presence  or  absence  of  the  ocular  preorbital  has  not  been  sur- 
veyed in  most  bothoid  groups.  However,  an  ocular  infraorbital 
lateral  line  is  present  in  most  bothoids.  In  addition  to  the  Both- 
idae, it  is  missing  in  Brachypleura  and  the  Cyclopsetta  group. 

Intermuscular  bones.— We  interpret  the  presence  of  at  least  two 
of  the  series  of  these  bones  (myorhabdoi)  as  a  derived  state 
unique  to  and  defining  the  Bothidae. 

First  neural  spine.— Although  the  first  neural  arch  is  present, 
the  neural  spine  is  missing  in  the  Bothidae.  It  is  present  in  all 
other  bothoids  except  the  Cyclopsetta  group. 

Vertebral  transverse  apophyses.— All  bothids  have  two  pairs  of 
transverse  apophyses  on  most  vertebrae.  As  previously  dis- 
cussed, how  to  interpret  these  on  the  pieuronectiform  level  and 
within  the  bothoid  group  is  questionable.  Within  the  bothoids 
well-developed  and  very  similar  structures  occur  only  in  the 
Bothidae  and  Scophthalmidae.  Very  small  transverse  apophyses 
also  occur  in  the  Cyclopsetta  group. 

Based  on  these  characters,  the  Bothidae  appear  to  be  mono- 
phyletic  and  definable  by  synapomorphies  in  at  least  three  char- 
acters or  complexes:  (1)  loss  of  the  preorbital  on  the  blind  side; 
(2)  presence  of  myorhabdoi;  and  (3)  asymmetrical  states  of  ven- 
tral-fin morphology. 

Since  Amaoka's  ( 1 969)  work,  we  have  examined  the  remain- 
ing genera  not  examined  by  him  that  have  been  considered 
bothids  (i.e.,  Graminatobothus.  Lophonectes,  Pelecanichlhys, 
Mancopsetta).  All  of  these  except  Mancopsetta  are  bothids. 
Mancopsetta  exhibits  the  following  character  states:  ( 1 )  hypural 
pattern  1,  i.e.,  the  most  primitive  type;  (2)  presence  of  pleural 
and  epipleural  ribs,  but  no  myorhabdoi  or  other  intermuscular 
bones  in  the  caudal  region;  (3)  at  least  one  free  epural  (none  in 
adult  bothids);  (4)  anus  on  midventral  line  (clearly  on  blind  side 
in  bothids);  (5)  no  vertebral  transverse  apophyses;  and  (6)  seven 
rays  in  the  ocular  ventral  fin,  5-7  in  that  of  the  blind  side  (six 
in  both  fins  in  bothids).  These  are  all  characters  in  which  Man- 
copsetta differs  from  the  Bothidae.  Due  to  the  primitive  hypural 
pattern,  it  is  not  a  bothoid  (see  Rhombosoleinae). 

Amaoka  (1969)  analyzed  intergeneric  relationships  of  Jap- 
anese bothids.  However,  his  analysis  was  eclectic  and  did  not 
include  all  genera  (i.e.,  Engyophrys.  Trichopsetta,  Monolene, 
Perissias.  Graminatobothus.  Lophonectes.  and  Pelecanichlhys 
were  not  examined).  He  recognized  two  subfamilies,  the  Tae- 
niopsettinae  and  Bothinae.  He  erected  the  first  subfamily  for 
Taeniopsetta.  Hensley  (1977),  Futch  (1977),  Evseenko  (1977, 
1981),  and  Amaoka  (1979)  implied  that  Engyophrys  and  Tri- 
chopsetta should  be  included  in  the  Taeniopsettinae.  This  was 
done  on  the  basis  of  larval  characters  and  ventral-fin  morphol- 
ogy. Most  of  the  slates  used  to  define  the  Taeniopsettinae  were 
considered  by  Amaoka  (1969)  to  be  plesiomorphic  at  the  family 
level.  Three  characters  were  emphasized:  (1)  degree  of  anterior 


extension  of  the  base  of  the  ocular  ventral  fin;  (2)  shape  of  the 
ventral  (sciatic)  area  of  the  urohyal;  and  (3)  number  of  suborbital 
bones  on  the  blind  side.  In  the  taeniopsettines,  the  origin  of  the 
blind  ventral  fin  is  at  the  same  transverse  level  as  the  second 
ray  of  the  ocular  ventral  fin,  i.e.,  the  base  of  the  ocular  fin  is 
only  slightly  elongated.  In  the  Bothinae,  extension  of  the  base 
of  the  ocular  fin  is  greater  and  the  origin  of  the  blind  fin  is  on 
the  same  transverse  level  as  the  third  or  fourth  ray  of  the  ocular 
fin.  Obviously,  the  taeniopsettine  state  here  is  the  more  plesio- 
morphic. Engyophrys,  Trichopsetta,  Monolene,  and  Perissias 
show  this  state.  Taeniopsetta  has  a  broad,  truncate  margin  on 
the  sciatic  part  of  the  urohyal.  In  bothines,  this  area  of  the 
urohyal  is  pointed.  Amaoka  (1969)  clearly  showed  that  the  ple- 
siomorphic state  for  bothoids  is  closer  to  the  condition  shown 
in  taeniopsettines.  Engyophrys,  Trichopsetta,  and  Perissias  show 
the  taeniopsettine  condition,  Monolene  the  bothine  state. 
Amaoka  (1969)  noted  an  apparent  trend  among  bothoids  in 
reduction  of  the  number  of  suborbital  bones  of  the  blind  side. 
This  reduction  may  have  occurred  in  several  bothoid  groups 
and  interpretation  of  this  character  is  not  clear.  Thus,  infraor- 
bital counts  for  bothoids  are  as  follows  (preorbital  -t-  suborbit- 
als): Scophthalmidae  1+5;  Brachypleura  1  +  0;  Paralichthys 
group  1  +  4-5;  Pseudorhombus  group  1  +  5-7;  Cyclopsetta 
group  1  +  5-6;  and  Bothidae  0  +  3-5.  Pleuronectines  were  not 
examined  for  this  character.  The  most  common  count  in  both- 
oids other  than  bothids  is  1  +  5-7.  Thus,  there  is  some  evidence 
that  the  basal  or  plesiomorphic  count  for  bothids  may  be  five 
suborbitals  on  the  blind  side.  Among  bothids  this  count  appar- 
ently occurs  only  in  Taeniopsetta  and  Pelecanichthys.  Engyo- 
phrys, Trichopsetta,  Perissias,  and  Monolene  have  three  sub- 
orbitals on  the  blind  side.  In  summary,  there  is  good  evidence, 
at  least  for  the  first  two  characters  discussed  above,  that  the 
Taeniopsettinae  show  states  that  are  plesiomorphic  for  the  fam- 
ily and  may  not  be  monophyletic. 

Pleuronectidae.— Norman  (1934)  considered  this  family  to  be 
one  of  the  "higher"  flatfish  groups,  i.e.,  those  with  a  mono- 
morphic  optic  chiasma.  Hubbs  (1945)  basically  followed  this 
interpretation,  but  showed  that  two  of  Norman's  pleuronectid 
genera,  Brachypleura  and  Lepidoblepharon.  possessed  some 
primitive  states  not  shown  in  other  pleuronectids.  These  two 
genera  were  removed  by  Hubbs  and  placed  in  his  family  Cith- 
aridae. 

Norman  (1934)  defined  the  Pleuronectidae  as  being  dextral 
and  having  eggs  without  oil  globules.  Basic  to  his  concept  of 
this  family  were  the  assumptions  that  all  members  were  mono- 
morphic  in  regard  to  the  optic  chiasma  and  that  nearly  all  species 
were  discriminately  dextral.  He  divided  the  family  into  five 
subfamilies.  All  members  of  the  Poecilopsettinae,  Paralichthod- 
inae,  Samarinae.  and  Rhombosoleinae,  as  presently  interpreted, 
are  discriminately  dextral,  i.e.,  sinistral  individuals  occur  so 
rarely  in  any  one  species  that  they  can  be  considered  anomalies. 
Most  species  of  Pleuronectinae  are  also  discriminately  dextral. 
The  few  exceptions  have  probably  returned  to  indiscriminate 
ocular  asymmetry  secondarily  (Hubbs  and  Hubbs,  1945).  We 
have  no  reason  to  doubt  Norman's  or  Hubbs'  assumption  that 
the  Pleuronectinae  have  a  monomorphic  optic  chiasma.  How- 
ever, as  previously  discussed,  there  are  no  data  showing  this  for 
the  other  pleuronectid  subfamilies.  Uniting  these  groups  in  the 
family  Pleuronectidae  appears  to  have  rested  only  on  ocular 
asymmetry.  We  have  surveyed  these  subfamilies  for  various 


684 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


characters  and  are  confident  that  the  Pleuronectidae  as  currently 
defined  are  not  monophyletic.  In  fact,  four  of  the  pleuronectid 
subfamilies  are  not  bothoids  as  we  define  the  group.  However, 
what  the  true  relationships  of  these  groups  are  is  unknown.  We 
discuss  these  subfamilies  individually: 

Poecilopsettinae.— We  have  examined  radiographs  of  speci- 
mens ofPoecilopsetta  and  Nematops.  These  genera  have  hypural 
pattern  1,  at  least  one  free  epural,  20  caudal  rays,  and  what 
appears  to  be  a  haemal-arch  remnant  on  the  parhypural.  The 
caudal  structure  here  is  primitive  compared  to  the  bothoids  and 
these  fishes  do  not  belong  to  that  group.  Poecilopsettines  are 
poorly  known  and  character  states  defining  the  group  or  relating 
it  to  others  have  not  been  investigated. 

?ar?i\ic)M\iod.mdie:. —Paralichthodes  algoensis  has  hypural  pat- 
tern 1  (Ahlstrom,  pers.  observ.)  and  does  not  belong  to  the 
bothoid  group.  Its  relationships  to  other  groups  are  unknown. 

Samarinae.— Since  Hubbs'  (1945)  removal  oi  Brachypleura  and 
Lepidoblepharon  from  this  group,  it  has  been  composed  of  Sa- 
maris  and  Samanscus.  We  have  not  done  a  detailed  study  of 
these  genera,  but  some  characters  we  have  examined  are  worthy 
of  note:  ( 1 )  These  genera  show  a  unique  hypural  pattern  (5;  Fig. 
364  upper).  We  interpret  this  pattern  as  derived  relative  to 
pattern  1  and  as  indicative  that  the  group  is  monophyletic.  Using 
this  pattern  to  relate  the  group  is  more  difficult:  however,  one 
of  us  (Ahlstrom)  noted  that  in  late-stage  larvae  of  Samanscus. 
hypural  pattern  1  is  present,  and  fusions  resulting  in  pattern  5 
must  occur  very  late  in  development.  This  is  evidence  that 
pattern  5  may  have  evolved  directly  from  pattern  1  and  does 
not  represent  a  modification  of  the  bothoid  pattern  6.  (2)  Sa- 
marines  are  the  only  pleuronectiforms  known  other  than  the 
Bothidae  to  have  intermuscular  bones,  although  they  do  not 
have  the  two  series  of  myorhabdoi  as  found  in  bothids.  We  have 
not  done  a  detailed  study  of  these  bones  in  samarines,  but  they 
appear  very  similar  to  the  epimerals,  epicentrals.  and  hypom- 
erals  of  bothids.  (3)  Samarines,  cynoglossids,  and  soleines  have 
an  anterior  pair  of  well-developed  transverse  apophyses  on  many 
vertebrae.  Two  pairs  of  these  structures  are  found  in  the  Both- 
idae and  Scophthalmidae.  (4)  The  Samannae,  Soleoidei,  and 
Mancopselta  lack  postcleithra,  at  least  in  adults.  How  to  inter- 
pret these  last  three  character  states  is  open  to  question.  Are 
three  of  the  series  of  intermuscular  bones  homologous  in  sa- 
marines and  bothids?  Are  the  anterior  vertebral  transverse 
apophyses  homologous  between  all  of  the  groups?  Do  some  of 
these  character  states  indicate  a  close  relationship  between  sa- 
marines and  some  soleoids  (i.e.,  cynoglossids  and  soleines)?  Our 
tentative  hypothesis  is  that  the  samarines  are  a  line  that  is  at 
least  independent  from  the  bothoids.  Here  we  are  obviously 
stressing  caudal  characters.  The  corollary  of  this  is  that  we  are 
interpreting  similarities  between  samarines  and  bothoids  in  in- 
termuscular bones  and  vertebral  transverse  apophyses  as  hom- 
oplasies. 

Rhombosoleinae.  — The  main  character  states  used  by  Norman 
(1926,  1934)  to  define  this  subfamily  were  the  high  degree  of 
asymmetry  in  the  ventral  fins  and  the  absence  of  pectoral  radials. 
The  ocular  ventral  fin  is  on  the  midventral  line  and  its  base  is 
considerably  extended.  The  blind  ventral  fin  is  short  based  or 
missing.  Another  interesting  characteristic  of  this  group  is  that 


several  genera  show  high  numbers  of  fin  rays  in  the  ocular  ven- 
tral fin.  There  is  a  great  deal  of  morphological  diversity  in  rhom- 
bosoleines.  Some  genera  appear  fairly  generalized  in  many  char- 
acters (Oncopterus.  Psammodiscus.  Rhombosolea.  Azygopus.  and 
Pelotretis);  others  are  more  specialized  (Colistiuin.  Peltorham- 
phits,  and  Ammotretis).  Many  of  the  specializations  in  the  latter 
genera  are  similar  to  those  in  some  soleoids.  This  has  been 
interpreted  as  parallel  evolution  (Norman,  1934;  Hubbs,  1945). 
Norman  apparently  had  some  doubts  about  aligning  this  group 
with  the  Pleuronectinae.  He  realized  that  Parker's  (1903)  ex- 
amination of  one  specimen  of  Oncopterus  darwinii  in  his  survey 
of  optic  chiasmata  did  not  prove  the  group  to  be  monomorphic 
in  this  character.  This  group  has  still  not  been  studied  in  detail. 
It  may  be  monophyletic,  but  its  relationship  to  other  flatfishes 
is  unknown. 

We  have  examined  the  caudal  skeleton  of  all  rhombosoleine 
genera  except  Psammodiscus.  They  show  hypural  patterns  1  and 
4  (Fig.  363  upper  and  lower).  Assuming  the  group  is  monophy- 
letic, there  are  two  implications  here:  (1)  The  primitive  pleu- 
ronectiform  hypural  pattern  1  is  also  plesiomorphic  for  the 
Rhombosoleinae,  and  the  derived  pattern  4  arose  within  the 
group  independently  from  the  same  pattern  in  the  Soleinae, 
Cynoglossidae,  and  Eucttharus.  (2)  The  Rhombosoleinae  are 
not  bothoids  and  should  not  be  aligned  with  the  Pleuronectinae. 

The  possibility  has  recently  become  apparent  that  Mancop- 
setla  may  be  most  closely  related  to  the  Rhombosoleinae.  All 
known  specimens  of  Mancopselta  are  sinistral  and  it  has  been 
considered  a  bothid.  However,  it  shares  certain  character  states 
with  at  least  some  rhombosoleines.  This  genus  has  ventral-fin 
ray  counts  of  7  on  the  ocular  side  and  5-7  on  the  blind  side. 
Although  not  strictly  limited  to  the  rhombosoleines,  these  high 
counts,  at  least  in  the  fin  of  the  ocular  side,  are  characteristic  of 
at  least  four  rhombosoleine  genera.  The  eyes  are  densely  scaled 
in  Mancopsetta  and  in  Azygopus  and  Pelotretis.  However,  scaled 
eyes  are  found  in  some  genera  of  other  groups  also  (e.g.,  some 
pleuronectines).  Andnashev  (1960)  and  Penrith  (1965)  have 
both  remarked  on  a  fleshy  lip-like  structure  which  overhangs 
the  anterior  end  of  the  upper  jaw  in  .Mancopsetta.  One  of  the 
soleoid-type  characteristics  exhibited  by  the  more  specialized 
rhombosoleines  is  the  dorsal  fin  originating  in  a  rostral  hook 
that  overhangs  the  mouth.  In  the  more  generalized  genera,  there 
is  no  rostral  hook  and  the  dorsal  fin  originates  at  some  posterior 
position.  In  at  least  one  of  these  generalized  genera  (Azygopus. 
the  only  one  examined  for  this  character)  there  is  a  fleshy  struc- 
ture (possibly  a  precursor  to  the  rostral  hook?)  overhanging  the 
anterior  end  of  the  upper  jaw  which  is  very  similar  to  that  in 
Mancopsetta.  Obviously  more  comparative  work  needs  to  be 
done  here.  However,  it  is  possible  that  Mancopsetta  and  the 
Rhombosoleinae  may  form  a  monophyletic  group  with  an  in- 
discriminately dextral  or  sinistral  common  ancestor. 

Pleuronectinae.  — Norman  (1934)  stressed  two  character  states 
in  defining  this  subfamily:  ( 1 )  lateral  line  well  developed  on  both 
sides  of  the  body;  and  (2)  olfactory  laminae  parallel  (except  in 
Atheresthes),  without  rachis.  A  well-developed  lateral  line  on 
both  sides  of  the  body  is  plesiomorphic  for  the  order  and  both- 
oids. We  have  not  examined  olfactory  laminae  or  attempted  to 
analyze  distributions  of  states  for  the  character. 

We  have  shown  that  the  Pleuronectidae  is  probably  not  mono- 
phyletic, due  to  the  inclusion  of  the  four  non-bothoid  subfam- 
ilies. The  subfamily  Pleuronectinae  is  the  only  bothoid  group 


HENSLEY  AND  AHLSTROM:  PLEURONECTIFORMES 


685 


in  Norman's  Pleuronectidae.  Members  of  this  subfamily  are 
dextral  or  apparently  secondarily  indiscriminate  (Hubbs  and 
Hubbs,  1945).  They  apparently  have  a  monomorphic  optic 
chiasma.  However,  most  character  states  which  species  of  this 
subfamily  share  appear  to  be  plesiomorphic  for  the  order  or 
bothoids.  e.g..  symmetrical  or  nearly  symmetrical  ventral-fin 
placement  and  fin-base  lengths,  anus  on  or  close  to  the  mid- 
ventral  line.  We  have  examined  the  caudal  osteology  of  about 
half  of  the  pleuronectine  genera.  All  have  the  bothoid  hypural 
pattern  (6)  and  one  or  possibly  two  free  epurals.  We  have  found 
no  synapomorphies  in  the  caudal  fin  for  this  group. 

Larval  characters 

In  the  previous  discussion,  many  doubts  were  raised  con- 
cerning pleuronectiform  interrelationships  as  expressed  in  the 
Regan-Norman  model.  Unfortunately,  larvae  for  many  of  these 
groups  are  unknown.  A  second  problem  is  that  surveys  for  many 
characters  where  larvae  are  known  have  been  incomplete  and 
inconsistent.  Most  descriptive  larval  research  has  dealt  with 
characters  useful  for  identification  and  has  not  involved  com- 
parative work  of  sufficient  detail  to  determine  homologous  states. 
Such  work  is  sorely  needed  before  distnbutions  of  homologous 
states  can  be  determined  for  many  characters. 

Below  is  a  list  and  discussion  of  certain  characters  and  com- 
plexes. Selection  of  these  was  based  mainly  on  the  amount  of 
available  information. 

Preopercidar  spines.  — The  presence  of  preopercular  spines  ap- 
pears to  be  plesiomorphic  for  the  order  and  some  pleuronecti- 
form groups.  This  is  based  on  the  observation  that  the  slate  is 
widespread  among  flatfish  and  percomorph  larvae. 

Neurocranial  spines. —Spines  occur  in  some  regions  of  the  neu- 
rocranium  in  some  pleuronectiform  larvae.  Most  of  these  are 
said  to  occur  in  the  otic  or  frontal  regions.  However,  determining 
homologies  here  is  difficult  due  to  a  general  lack  of  detailed 
osteological  study  of  the  bones  carrying  these  spines.  Spines  in 
the  otic  and  frontal  regions  appear  to  be  of  two  types.  One  of 
these  is  where  spines  are  associated  with  neurocranial  ridge 
systems.  These  are  known  for  larvae  of  achirines  (Houde  et  al., 
1970;  Futch  et  al.,  1972),  some  scophthalmids  (Jones,  1972), 
and  some  pleuronectines  (Pertseva-Ostroumova,  1961).  In  the 
second  type,  spines  occur  singly  or  in  small  groups  but  are  not 
part  of  a  pronounced  ridge.  These  have  been  said  to  occur  on 
various  bones  of  the  otic  region  (epiotics,  autosphenotics,  au- 
topterotics)  or  on  the  frontals.  Tucker  (1982)  was  not  able  to 
determine  the  origin  of  such  spines  in  the  larvae  of  Citharichthys 
and  Etropus  and  referred  to  them  as  frontal-sphenotic  spines. 
Although  thorough  studies  are  needed  before  neurocranial  spines 
can  be  used  to  infer  or  test  pleuronectiform  interrelationships, 
certain  patterns  are  noteworthy:  (1)  Spines  that  are  not  part  of 
some  pronounced  ridge  system  appear  to  be  limited  to  some 
bothoids  (some  species  of  the  Paralichthys  group,  Cyclopsetta 
group,  Pseudorhombus  group.  Scophthalmidae,  Pleuronectinae, 
and  Bothidae).  (2)  Within  the  Bothidae,  only  the  larvae  of  En- 
gyophrys.  Taeniopsetla,  and  Trichopselta  (Taeniopsettinae;  lar- 
vae of  Perissias  are  unknown)  are  known  to  have  otic  spines 
(Amaoka.  1979).  In  these  genera,  the  spines  are  on  the  same 
bones  (epiotics  and  autosphenotics)  and  are  probably  homol- 
ogous. (3)  Within  the  Cyclopsetta  group,  a  relatively  well-de- 


veloped otic  or  frontal  spine  occurs  in  Cyclopsetta  and  Syacium 
(Aboussouan,  1968b;  Gutherz,  1970;  Ahlstrom,  1971;  Futch 
and  HoflT,  1971;  Evseenko.  1979),  while  series  of  small  spines 
occur  in  Citharichlfiys  and  Etropus  (Tucker.  1982). 

Urohyal,  basipterygial.  and  cleithral  spines.  Spines  on  these 
bones  are  limited  to  certain  genera  of  the  Bothidae.  Thus,  they 
are  considered  apomorphic  at  the  pleuronectiform  and  bothoid 
levels  of  universality. 

Early-forming  elongated  dorsal-fin  rays.  —  The  presence  of  elon- 
gated dorsal-fin  rays  in  pleuronectiform  larvae  has  been  exten- 
sively and  justifiably  used  for  identification  purposes.  However, 
use  of  these  structures  for  phylogenetic  interpretations  is  pres- 
ently difficult  and  generally  premature.  There  are  several  reasons 
for  this.  Surveys  for  these  characters  are  inadequate,  since  larvae 
for  many  groups  are  unknown.  Characters  and  character  states 
have  never  been  adequately  defined  to  allow  proper  compari- 
sons to  be  made.  The  only  pattern  here  that  is  clear  and  phy- 
logenetically  interpretable  is  the  state  in  bothids.  All  species  of 
this  family  for  which  larvae  are  known  show  elongation  of  only 
the  second  dorsal-fin  ray.  This  state  is  known  only  in  this  family 
and  thus  appears  to  be  apomorphic  within  the  order  and  both- 
oids. 

Early-forming  elongated  ventral- fin  rari.— Ocular  ventral-fin 
rays  which  are  elongated  relative  to  those  of  the  blind  side  are 
limited  to  certain  species  of  the  Cyclopsetta  group.  Due  to  the 
restricted  occurrence  of  these,  they  are  probably  apomorphic 
for  the  order  and  bothoids.  However,  within  the  Cyclopsetta 
group,  the  distribution  of  elongated  ocular  ventral-fin  rays  does 
not  conform  to  generic  groups  based  on  adult  morphology.  At 
least  one  species  of  cynoglossid  is  known  to  have  elongated  rays 
in  the  ventral  fin  of  the  blind  side  (Kyle,  1913;  Padoa,  1956k). 

Size  at  metamorphosis.  — MosX  flatfishes  metamorphose  in  the 
size  range  of  ca.  10-25  mm.  When  size  at  metamorphosis  has 
been  discussed  in  regard  to  evolution  in  pleuronectiforms,  the 
usual  hypothesis  has  been  that  certain  species  and  groups  have 
evolved  mechanisms  for  prolonging  larval  life  for  greater  dis- 
persal, and  others  have  actually  shortened  larval  life  for  re- 
cruitment to  limited  habitats  (Amaoka,  1979;  Moser,  1981). 
There  are  several  implications  in  this  hypothesis  that  are  rele- 
vant here;  (1)  There  is  some  size  range  for  transformation  that 
is  plesiomorphic  for  the  order.  This  is  usually  implied  to  be  ca. 
10-25  mm  because  most  pleuronectiforms  metamorphose  in 
this  range.  (2)  Metamorphosis  at  markedly  smaller  (e.g.,  Achir- 
inae)  or  larger  (e.g.,  Bothidae,  some  pleuronectines)  sizes  are 
derived  states.  (3)  According  to  the  Regan-Norman  model,  pro- 
longed larval  development  must  have  developed  independently 
in  several  lines.  Although  metamorphosis  at  large  sizes  is  most 
common  in  bothids,  it  is  also  known  for  some  Pleuronectinae, 
the  Poecilopsettinae,  some  species  of  the  Cyclopsetta  group,  and 
some  cynoglossids. 

Size  at  metamorphosis  is  an  important  character  for  larval 
identification,  but  its  use  for  inferring  phylogenetic  relationships 
in  most  instances  is  premature.  Exceptions  may  exist  in  the 
Bothidae,  where  the  extremely  long  premetamorphic  lengths 
exhibited  by  some  genera  are  probably  apomorphic  within  the 
family  and  can  be  used  for  phylogenetic  information. 


686 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Relative  time  of  caudal-fin  formation.— In  most  known  larvae 
of  flatfishes  and  other  teleosts,  formation  of  the  caudal  fin  pre- 
cedes or  occurs  with  that  of  the  dorsal  and  anal  fins.  The  only 
exceptions  known  in  pleuronectiforms  are  the  cynoglossids.  In 
this  family,  the  caudal  fin  does  not  develop  until  the  dorsal  and 
anal  fins  are  nearly  completely  developed.  This  pattern  of  de- 
velopment is  considered  apomorphic  in  pleuronectiforms. 

Eye  migration  and  dorsal-fin  position  at  tnetamorphosis.  —  Eye 
migration  has  been  observed  in  some  flatfish  groups.  In  the 
Psettodidae,  Pleuronectinae,  Paralichthyidae  (excluding  the  Cy- 
clopsetta  group),  Scophthalmidae,  and  apparently  some  Sole- 
idae,  the  first  ray  of  the  dorsal  fin  is  above  or  posterior  to  the 
eyes.  At  metamorphosis,  the  migrating  eye  crosses  anterior  to 
the  dorsal-fin  origin.  These  types  of  eye  migration  and  dorsal- 
fin  position  appear  to  be  plesiomorphic  for  the  order.  Several 
derived  states  for  these  characters  occur.  In  at  least  one  species 
of  cynoglossid,  a  fleshy  rostral  beak  is  formed  anterior  to  the 
dorsal-fin  origin.  Eye  migration  takes  place  between  the  rostral 
beak  and  the  interorbital  region.  In  some  soleids,  the  dorsal-fin 
origin  projects  above  the  snout  and  the  eye  migrates  between 
this  projection  and  the  neurocranium.  In  the  Bothidae,  the  dor- 
sal fin  is  anterior  to  the  eye  and  attached  to  the  ethmoid  region. 
During  migration,  the  eye  goes  between  the  base  of  the  dorsal 
fin  and  the  ethmoid  region.  A  path  for  the  migrating  eye  is 
created  by  detachment  of  the  anterior  section  of  the  dorsal  fin 
from  the  ethmoid  region  so  that  a  narrow  slit  is  formed,  or  some 
tissue  in  the  path  of  the  migrating  eye  is  absorbed.  A  very  similar 
type  of  eye  migration  occurs  in  some  species  of  the  Cyclopsetta 
group.  However,  in  other  members  of  this  group,  the  eye  mi- 
grates around  the  dorsal-fin  origin  (Gutherz,  1970;  Tucker,  1982). 

Phylogenetic  information  provided  by 
larval  characters 

Although  larvae  of  some  critical  groups  are  unknown  or  poor- 
ly known,  some  comments  about  phylogenetic  relationships  can 
be  made  in  regard  to  groups  where  our  knowledge  is  on  a  higher 
level. 

Bothoids. —Spines  in  the  otic  or  frontal  regions  of  the  neuro- 
cranium which  are  isolated  or  in  small  clusters  appear  to  be 
limited  to  various  groups  of  bothoids.  If  these  spines  prove  to 
be  homologous  between  these  groups,  they  may  be  apomorphic 
within  the  order.  In  this  interpretation,  they  would  be  primitive 
for  bothoids  and  lost  in  various  lines. 

Paralichthyidae.— As  discussed  in  the  section  on  adult  charac- 
ters, this  family  as  currently  interpreted  is  polyphyletic  due  to 
the  inclusion  of  Tephrinectes  and  Thysanopsetta.  We  do  not 
consider  these  genera  bothoids  as  defined  by  the  caudal-fin  com- 
plex. Their  larvae  are  unknown. 

We  have  interpreted  the  Cyclopsetta  group  as  monophyletic 
based  on  some  adult  character  states  which  are  probably  apo- 
morphic. Although  larvae  of  this  group  show  certain  states  which 
appear  to  be  apomorphic  within  bothoids  (e.g.,  elongated  left 
ventral-fin  rays),  not  all  species  in  this  group  show  these. 

The  Pseudorhombus  group  is  possibly  definable  by  adult  syn- 
apomorphies.  In  larvae  of  this  group,  we  see  no  character  states 
that  are  presently  interpretable  with  certainty  as  synapomor- 
phies. 

In  examining  adult  characters  of  the  Paralichthys  group,  it 
appeared  likely  that  this  group  had  no  synapomorphies.  Larvae 


tend  to  support  this.  They  show  the  following  character  states 
which  appear  to  be  plesiomorphic  for  the  order:  (1)  presence  of 
preopercular  spines;  (2)  origin  of  the  dorsal  fin  behind  the  eyes; 
(3)  metamorphosis  in  a  size  range  of  7.5-14.2  mm;  and  (4)  eye 
migration  anterior  to  the  dorsal  fin.  In  addition,  at  least  some 
species  show  the  following  states  which  may  prove  to  be  ple- 
siomorphic at  least  within  the  bothoids:  (1)  four  or  five  elon- 
gated, early-forming  dorsal-fin  rays;  and  (2)  presence  of  otic 
spines. 

Bothidae.  —  'With  the  exclusion  of  Mancopsetta  and  inclusion  of 
Perissias,  this  family  is  definable  by  adult  synapomorphies.  Lar- 
vae of  the  Bothidae  are  probably  better  known  than  for  any 
other  family  of  flatfishes.  However,  larvae  of  many  genera  are 
still  unknown  (i.e.,  Parabothus,  Asterorhombus,  Tosarhombus. 
Neolaeops.  Japonolaeops,  and  Perissias).  Amaoka  (1979)  re- 
viewed larval  characters  of  most  genera  for  which  larvae  are 
known.  Known  bothid  larvae  show  the  following  character  states 
which  are  interpreted  as  synapomorphies:  (1)  metamorphosis 
at  a  relatively  large  size  (ca.  15-120  mm);  (2)  eye  migration 
below  the  dorsal  fin;  (3)  dorsal-fin  origin  anterior  to  eyes  just 
prior  to  metamorphosis;  (4)  elongated,  early-forming  second 
dorsal-fin  ray;  and  (5)  lack  of  preopercular  spines. 

Larvae  of  some  bothid  genera  have  various  combinations  of 
otic-region,  urohyal,  cleithral,  and  basipterygial  spines.  It  is 
tempting  to  use  the  presence  of  these  spines  to  define  bothid 
groups,  and  therefore,  assume  that  they  are  apomorphic  within 
the  family.  Spines  in  the  otic  region  within  the  Bothidae  are 
limited  to  the  Taeniopsettinae  as  presently  defined.  However, 
spines  in  this  region  occur  in  other  bothoid  groups.  Although 
sufficient  comparative  osteological  work  has  not  been  done  to 
show  that  these  spines  are  homologous  between  taeniopsettines 
and  other  bothoids,  use  of  these  spines  to  infer  close  relation- 
ships between  Engyophrys.  Taeniopsetta.  and  Trichopsetta  is 
questionable.  Urohyal,  cleithral,  and  basipterygial  spines  are 
known  only  from  larvae  of  nine  bothid  genera.  They  occur  in 
various  combinations  inter-  and  intragenerically.  Amaoka  (1969) 
presented  a  model  of  intergeneric  relationships  for  Japanese 
bothids  based  on  adult  characters.  Occurrence  of  these  larval 
spines  is  scattered  among  the  bothid  lines  hypothesized  by 
Amaoka.  This  could  indicate  two  possibilities:  ( 1 )  the  spines  are 
apomorphic  within  the  family,  and  Amaoka's  model  is  incor- 
rect; or  (2)  Amaoka's  model  is  correct  and  the  spines  are  ple- 
siomorphic within  the  family  and  have  been  lost  in  several  lines. 
Two  major  problems  exist  with  Amaoka's  phytogeny  based  on 
adult  characters;  it  was  constructed  using  eclectic  methods  and 
it  did  not  include  all  genera.  Interpretation  of  urohyal,  basip- 
terygial, and  cleithral  spines  should  await  a  cladistic  analysis  of 
bothid  interrelationships  based  on  adult  characters. 

Pleuronectidae.  — Based  on  adult  characters,  we  interpret  this 
family  as  polyphyletic.  Larvae  of  the  four  non-bothoid  subfam- 
ilies are  poorly  known,  and  hence,  of  little  aid  in  determining 
relationships  of  these  groups.  However,  there  are  certain  simi- 
larities in  general  body  morphology  between  the  few  known 
samarine  and  poecilopsettine  larvae.  In  regard  to  the  Pleuro- 
nectinae, many  adult  states  that  are  shared  are  plesiomorphic 
for  pleuronectiforms  or  bothoids.  This  also  appears  to  be  true 
for  most  larval  characters.  The  position  of  the  dorsal-fin  origin 
(posterior  to  the  eyes)  and  the  type  of  eye  migration  (anterior 
to  the  dorsal-fin  origin)  are  plesiomorphic  for  the  order.  Some 
pleuronectine  larvae  have  preopercular  spines,  which  again,  are 


HENSLEY  AND  AHLSTROM:  PLEURONECTIFORMES 


687 


probably  plesiomorphic  for  flatfishes.  Some  genera  show  spines 
in  the  otic  region  of  the  neurocranium;  these  are  possibly  ple- 
siomorphic for  bothoids.  All  known  pleuronectine  larvae  lack 
elongated  dorsal-fin  rays.  However,  this  state  is  not  limited  to 
this  group  and  a  phylogenetic  interpretation  of  it  would  be  pre- 
mature. In  short,  at  present,  we  know  of  no  character  states  that 
are  unique  to  pleuronectine  larvae  or  that  can  confidently  be 
interpreted  as  apomorphic. 

Egg  characters 

Except  in  certain  groups,  eggs  of  flatfishes  are  still  too  poorly 
known  to  be  of  much  value  in  phylogenetic  studies.  One  char- 
acter of  pleuronectiform  eggs  was  used  by  Regan  (1910)  and 
Norman  (1934)  to  interpret  phylogeny,  the  presence  of  one  oil 
globule  in  bothid  eggs  to  separate  them  from  those  of  pleuro- 
nectids  which  lack  oil  globules.  We  now  have  more  information 
about  the  occurrence  of  oil  globules  in  flatfish  eggs,  and  the 
distribution  of  these  character  states  is  not  exactly  that  predicted 
by  the  Regan-Norman  model  (preceding  article,  this  volume). 
The  obvious  pattern  here  is  that  bothoids  have  0-1  and  soleoids, 
rhombosoleines,  and  Mancopsetta  multiple  oil  globules.  There 
are  published  exceptions  to  this.  Watson  and  Leis  (1974)  iden- 
tified three  types  of  eggs  with  multiple  oil  globules  as  those  of 


bothids.  However,  these  authors  expressed  some  doubt  about 
the  identifications  of  at  least  two  of  these  egg  types.  These  eggs 
are  probably  some  other  group  (poecilopsettines  or  samarines?). 
Brownell  (1979)  identified  some  eggs  which  lacked  oil  globules 
as  the  soleid  Heteromycteris  capensis.  This  is  the  only  soleid  we 
are  aware  of  that  lacks  multiple  oil  globules. 

It  is  probably  premature  to  use  the  oil-globule  character  for 
phylogenetic  information  until  eggs  from  other  groups  are  known. 
However,  it  is  interesting  and  possibly  significant  that  the  so- 
leoids, rhombosoleines,  and  Mancopsetta  are  so  sharply  sepa- 
rable from  the  bothoids  in  this  character.  One  oil  globule  appears 
to  be  the  most  common  state  in  the  eggs  of  percomorph  fishes 
(based  on  accounts  in  Watson  and  Leis,  1974;  Russell,  1976; 
Fritzsche,  1978;  Hardy,  1978b;  Johnson,  1978;  and  Brownell, 
1979).  This  may  indicate  that  this  state  is  plesiomorphic  for 
pleuronectiforms.  Corollaries  of  this  would  be  that  oil  globules 
were  lost  in  most  pleuronectines,  and  multiple  oil  globules  de- 
veloped in  a  line  leading  to  the  soleoids,  rhombosoleines,  and 
Mancopsetta. 

(D.A.H.)  Department  of  Marine  Sciences,  University  of 
Puerto  Rico,  Mayaguez,  Puerto  Rico  00708. 


LITERATURE  CITED 


Abe,  T.    1954.    New,  rare  or  uncommon  fishes  from  Japanese  waters. 

IV.  Records  of  rare  fishes  of  Ihe  families  Lophotidae,  Nomeidae, 

Icosteidae.  Jpn.  J.  Ichthyol.  3:90-95. 
.    1957.    Notes  on  fishes  from  the  stomachs  of  whales  taken  in 

the  Antarctic.  Sci.  Rep.  Whales  Res.  Inst.  (Tokyo)  No.  12:225-233. 
,  andS.  Kaji.    1972.   A.  record  of  Oreoso?na  atlanlicuni(OTeoso- 


matidae,  Zeiformes,  Teleostei)  from  Tasman  Sea.  U.O.  No.  8:5-7. 
Able,  K.  W.    1984.   Variation  m  spawning  site  preference  of  the  mum- 

michog,  Fundulus  heleroclilus.  Copeia  1984:522-525. 
.   MS.   Comments  on  the  distribution  and  reproductive  season- 
ality of  snailfishes  and  lumpfishes  in  the  estuary  and  Gulf  of  St. 

Lawrence. 
,  J.  D.  Felley  and  D.  Powers.    MS.    Taxonomy  of  Bermuda 

Fundulus  (Pisces:  Fundulidae). 
,  D.  F.  Markle  AND  M.  P.  Fahay.    MS.    Development  of  larval 

liparidine  snailfishes  from  the  western  North  Atlantic. 
,  and  D.  E.  McAllister.    1980.    Revision  of  the  snailfish  genus 

Liparis  from  Arctic  Canada.  Can.  Bull.  Fish.  Aquat.  Sci.  No.  208. 
,  and  J.  A.  Mlisick.    1976.    Life  history,  ecology  and  behavior 


oi  Lipans  inquilmus  (Pisces:  Cycloptendae)  associated  with  the  sea 

scallop,  Placopecten  magetlanicus.  Fish.  Bull.  U.S.  74:409-421. 
Aboussouan,  A.    1964.   Contribution  a  I'etude  des  oeufs  et  larves  pe- 

lagiques  des  poissons  teleosteens  dans  la  Golfe  de  Marseille.  Rec. 

Trav.  Stn.  Mar.  Endoume  Fac.  Sci.  Mars.  32(48):87-l  73. 
.     1965.    Oeufs  et  larves  de  teleosteens  de  I'ouest  africain.   1. 

Acanthurus  monroviae  Sleind.  Bull.  Inst.  Fondam.  Afr.  Noire,  Ser. 

A  Sci.  Nat.  27:1183-1187. 
.    1966a.   Sur  quelques  larves  pelagiques  de  teleosteens  recoltees 

a  Tulear.  Reel.  Trav.  Stn.  Mar.  Endoume-Mars.  Suppl.  5:149-164. 
.    1966b.   Sur  quelques  larves  pelagiques  de  teleosteens  recoltees 

a  Tulear.  Cent.  Oceanogr.  Fac.  Sci.  Mars.  1966:147-159. 
.    1966c.    Oeufs  et  larves  de  teleosteens  de  I'ouest  africain.  III. 

Larves  de  Monacanthus  hispidus  (L.)  et  de  Balistes  forcipalus  Gm. 

Bull.  Inst.  Fondam.  Afr.  Noire,  Ser.  A  28:276-282. 
.    1966d.    Oeufs  et  larves  de  teleosteens  de  fouest  africain.  IV. 

Galeoides  polydactylus  (Vahl)  [Polynemidae].  Bull.  Inst.  Fondam. 

Afr.  Noire,  Ser.  A  Sci.  Nat.  28:1037-1040. 
.     1967.    Oeufs  et  larves  de  teleosteens  de  I'ouest  africain.  V. 

Caranx  rhonchus  GeofTroy  Saint-Hilaire  (Carangidae).  Affinites  avec 

Trachurus  trecae  Cadenat.  Bull.  Inst.  Fondam.  Afr.  Noire,  Ser.  A 

Sci.  Nat.  29:1039-1050. 
.    1968a.    Oeufs  et  larves  de  teleosteens  de  I'ouest  africain.  VI. 

Larves  de  Chloroscombrus  chrysurus  (Linne)  et  de  Blepharius  crin- 

itis  (Mitchill)  (Carangidae).  Bull.  Inst.  Fondam.  Afr.  Noire,  Ser.  A 

Sci.  Nat.  30:226-237. 
.    1968b.   Oeufs  et  larves  de  teleosteens  de  I'ouest  africain  VII. 


Larves  de  Syacium  giiineensis  (Blkr.)  [Bothidae].  Bull.  Inst.  Fon- 
dam. Afr.  Noire,  Ser.  A  Sci.  Nat.  30:1  188-1 197. 

— .  1968c.  Oeufs  et  larves  de  teleosteens  de  I'ouest  africain.  VIll. 
Larves  de  Bregmaceros  macclellandi  Thompson  (Bregmaceroti- 
dae).  Bull.  Inst.  Fondam.  Afr.  Noire,  Ser.  A  Sci.  Nat.  30:1590- 
1602. 

— .  1 969.  Sur  une  petite  collection  de  larves  de  teleosteens  recoltee 
au  large  du  Bresil  (Campagne  "Calypso"  1962).  Vie  Milieu,  Ser.  A 
Biol.  Mar.  20:595-610. 

— .  1972a.  Oeufs  et  larves  de  teleosteens  de  I'ouest  africain.  X. 
Larves  d'Ophidioidei  (Oligopus.  Ophidian  et  Carapus)  et  de  Per- 
coidei  (Pseudupeneus).  Bull.  Inst.  Fondam.  Afr.  Noire,  Ser.  A  Sci. 
Nat.  34:169-178. 

— .  1972b.  Oeufs  et  larves  de  teleosteens  de  I'ouest  africain.  XI. 
Larves  serraniformes.  Bull.  Inst.  Fondam.  Afr.  Noire,  Ser.  A  Sci. 
Nat.  34:485-502. 

— .  1972c.  Oeufs  et  larves  de  teleosteens  de  I'ouest  africain.  Xll. 
Les  larves  d'Heterosomata  recoltees  aux  environs  de  I'ile  de  Goree 
(Senegal).  Bull.  Inst.  Fondam.  Afr.  Noire,  Ser.  A  Sci.  Nat.  34:974- 
1003. 

.    1975.    Oeufs  et  larves  de  teleosteens  de  I'ouest  africain.  XIII. 


Contribution  a  I'identification  des  larves  de  Carangidae.  Bull.  Inst. 

Fondam.  Afr.  Noire,  Ser.  A  Sci.  Nat.  37:899-938. 
.    1980.   Description  d"unelarvegeante"rubaniforme,"attribuee 

au  genre  Brotulataenia  (Pisces,  Gadiformes,  Ophidioidei,  Ophi- 

diidae)  et  recoltee  a  Test  des  nouvelles  Hebrides.  Cybium,  3rd  Ser., 

(IO):51-64. 
Adams,  J.  A.    1960.    A  contribution  to  the  biology  and  postlarval  de- 
velopment of  the  sargassum  fish,  Histno  hisino  (Linnaeus),  with  a 

discussion  of  the  Sargassum  complex.  Bull.  Mar.  Sci.  Gulf  Caribb. 

10:55-82. 
Agassiz,  A.    1882.    On  the  young  stages  of  some  osseous  fishes.  Proc. 

Am.  Acad.  Arts  Sci.  16(Pt.  3):27 1-303. 
,  AND  C.  O.  Whitman.     1885.    The  development  of  osseous 

fishes.  I.  The  pelagic  stages  of  young  fishes.  Mem.  Mus.  Comp. 

Zool.  Harv.  14(1):  1-56. 
Ahlstrom,  E.  H.     1943.    Studies  on  the  Pacific  pilchard  or  sardine 

(Sardinops  caerulea).  4.  Influence  of  temperature  on  the  rate  of 

development  of  pilchard  eggs  in  nature.  U.S.  Fish  Wildl.  Serv.  SSRF 

23. 
.    1956.   Eggs  and  larvae  of  anchovy,  jack  mackerel,  and  Pacific 

mackerel,  p.  33-42.  In:  Progress  Rep.,  Calif  Coop.  Oceanic  Fish. 

Invest.,  1  April  1955-30  June  1956,  Sacramento. 
.    1965.  Kinds  and  abundance  of  fishes  in  the  California  Current 

region  based  on  egg  and  larval  surveys.  Calif  Coop.  Oceanic  Fish. 

Invest.  Rep.  10:31-52. 
.    1968.    Review  of  "Development  of  fishes  of  the  Chesapeake 

Bay  region,  an  atlas  of  egg,  larval  and  juvenile  stages.  Part   1". 

Copeia  1968:648-651. 
.    1969.   Remarkable  movements  of  oil  globules  in  eggs  of  bathy- 

lagid  smelts  during  embryonic  development.  J.  Mar.  Biol.  Assoc. 

India  11:206-217. 
.    1971.   Kinds  and  abundance  of  fish  larvae  in  the  eastern  trop- 
ical Pacific,  based  on  collections  made  on  EASTROPAC  1.  Fish. 

Bull.  U.S.  69:3-77. 
.     1972a.    Kinds  and  abundance  of  fish  larvae  in  the  eastern 


tropical  Pacific  on  the  second  multivessel  EASTROPAC  survey, 
and  observations  on  the  annual  cycle  of  larval  abundance.  Fish. 
Bull.  U.S.  70:1153-1242. 

.     1972b.    Distributional  atlas  of  fish  larvae  in  the  California 

Current  region:  six  common  mesopelagic  fishes.  Calif  Coop.  Oceanic 
Fish.  Invest.  Atlas  17. 

— .  1974.  The  diverse  patterns  of  metamorphosis  in  gonostomatid 
fishes  -  an  aid  to  classification,  p.  659-674.  In:  The  early  life  history 
offish.  J.  H.  S.  Blaxter  (ed.).  Springer-Verlag,  Berlin. 

— .  1 976.  Maintenance  of  quality  in  fish  eggs  and  larvae  collected 
during  plankton  hauls,  p.  313-318.  //;.  Zooplankton  fixation  and 
preservation.  H.  F.  Steedman  (ed).  LInesco  Press,  Paris. 

— ,  and  O.  P.  Ball.  1954.  Description  of  eggs  and  larvae  of  jack 
mackerel  ( Trachurus  symmelricus)  and  distribution  and  abundance 
of  larvae  in  1950  and  1951.  Fish.  Bull.  U.S.  56:209-245. 

— ,  J.  L.  Butler  and  B.  Y.  Sumida.  1976.  Pelagic  stromateoid 
fishes  (Pisces,  Perciformes)  of  the  eastern  Pacific:  kinds,  distribu- 
tions and  early  life  histories  and  observations  on  five  of  these  from 
the  northwest  Atlantic.  Bull.  Mar.  Sci.  26:285-402. 

— ,  AND  R.  C.  Counts.  1955.  Eggs  and  larvae  of  the  Pacific  hake 
Merlucaus  produclus.  Fish.  Bull.  U.S.  56:295-329. 

— ,  and .    1958.    Development  and  distribution  of  iinci- 

guerria  lucetia  and  related  species  in  the  eastern  Pacific.  U.S.  Fish 
Wildl.  Serv.  Fish.  Bull.  58:363-416. 

— ,  AND  H.  G.  Moser.    1969.    A  new  gonostomatid  fish  from  the 


tropical  eastern  Pacific.  Copeia  1969:493-500. 
— ,  AND .    1975.    Distributional  atlas  offish  larvae  in  the 

California  Current  region:  Flatfishes,   1955  through   1960.  Calif 

Coop.  Oceanic  Fish.  Invest.  Atlas  23. 
— ,  AND .    1976.    Eggs  and  larvae  of  fishes  and  their  role  in 

systematic  investigations  and  in  fisheries.  Rev.  Trav.  Inst.  Peches 

Mant.  40:379-398. 


688 


LITERATURE  CITED 


689 


.AND .    1980.  Characters  useful  in  identification  of  pelagic 

manne  fish  eggs.  Calif.  Coop.  Oceanic  Fish.  Invest.  Rep.  21:121- 
131. 

,  AND .    1981.    Systematics  and  development  of  early  life 

history  stages  of  marine  fishes:  achievements  during  the  past  cen- 
tury, present  status  and  suggestions  for  the  future.  Rapp.  P.  V.  Reun. 
Cons.  Int.  Explor.  Mer  178:541-546. 

— , AND  M.  J.  O'TooLE.  1976.  Development  and  distri- 
bution of  larvae  and  early  juveniles  of  the  commercial  lantemfish, 
Lampanyctodes  hecloris  (Gijnther).  off  the  west  coast  of  southern 
Africa  with  a  discussion  of  phylogenetic  relationships  of  the  genus. 
Bull.  South.  Calif  Acad.  Sci.  75:138-152. 

— ,  AND  E.  G.  Stevens.    1977.    Report  of  neuslon  (surface)  collec- 
tions made  on  an  extended  CALCOFI  cruise  dunng  May  1972. 
Calif  Coop.  Oceanic  Fish.  Invest.  Rep.  18:167-180. 
-,  AND  B.  Y.  Slimida.    MS.    Early  life  history  of  eastern  Pacific 


carangids.  NMFS,  U  Jolla,  CA  92038. 

Akatsu.  S.,  Y.  OciASAWARA  andF.  Yasuda.  1977.  Spawning  behavior 
and  development  of  eggs  and  larvae  of  the  striped  fingerfish,  Mono- 
daaylus  sebac.  Jpn.  J.  Ichthyol.  23:208-214. 

Akazaki.  M..  H.  Nakajima,  H.  Kawahara  and  S.  Takamatsii.  1 976. 
Embryonic  development  and  metamorphosis  after  hatching  in  the 
sharksucker.  Echeneis  naucrates.  Jpn.  J.  Ichthyol.  23:153-159.  (In 
Japanese,  English  abstract.). 

Akihito,  p.  1963.  On  the  scapula  of  gobiid  fishes.  Jpn.  J.  Ichthyol. 
11:1-26. 

.    1967.   Additional  research  on  the  scapula  of  gobiid  fishes.  Jpn. 

J.  Ichthyol.  14:167-182. 

.    1969.   A  systematic  examination  of  the  gobiid  fishes  based  on 

the  mesoptcrygoid,  postcleithra,  branchiostegals.  pelvic  fins,  scap- 
ula, and  suborbital.  Jpn.  J.  Ichthyol.  16:93-1  14. 

,  ANI5  K.  Mecjuro.    1981.    A  gobiid  fish  belonging  to  the  genus 

Hetereteolris  coWecleA  in  Japan.  Jpn.  J.  Ichthyol.  28:329-339. 

Alcock,  a.  1889.  Natural  history  notes  from  H.  M.  Indian  Manne 
Survey  Steamer 'Investigator,' Commander  Alfred  Carpenter.  R.N., 
D.S.O.,  Commanding. —  No.  13.  On  the  bathybial  fishes  of  the  Bay 
of  Bengal  and  neighbouring  waters,  obtained  dunng  the  seasons 
1885-1889.  Ann.  Mag.  Nat.  Hist.,  Ser.  6.  4:450-461. 

Alderdice,  D.  F..  AND  C.  R.  Forrester.  1971.  Effects  of  salinity  and 
temperature  on  embryonic  development  of  petrale  sole  (Eopsetia 
jordani).  J.  Fish.  Res.  Board  Can.  28:727-744. 

Alexander,  E.  C  1961.  A  contribution  to  the  life  history,  biology 
and  geographical  distribution  of  the  bonefish,  Alhtilii  vulpes  (Lin- 
naeus). Dana-Rep.  Carlsberg  Found.  53. 

Allen,  G.  H.  1968.  Fecundity  of  the  brown  ragfish,  Icosleus  aenig- 
maticus  Lockington,  from  northern  California.  Calif  Fish  Game 
54:207-214. 

Allen,  G.  R.  1972.  The  anemonefishes.  Their  classification  and  bi- 
ology. T.F.H.  Publications,  Inc..  Neptune  City,  NJ. 

.     1975a.    Damselfishes  of  the  south  seas.  T.F.H.  Publ..  Inc., 

Neptune  City,  NJ. 

.    1975b.   The  biology  and  taxonomy  of  the  cardinalfish  5p/iae- 

ramia  orbicularis  (Piices:  Apogomdae).  J.  R.  Soc.  West.  Aust.  58(3): 
86-92. 

.     1980.    A  genenc  classification  of  the  rainbowfishes  (family 

Melanotaeniidae).  Rec.  West.  Aust.  Mus.  8:449-490. 

.     1981.    Polynemidae.  In:  FAO  species  identification  sheets. 

Eastern  Central  .Atlantic,  Fishing  Areas  34,  37.  W.  Fischer  and  W. 
B.  Scott  (eds.).  FAO,  Rome. 

,  and  H.  J.  Cross.    1980.   Descriptions  of  five  new  rainbowfishes 

(Melanotaeniidae)  from  New  Guinea.  Rec.  West.  Aust.  Mus.  8: 
377-396. 

,  and  R.  H.  KtUTFR.    1976.   A  review  of  the  plesiopid  fish  genus 


Assesor.  with  descriptions  of  two  new  species.  Rec.  West.  Aust. 
Mus.  4(3):20l-215. 
Allen.J.  M.,  J.  H.S.  Blaxter  andE.  J.  Denton.  1976.  The  functional 
anatomy  and  development  of  the  swim  bladder— inner  ear— lateral 
line  system  in  herring  and  sprat.  J.  Mar.  Biol.  Assoc.  U.K.  56:471- 
486. 


Allen,  L.  G.  1979.  Lar\a\  deve\opmeM  of  Gohiesox  rhessodon  (Go- 
biesocidae)  with  notes  on  the  larva  of  Rimicola  muscarum.  Fish. 
Bull.  U.S.  77:300-304. 

.  and  J.  M.  Ilg.     1983.    Larval  development  of  the  northern 

clingflsh,  Gobiesox  maeandricus.  Copeia  1983:551-554. 

Altukhov,  K,  A.  1978.  On  the  reproduction  and  abundance  of  the 
lesser  sand  lance,  Ammodyles  marinus.  in  the  White  Sea.  J.  Ich- 
thyol. (Engl.  Transl.  Vopr.  Ikhtiol.)  18:560-567. 

.    1979.    A  contribution  to  the  biology  of  non-commercial  and 

rare  fish  species  from  the  White  Sea  in  the  early  stages  of  devel- 
opment. J.  Ichthyol.  (Engl.  Transl.  Vopr.  Ikhtiol.)  19(6):105-1  15. 

Alvarez-Lajonchere,  L.  L.  1975.  Estudio  sistematico  de  Afugil  hra- 
siliensis.  Mugil gaimardianus,  y  Mugil curema.  Cienc.  Ser.  8  Invest. 
Mar.  (Havana)  No.  14:1-18. 

Amaoka,  K.  1964.  Development  and  growth  of  the  sinistral  flounder, 
Bothus  myriaster  (TemmmcV.  and  Schlegel)  found  in  the  Indian  and 
Pacific  Oceans.  Bull.  Misaki  Mar.  Biol.  Inst.  Kyoto  LIniv.  5:1 1-29. 

.    1969.    Studies  on  the  sinistral  flounders  found  in  the  waters 

around  Japan.  Taxonomy,  anatomy,  and  phylogeny.  J.  Shimono- 
seki  Univ.  Fish.  18:65-340. 

.  1970.  Studies  on  the  larvae  and  juveniles  of  the  sinistral  floun- 
ders. I.  Taeniopseua  ocellata  (Gunther).  Jpn.  J.  Ichthyol.  17:95- 
104. 

.  1971.  Studies  on  the  larvae  and  juveniles  of  the  sinistral  floun- 
ders. II.  Chascanopsella  lugiibris.  Jpn.  J.  Ichthyol.  18:25-32. 

.     1972a.    Studies  on  the  larvae  and  juveniles  of  the  sinistral 

flounders.  III.  Laeops  kitaharae.  Jpn.  J.  Ichthyol.  19:154-165. 

.    1972b.    Osteology  and  relationships  of  the  citharid  flatfish 

Brachypleura  novaezeetandiae.  Jpn.  J.  Ichthyol.  19:263-273. 

.  1973.  Studies  on  the  larvae  and  juveniles  of  the  sinistral  floun- 
ders. IV.  Arnoglossus  japonicus.  Jpn.  J.  Ichthyol.  20:145-156. 

.  1974.  Studies  on  the  larvae  and  juveniles  of  the  sinistral  floun- 
ders. V.  Arnoglossus  tenuis.  Jpn.  J.  Ichthyol.  21:153-157. 

.    1976.   Studies  on  the  larvae  and  juveniles  of  the  sinistral  floun- 


ders. VI.  Psetlma  iijimae.  P  tosana.  and  P  gigantea.  Jpn.  J.  Ich- 
thyol. 22:201-206. 

.    1979.    Phylogeny  and  larval  morphology  of  pleuronectiform 

fishes.  (Psettodidae.  Citharidae,  Paralichthyidae  and  Bothidae). 
Kaiyo  Kagaku  1 1(2):I00-I  10.  (English  transl.,  1981.  Edited  by  J. 
R.  Dunn,  Translated  by  A.  Shiga.  Natl.  Mar.  Fish.  Serv.,  Northwest 
and  Alaska  Fish.  Center,  Seattle). 

,  M.  ToKOSHiMA  and  T.  Inada.     1977.    New  records  of  the 


stichaeid  fish  Ascoldia  variegata  knipowitschi  and  the  zoarcid  fish 
Puzanovia  rubra  from  Japan.  Jpn.  J.  Ichthyol.  24:91-97. 

Ambrose.  D.  A.,  J.  L.  Butler,  H.  G.  Moser,  B.  Y.  Sumida,  E.  M. 
Sandknop  and  E.  G.  Stevens.  1983.  Description  of  the  larvae 
of  the  cusk-eels  Ophidian  scrippsae  and  Chilara  taylori  (Ophidi- 
idae).  Calif  Coop.  Oceanic  Fish.  Invest.  Rep.  24:226-234. 

Andersen,  N.  C  1984.  Genera  and  subfamilies  of  the  family  No- 
totheniidae  (Pisces,  Perciformes)  from  the  antarctic  and  subant- 
arctic.  Steenstrupia  10:1-34. 

,  AND  J.  C  HiREAU.  1979.  Proposition  pour  une  nouvelle  clas- 
sification des  Nototheniinae  (Pisces,  Perciformes,  Nototheniidae). 
Cybium,  3rd  Ser.,  3(6):47-53. 

Anderson,  M.  E.  1981.  Aspects  of  the  natural  history  of  the  midwater 
fish  Lycodapus  mandibulans  (Zoarcidae)  in  Monterey  Ba\.  Cali- 
fornia. Pac.  Sci.  34:181-194. 

.     1982.    Revision  of  the  fish  genera  Gymnetus  Reinhardt  and 

Gymnetopsis  Soldatov  (Zoarcidae),  with  two  new  species  and  com- 
parative osteology  of  Gymnclus  vindis.  Natl.  Mus.  Nat.  Sci.  (Ot- 
tawa) Publ.  Zool.  17:1-76. 

.    1984.   On  the  anatomy  and  phylogeny  of  Zoarcidae  (Teleostei: 

Perciformes).  LInpubl.  PhD  dissertation.  Coll.  William  and  Mary, 
Williamsburg,  VA. 

,  andG.  M.  Cauliet.    1974.   Crab  and  snailfish  commensalism 

in  Monterey  Bay.  Underwater  Nat.  8(3):29-31. 

,  AND  C  L.  Hi  BBS.    1981.    Redescription  and  osteology  of  the 

northeastern  Pacific  fish  Derepodichlhys  alepidotus  (Zoarcidae).  Co- 
peia 1981:341-352. 


690 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Anderson,  W.  D.,  Jr..  and  P.  C.  Heemstra.    1980.    Two  new  species 

of  western  Atlantic  Anlhias  (Pisces:  Serranidae),  redescription  of 

A.  asperilinguis  and  review  of  Holanthias  martinicensis.  Copeia 

1980:72-87. 
Anderson,  W.  W.    1957.    Early  development,  spawning,  growth,  and 

occurrence  of  the  silver  mullet  (Mugil  curema)  along  the  south 

Atlantic  coast  of  the  United  States.  Fish.  Bull.  U.S.  57:397-414. 
,  J.  W.  Gehringer  and  F.  H.  Berry.    1966.    Family  Synodon- 

tidae.  p.  30-102.  In:  Fishes  of  the  western  North  Atlantic.  G.  W. 

Mead  (ed.).  Mem.  Sears  Found.  Mar.  Res.  No.  1  (Pt.  5). 
Anderton,  T.    1907.   Observations  on  New  Zealand  fishes,  etc.,  made 

at  the  Portobello  marine  fish-hatchery.  Trans.  NZ  Inst.  39:477- 

496. 
Andrew,  W.    1959.  Textbook  of  comparative  histology.  Oxford  Univ. 

Press,  New  York. 
Andriashev,  a.  p.    1938.   Review  of  the  genus  Krusensterniella  Schmidt 

(Pisces,  Zoarcidae),  with  description  of  a  new  species  from  the  Japan 

Sea.  Bull.  Far  East.  Br.  Acad.  Sci.  USSR  32:1 17-121. 
.     1952.    A  new  fish  of  the  eelpout  family  (Pisces.  Zoarcidae) 

from  the  Benng  Sea.  Zool.  Inst.  Acad.  Nauk  SSSR  12:415-417. 
.    1954.   Fishes  of  the  northern  seas  of  the  USSR.  Tr.  Zool.  Inst. 

Akad.  Nauk  SSSR,  No.  53.  (In  Russian,  translated  by  Isr.  Program 

for  Sci.  Transl.,  Jerusalem,  1 964.  Available  from  NTIS,  Springfield. 

VA;  OTS  63-11 160. 
.   1955a.  A  fish  newto  thefaunaoftheUSSR— fn/ep/iliin/epM 

zonifer (LocV.)  Pisces,  Anoplopomidae]  from  the  Kamchatkan  waters 

of  the  Pacific  Ocean.  Vopr.  Ikhtiol.  4:3-9. 
.    1955b.   Observations  on  the  eel-like  likods  (Lr(v«i:-/)Wv5  Gill 

(Pisces,  Zoarcidae)  and  related  forms)  in  the  seas  of  the  USSR  and 

adjacent  waters.  Tr.  Zool.  Inst.  Acad.  Nauk  SSSR  18:349-384. 
.    1959.    On  the  number  of  vertebrae  and  certain  other  osteo- 

logical  characters  of  Antarctic  fishes.  Vopr.  Ikhtiol.   12:3-7.  (In 

Russian.) 
.    1960.    Families  of  fishes  new  to  the  Antarctic.  Communique 

3.  Pelagic  young  of  flatfish  (Pisces,  Bothidae)  off  the  Antarctic  coast. 

Zool.  Zh.  39:1056-1061. 
.    1962.    Bathypelagic  fishes  of  the  Antarctic.  1.  Family  Myc- 

tophidae.  Akad.  Nauk  SSSR  Issled.  Fauny  Morei  l(9):2l6-294.  (In 

Russian;  Bur.  Comm.  Fish.  Ichthyol.  Lab.  Transl.  29.) 
.    1965.    A  general  review  of  the  Antarctic  fish  fauna.  Monogr. 


Biol.  15:491-550. 

— .  1975.  A  new  ultra-abyssal  fish,  Notolipans  kurchatovi  gen.  et 
sp.  n.  (Liparidae)  from  the  South  Orkney  Trench  (.Antarctica).  Tr. 
Inst.  Okeanol.  Akad.  Nauk  SSSR  103:313-319. 

— ,  and  a.  V.  Neelov.  1976.  Geniolipans  lindhergi,  gen.  et  sp. 
n.— a  new  fish  of  the  family  of  sea  snails  (Liparidae)  from  the  bathyal 
depths  of  Antarctica,  p.  68-77.  In:  Zoogeography  and  systematics 
of  fishes.  V.  M.  Korovina  (ed).  Zool.  Inst.  Acad.  Sci.  USSR,  Len- 
ingrad. 

— , ,  and  V.  P.  Prirodina.    1977.   On  methods  of  study  of 


the  morphology  and  systematics  of  the  fish  family  of  sea  snails 

(Liparidae).  Zool.  Zh.  50:141-147. 
AoYAMA,  T.    1959.    On  the  egg  and  larval  stages  of  Liparis  tanakae 

(Gilbert  and  Burke).  Bull.  Seikai  Reg.  Fish.  Res.  Lab.  18:69-73. 
,  and  N.  Sotogak.1.    1955.    On  the  development  of  the  egg  of 

Nemtplerus  virgalus  (Houttuyn).  Jpn.  J.  Ichthyol.  6:130-132. 
Aprieto,  V.  L.    1974.   Early  development  of  five  carangid  fishes  of  the 

Gulf  of  Mexico  and  the  south  Atlantic  coast  of  the  United  States. 

Fish.  Bull.  U.S.  72:415-443. 
Apsangikar,  D.  K.    1953.   The  systematic  position  of  5/TO»Mre«i  «/- 

ger.  J.  Univ.  Bombay,  N.S.  21B(5):41-50. 
Apstein,  C.    1909.   Die  Bestimmungdes  Alters  pelagisch  lebenderSee- 

fischeier.  Mitt.  Dtsch.  Seefischereiver  25:364-373. 
Arata,  G.  F..  Jr.     1954.    A  contribution  to  the  life  history  of  the 

swordfish.  Xiphias  gladius  Linnaeus,  from  the  South  Atlantic  coast 

of  the  United  States  and  the  Gulf  of  Mexico.  Bull.  Mar.  Sci.  Gulf 

Caribb.  4:183-243. 
Arbault,  S.,  and  N.  Boutin.    1968a.   Ichthyoplancton.Oeufset  larves 

de  poissons  teleosteens  dans  le  golfe  de  Gascogne  en  1964.  Rev. 

Trav.  Inst.  Peches  Marit.  32:413-476. 


,  and .  1968b.  Ichthyoplancton.  Oeufs  et  larves  de  pois- 
sons teleosteens  dans  le  Golfe  Gascogne  en  1964.  Rev.  Trav.  Inst. 
Peches  Marit.  32:477-506. 

,  AND .  1968c.  Ichthyoplancton.  Oeufs  et  larves  de  pois- 
sons teleosteens  dans  le  Golfe  de  Gascogne  en  1965-1966  (Oeufs 
et  larves).  Rev.  Trav.  Inst.  Peches  Marit.  33:181-202. 

Armstrong,  P.  B.  1962.  Stages  in  the  development  of /c7a/Hn« /ic/)- 
ulosus.  Syracuse  University  Press,  Syracuse,  New  York. 

,  AND  J.  S.  Child.    1965.    Stages  in  the  normal  development  of 

Fundutus  heteroditus.  Biol.  Bull.  (Woods  Hole)  128(2):143-168. 

Arnold,  D.C.  1956.  A  systematic  revision  of  the  teleost  family  Carap- 
idae  (Percomorphi.  Blennioidea),  with  descriptions  of  two  new 
species.  Bull.  Bnt.  Mus.  (Nat.  Hist.)  Zool.  4:245-307. 

Aronovich,  T.  M.,  S.  I.  DoROSHEv,  L.  V.  Spectorova  and  V.  M. 
Makhotin.  1975.  Egg  incubation  and  larval  rearing  of  navaga 
(Eleginus  navaga  Pall)  polar  cod  (Boreogadus  saida  Lepechin)  and 
arctic  flounder  (Liopsetta  glacialis  Poll.)  in  the  laboratory.  Aqua- 
culture  6:233-242. 

AsANo,  H.  1962.  Studies  on  the  congrid  eels  of  Japan.  Bull.  Misaki 
Mar.  Biol.  Inst.  Kyoto  Univ.  1:1-143. 

Astakhov,  D.  A.  1980.  Cranial  lateral-line  canals  of  the  beloniform 
fishes  (Beloniformes,  Osteichthyes).  Tr.  Inst.  Okeanol.  Akad.  Nauk 
SSR  97:174-21 1.  (In  Russian,  English  abstract.) 

Auer,  N.  a.  (ed.).  1982.  Identification  of  larval  fishes  of  the  Great 
Lake  Basin  with  emphasis  on  the  Lake  Michigan  Drainage.  Great 
Lakes  Fish.  Comm.  Spec.  Publ.  82-3. 

Auten,  M.,  AND  F.  Bonhomme.  1980.  Elements  de  systematique  bio- 
chimique  chez  les  Mugilides  de  Mediterranee.  Biochim.  Syst.  Ecol. 
8(3):305-308. 

AziiMA,  H.  1975.  Spawning  the  red-bellied  piranha,  Serrasabnus  nat- 
tereri.  Trop.  Fish  Hobby.  23(ll):4-6.  8-12,  14-15,  96-97. 

.    1982.    Spawning  silver.  Trop.  Fish  Hobby.  30(7):70-73. 

Backhouse,  G.  N.,  and  R.  W.  V.anner.  1978.  Observations  on  the 
biology  of  the  dwarf  galaxiid  Galaxiella  pusilla  (Mack)  (Pisces: 
Galaxiidae).  Victorian  Nat.  95:128-132. 

Backl'S,  R.  H..  J.  E.  Craddock.  R.  L.  Haedrich  and  B.  H.  Robison. 
1977.  Atlantic  mesopelagic  zoogeography,  p.  266-287.  In:  Fishes 
of  the  western  North  Atlantic.  R.  H.  Gibbs,  Jr.  (ed.).  Mem.  Sears 
Found.  Mar.  Res.  No.  1  (Pt.  7). 

,  G.  W.  Mead,  R.  L.  Haedrich  AND  A.  W.  Ebeling.    1965.  The 

mesopelagic  fishes  collected  during  Cruise  1 7  of  the  R/V  Chain, 
with  a  method  for  analyzing  faunal  transects.  Bull.  Mus.  Comp. 
Zool.  134:139-157. 

Badcock,  J.  R.  1982.  A  new  species  of  the  deep-sea  fish  genus  Cy- 
clothone  Goode  and  Bean  (Stomiatoidei,  Gonostomatidae)  from 
the  tropical  Atlantic.  J.  Fish  Biol.  20:197-21 1. 

,  .\ND  N.  R.  Merrett.    1972.    On  Argynpnus  allanlicus.  Maul 

1952  (Pisces,  Stomiatoidei),  with  a  description  of  post-larval  forms. 
J.  Fish  Biol.  4:277-287. 

,  AND .    1976.    Midwater  fishes  in  the  eastern  North  At- 


lantic. I.  Vertical  distribution  and  associated  biology  in  30°N,  23°W, 
with  developmental  notes  on  certain  myctophids.  Prog.  Oceanogr. 
7:3-58. 

Bagarinao.T.,  AND  J.  R.  Hunter.  1983.  The  visual  feeding  threshold 
and  action  spectrum  of  northern  anchovy  (Engraulis  mordax  G.) 
larvae.  Calif  Coop.  Oceanic  Fish.  Invest.  Rep.  24:245-254. 

Bailey.  C.  F.  1973.  Lipids  of  the  fertilized  egg  and  adult  brain  of 
Fundulus  heteroditus.  J.  Exp.  Zool.  185:265-276. 

Bailey,  K.  M.  1982.  The  early  life  history  of  the  Pacific  hake  Mer- 
lucaus  productus.  Fish.  Bull.  U.S.  80:589-598. 

,  AND  R.  S.  Batty.    1983.    A  laboratory  study  of  predation  by 

Aurelia  aurita  on  larval  herring  (Clupea  harengits).  Experimental 
observations  compared  with  model  predictions.  Mar.  Biol.  72:295- 
301. 

Bailey,  R.  G.  1972.  Observations  on  the  biology  oi  Nolhobranchius 
guentheri  (PfefCer)  (Cyprinodontidae),  an  annual  fish  from  the  coast- 
al region  of  East  Afnca.  Afr.  J.  Trop.  Hydrobiol.  Fish.  2:33-47. 

Bainbridoe,  V.  1961.  The  early  life  history  of  the  bonga  Ethmalosa 
dorsalis  (Cuvier  and  Valenciennes).  J.  Cons.  Int.  Explor.  Mer  26: 
347-353. 


LITERATURE  CITED 


691 


.    1962.    The  larvae  of  fe/ZoMi^/a  vora.v  Gunther  (Clupeidae)  in 

Sierra  Leone  coastal  waters.  Bull.  Inst.  Fr.  Afr.  Noire,  Ser.  A  24: 
262-269. 

Baird,  R.  C.  1971.  The  systematics.  distribution,  and  zoogeography 
of  the  marine  hatchetfishes  (Family  Stemoptychidae).  Bull.  Mus. 
Comp.  Zool.  142:1-128. 

,  D.  F.  Wilson,  R.  C.  Beckett  .-".nd  T.  L.  Hopkins.    1974.   Dia- 

phus  taamngi  Norman,  the  principal  component  of  a  shallow  sound- 
scattering  layer  in  the  Cariaco  Trench,  Venezuela.  J.  Mar.  Res.  32: 
301-312. 

, AND  D.  M.  MiLLiKEN.     1973.    Observations  on  Breg- 


maceros  nectabanus  Whitley  in  the  anoxic,  sulfurous  waters  of  the 

Cariaco  Trench.  Deep-Sea  Res.  20:503-504. 
Baker,  A.  N.    1973.    Spawning  and  development  of  the  New  Zealand 

sprat  Spratlus  antipodum.  Zool.  Publ.  Victoria  Univ.  Wellington 

62:1-12. 
Balakrishnan,  K.  p.     1961.    Fish  eggs  and  larvae  collected  by  the 

research  vessel  "Conch"  during  1957-61.  1.  Larvae  of  C'v«og/oMi« 

semifasciatus  Day.  Bull.  Cent.  Res.  Inst.  Univ.  Kerala  India  Ser.  C 

Nat.  Sci.  8:131-139. 
.    1963.    Fish  eggs  and  larvae  collected  by  the  research  vessel 

"Conch":   I.  Larvae  oi  Arnoglossus  tapaemosoma  Blkr.,  Bolhus 

ocellatus  Agassiz,  Laeops  giienlhen  Ale,  Solea  ovala  Rich.,  and 

Cynoglossus  monopus  Blkr.  Bull.  Dep.  Mar.  Biol.  Oceanogr.  Univ. 

Kerala  1:81-96. 
,  AND  C.  B.  L.  Devi.    1974.    Larvae  of  some  Hal  fishes  from  a 


tropical  estuary,  p.  677-684.  In:  The  early  life  history  offish.  J.  H. 

S.  Blaxter  (ed.).  Springer- Verlag,  New  York. 
Balbontin,  P.,  I.  Campodonico  AND  L.  Guzman.    1979.   Descripcion 

de  huevos  y  larvas  de  especies  de  Careproctus  comensales  de  Par- 

alomis granutosay  Lithodes anlarctica.  An.  Inst.  Patagonia  10:235- 

243. 
,  AND  R.  Perez.    1980.    Descripcion  de  los  estados  larvales  de 

Normanichthys  crocken  Clark  (Perciformes:  Normanichthyidae)  del 

area  de  Valparaiso.  Rev.  Biol.  Mar.  17(l):81-95. 
Ballard,  J.     1970.    Aquanum  briefs.  Bull.  S.  Afr.  Assoc.  Mar.  Biol. 

Res.  8:31-34. 
Balon,  E.  K.    1956.    Development  of  the  huchen  [Hiicho  huclio  (L.)] 

during    the    period    of  exogenous    feeding    after    hatching. 

Pornohospodarstvo  3:433-455.  (In  Slovak.) 
.    1959.    Die  Entwicklung  der  Texas— Cichlide  (//cnc/»/n'i  cvfl- 

nogultalus  Baird  et  Girard)  nach  dem  Schliipfen.  Zool.  Anz.  162 

(ll-l2):339-355. 
.    1975a.    Reproductive  guilds  of  fishes:  a  proposal  and  defini- 
tion. J.  Fish.  Res.  Board  Can.  32:821-864. 
.    1975b.   Terminology  of  intervals  in  fish  development.  J.  Fish. 

Res.  Board  Can.  32:1663-1670. 
.    1977.    Early  onxogeny  of  Labeotropheus  Ahl.  1927  (Mbuna, 

Cichlidae,  Lake  Malawi),  with  a  discussion  on  advanced  protective 

styles  in  fish  reproduction  and  development.  Environ.  Biol.  Fishes 

2:147-176. 
(ed.).    1980.    Charrs:  salmonid  fishes  of  the  genus  Salvelinus. 

Perspectives  in  vertebrate  science  I.  Dr.  W,  Junk,  The  Hague. 
.     1981a.    Additions  and  amendments  to  the  classification  of 

reproductive  styles  in  fishes.  Environ.  Biol.  Fishes  6:377-389. 
1981b.    Saltatory  processes  and  altncial  to  precocial  forms  in 


the  ontogeny  of  fishes.  Am.  Zool.  21:573-596. 
BANARESfU,  P.    1964.    Pisces.  Osteichthyes.  Fauna  Repub.  Pop.  Rom. 

(Bucuresti)  13. 
Banister,  K.  E.     1967.    The  anatomy  and  classification  of  the  order 

Gasterosteiformes  (Pisces).  Unpubl.  PhD  dissertation.  Univ.  New- 
castle upon  Tyne. 
Bapat,  S.  V.    1955.    A  preliminary  study  of  the  pelagic  fish  eggs  and 

larvae  of  the  Gulf  of  Mannar  and  the  Palk  Bay.  Indian  J.  Fish. 

2:231-255. 
,  and  R.  R.  Prasad.    1952.    On  some  developmental  stages  of 

Caranx  kalla  Cuv.  &  Val.  J.  Bombay  Nat.  Hist.  Soc.  51:111-115. 
Baracloi'gh,  W.  E.    1964.    Contribution  of  the  marine  life  history  of 

Thateichlhys  paaficus.  J.  Fish.  Res.  Board  Can.  21:1333-1337. 


Barans.  C.  A.,  AND  A.  C.  Barans.  1972.  Eggs  and  early  larval  stages 
of  the  spotted  hake,  Urophycis  regius.  Copeia  1972:188-190. 

Barham,  E.  G.  1966.  An  unusual  pelagic  flatfish  observed  and  pho- 
tographed from  a  diving  saucer.  Copeia  1966:865-867. 

Barkman,  R.  C.  1978.  The  use  of  otolith  growth  rings  to  age  young 
Atlantic  silversides,  Menidia  menidia.  Trans.  .Am.  Fish.  Soc.  107: 
790-792. 

Barnard,  K,  H.  1926.  The  eggs  of  the  pilot-fish  (A'aiicTa/M  c/wcvor). 
Nature  (Lond.)  118:228. 

.    1927.   A  monograph  of  the  marine  fishes  of  South  Africa.  Pari 

II.  Ann.  S.  Afr.  Mus.  21:419-1065. 

Barnhart,  P.  S.  1927.  Pelagic  fish  eggs  of  La  Jolla,  California.  Bull. 
Scripps  Inst.  Oceanogr.  Univ.  Calif.  1:91-92. 

.    1932.    Notes  on  the  habits,  eggs,  and  young  of  some  fishes  of 

southern  California.  Bull.  Scripps  Inst.  Oceanogr.  Univ.  Calif.  3(4): 
87-99. 

Barrington,  E.  J.  W.  1937.  The  structure  and  development  of  the 
tail  in  the  plaice  (Pleiironecles  platessa)  and  the  cod  (Gadus  mor- 
rhua).  Q.  J.  Microsc.  Sci.  79:447-469. 

Barsukov,  V.  V.  1959.  The  wolffish  (Anarhichadidae).  Fauna  SSSR 
Moscow  N.S.  73.(Transl.  by  Smithson.  Inst.,  1972.  Available  from: 
NTIS,  Spnngfield,  VA.:  TT  67-59074.) 

.    1973.    The  species  composition  of  the  genus //e&o/e^wi  (Se- 

bastinae.  Scorpaenidae:  Pisces)  and  a  description  of  a  new  species. 
J.  Ichthyol.  (Engl.  Transl.  Vopr.  Ikhtiol.)  13:161-167. 

.    1981.   A  brief  review  of  the  subfamily  Sebastinae.  J.  Ichthyol. 

(Engl.  Transl.  Vopr.  Ikhtiol.)  21(l):l-26. 

Bartlett,  M.  R.,  and  R.  L.  Haedrich.  1966.  Techniques  in  the 
radiography  of  fishes.  Trans.  Am.  Fish.  Soc.  95:99-101. 

Bassot,  J.-M.  1966.  On  the  comparative  morphology  of  some  lu- 
minous organs,  p.  557-610.  In:  Bioluminescence  in  Progress.  F. 
Johnson  and  Y.  Haneda  (eds.).  Princeton  Univ.  Press.  Princeton, 
N.J. 

.    1971.    Structure  and  evolution  of  the  light  organs  of  the  sto- 

miatoidei  (Abstract),  p.  253-254.  In:  The  ocean  world.  M.  Uda 
(ed.).  Proc.  Joint  Oceanogr.  Assembly. 

Bath,  H.  1973.  Wiederbeschreibung  und  neuer  Nachweis  von  Tnpre- 
rygion  melanurus  (Guichenot)  (Pisces:  Blennioidea:  Tripterygi- 
idae).  Senckenb.  Biol.  54:47-56. 

.    1976.    Revision  der  Blenniini  (Pisces:  Blenniidae).  Senckenb. 

Biol.  57:167-234. 

.    1978.    Erganzungen  zur  Revision  der  Blenniini  mit  zwei  neun 

Gattungen  (Pisces:  Blenniidae).  Senckenb.  Biol.  59:183-190. 

Badchot,  M.  L.  1959.  Etude  des  larves  leptocephales  du  group  Lep- 
lochephalus  lanceolalus  Stromman  et  identification  a  la  famille  des 
Serrivomeridae.  Dana-Rep.  Carlsberg  Found.  48:1-148. 

Baugh,  T.  M.  1980.  Spawning  and  hatching  the  Olympic  mudminnow 
{NoYumhra  hubbsi)  in  aquaria.  J.  Aquaric.  l(3):63-64. 

Baxter,  J.  L.  1 960.  A  study  of  the  yellowtail  Seriola  dorsalis  (Gill). 
Calif.  Dep.  Fish  Game  Fish  Bull.  1 10. 

Beck,  A.  D.,  D.  A.  Benotson  AND  W.  H.  Howell.  1980.  International 
study  on  Anemia.  V.  Nutritional  value  of  five  geographical  strains 
of  Anemia:  effects  on  survival  and  growth  of  larval  Atlantic  sil- 
verside,  Menidia  menidia.  p.  249-259.  In:  The  brine  shrimp  .Ir- 
lemia.  Vol.  3.  Ecology,  cultunng,  use  in  aquaculture.  G.  Persoone, 
P.  Sorgeloos,  O.  Roels  and  E.  Jaspers  (eds.).  Universa  Press.  Wet- 
teren,  Belgium. 

Beebe,  W.  1932.  Nineteen  new  species  and  four  post-larval  deepsea 
fish.  Zoologica  (N.Y.)  13:47-107. 

.  1933a.  Deep-sea  fishes  of  the  Bermuda  Oceanographic  Ex- 
peditions. No.  2.  Family  Alepocephalidae.  Zoologica  (N.Y.)  1 6(2): 
15-93. 

.  1933b.  Deep-sea  fishes  of  the  Bermuda  Oceanographic  Ex- 
peditions. No.  3.  Family  Argentinidae.  Zoologica  (N.Y.)  I6(3):97- 
147. 

.  1934.  Deep-sea  fishes  of  the  Bermuda  Oceanographic  Expe- 
ditions. Family  Idiacanthidae.  Zoologica  (N.Y.)  16(4):  149-241. 

,  AND  J.  Crane.  1939.  Deep-sea  fishes  of  the  Bermuda  Ocean- 
ographic Expeditions.  Family  Melanostomiatidae.  Zoologica  (N.Y.) 
24(2):65-238. 


692 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Behnke,  R.  J.  1968.  A  new  subgenus  and  species  of  trout.  Salmo 
(Platysalmo)  platycephalus.  from  southcentral  Turkey,  with  com- 
ments on  the  classification  of  the  subfamily  salmoninae.  Mitt.  Hamb. 
Zool.  Mus.  Inst.  66. 

Bekker,  V.  E.  1963.  Taxonomy  and  distribution  of  Tarletonheania 
(Myctophidae,  Pisces).  Tr.  Inst.  Okeanol.  Akad  Nauk  SSSR  62: 
145-163.  (In  Russian.) 

.  1964.  Slendertailed  myctophids  (genera  Loweina.  Tarleton- 
heania. Gonichthys  and  Centrobranchus)  of  the  Pacific  and  Indian 
Oceans,  systematics  and  distribution.  Tr.  Inst.  Okeanol.  Akad.  Nauk 
SSSR  73:11-75  (In  Russian.) 

.     1965.    Lantemfishes  of  the  genus  A/i',?cip/!i/»;  (Myctophidae, 

Pisces),  systematics  and  distribution.  Tr.  Inst.  Okeanol.  Akad.  Nauk 
SSSR  80:62-103.  (In  Russian;  Bur.  Comm.  Fish.  Ichthyol.  Lab. 
Transl.  45.) 

.    1967.    The  lantemfishes  (Myctophidae)  from  the  "Peter  Le- 

bedev"  Atlantic  Expeditions  1961-1964.  Tr.  Inst.  Okeanol.  Akad. 
Nauk  SSSR  84:84-124.  (In  Russian.) 

,  AND  O.  D.  BoRODLiLiNA.     1968.    Lantcmfishes  of  the  genus 

Ceratoscopelus  Gunther,  systematics  and  distribution.  Vopr.  Ikh- 
tiol.  8(5):779-798.  (In  Russian.) 

,  AND  .     1978.    "Myclophum  asperum"  species— group 

with  description  of  a  new  species,  and  Myclophum  selenops  TSning 
(Myctophidae,  Osteichthys),  taxonomy  and  distribution.  Tr.  Inst. 
Okeanol.  Akad.  Nauk  SSSR  111:108-128.  (In  Russian.) 

Belanger,  L.  F.  1973.  The  skeletal  tissues,  p.  175-215. /«.  Histology. 
R.  O.  Creep  and  L.  Weiss  (eds.).  McGraw  Hill,  New  York. 

Bell,  M.  A.  1976.  Evolution  of  phenotypic  diversity  in  Gasterosleus 
aculeatus  superspecies  on  the  Pacific  Coast  of  North  America.  Syst. 
Zool.  25:211-227. 

Belman,  B.  W.,  AND  M.  E.  Anderson.  1979.  Aquanum  observations 
on  feeding  by  Melanostigma  pammelas  (Pisces.  Zoarcidae).  Copeia 
1979:365-369. 

Belyanina,  T.  N.  1974.  Materials  on  the  development,  systematics 
and  distribution  of  fish  of  the  family  Bregmacerotidae.  Tr.  Inst. 
Okeanol.  Akad.  Nauk.  SSSR  96:143-188. 

.    1977.    Materials  on  development  of  C/ia«/(0(^Mi  fishes  (Chau- 

liodontidae,  Pisces).  Tr.  Inst.  Okeanol.  Akad.  Nauk.  SSSR  109: 
113-132. 

.    1 980.    Codlets  (Bregmacerotidae,  Osteichthyes)  of  the  Carib- 


bean Sea  and  the  Gulf  of  Mexico.  J.  Ichthyol.  (Engl.  Transl.  Vopr. 

Ikhtiol.)  20(0:138-141. 
.     1981.    The  larvae  of  some  rare  mesopelagic  fishes  from  the 

Caribbean  and  the  Gulf  of  Mexico.  J.  Ichthyol.  (Engl.  Transl.  Vopr. 

Ikhtiol.)  21(l):82-95. 
.    1982a.   Larvaeof  the  family  Scopelarchidae  (collections  of  the 

Institute  of  Oceanology  Academy  of  Sciences  of  USSR).  Tr.  Inst. 

Okeanol.  Akad.  Nauk  SSSR.  1  18:133-150.  (In  Russian.) 
.    1982b.    Larvae  of  the  midwater  fishes  in  the  western  tropical 

Pacific  Ocean  and  the  seas  of  the  Indo-Australian  Archipelago.  Tr. 

Inst.  Okeanol.  Akad.  Nauk  SSSR  1 18:5-42. 
,  AND  N.  V.  KovALEvsicAVA.    1979.    Materials  of  development 

and  distribution  of  myctophid  larvae  (Family  Myctophidae)  of  the 

sub-antarctic  waters  of  the  Australian-New  Zealand  waters.  Tr.  Inst. 

Okeanol.  Akad.  Nauk  SSSR  106:69-96. 
,  AND  G.  R.  Lopes.    1974.   The  occurrence  oi  Bregmaceros  nec- 

tabanus  ( Pisces,  Bregmacerotidae)  in  the  Caribbean  Sea.  J.  Ichthyol. 

(Engl.  Transl.  Vopr.  Ikhtiol.)  14:447-449. 
Bennett,  H.  S.,  A.  D.  Wyrick,  S.  W.  Lee  and  J.  H.  McNeil.    1976. 

Science  and  art  in  preparing  tissue  embedded  in  plastic  for  light 

microscopy,  with  special  reference  to  glycol  methacrylate,  glass 

knives  and  simple  stains.  Stain  Technol.  51:71-97. 
Bensam,  p.    1 967.   The  embryonic  and  early  larval  development  of  the 

long-finned  herring,  Opisthoplenis  lardoore  (Cuvier).  J.  Mar.  Biol. 

Assoc.  India  9:76-83. 
.    1970.    Notes  on  the  eggs,  larvae  and  juveniles  of  the  Indian 

Sprat,  Sardinella  jussieui  (Lacepede).  Indian  J.  Fish.  13:219-231. 
Ben-tlivia,  A.    1975.    Mugilid  fishes  of  the  Red  Sea  with  a  key  to  the 

Mediterranean  and  Red  Sea  species.  Bamidgeh  27(1):  14-20. 


Benzie,  V.  1968a.  Stages  in  the  normal  development  of  Galaxias 
maculatus  allenualus  Uenyn^).  N.Z.  J.  Mar.  Freshwater  Res.  2:606- 
627. 

.    1968b.    The  life  history  of  Galaxias  vulgaris  Stokell,  with  a 

comparison  with  G.  maculalus  allenualus.  N.Z.  J.  Mar.  Freshwater 
Res.  2:628-653. 

Berg,  L.  S.  1940.  Classification  of  fishes,  both  recent  and  fossil.  Tr. 
Zool.  Inst.  Akad.  Nauk  SSSR  5(2).  (In  Russian.) 

,     1947.    Classification  of  fishes  both  recent  and  fossil.  J.  W. 

Edwards,  Ann  Arbor,  Michigan. 

.    1948.   Freshwater  fishes  of  the  U.S.S.R.  and  adjacent  countries. 

Vol.  I.  Acad.  Sci.  U.S.S.R.  Zool.  Inst..  Keys  to  the  fauna  of  the 
U.S.S.R.,  No.  27.  (Translated  from  the  Russian,  1962,  Israel  Pro- 
gram for  Scientific  Translations.) 

Berkeley,  S.  A.,  and  E.  D.  HouDE.  1978.  Biology  of  two  exploited 
species  of  halfbeaks,  Hemiramphus  brasiliensis  and  H.  balao  from 
southeast  Flonda.  Bull.  Mar.  Sci.  28:624-644. 

Berra,  T.  M.  1982.  Life  history  of  the  Australian  grayling  Pro/o/rortM 
maraena  (Salmoniformes:  Prototroctidae)  in  the  Tambo  River, 
Victoria.  Copeia  1982:795-805. 

Berrien,  P.  L.,  N.  A.  Naplin  and  M.  R.  Pennington.  1981.  Atlantic 
mackerel  Scomber  scombrus.  egg  production  and  spawning  popu- 
lation estimates  for  1977  in  the  Gulf  of  Maine,  Georges  Bank  and 
Middle  Atlantic  Bight.  Rapp.  P.  V.  Reun.  Cons.  Int.  Explor.  Mer 
178:279-288. 

Berr'i,  F.  H.  1958.  A  new  species  of  fish  from  the  western  North 
Atlantic,  Dikellorhynchus  tropidnlepis.  and  relationships  of  the  gen- 
era Dikellorhynchus  and  Malacanthus.  Copeia  1958:1  16-125. 

.  1959a.  Boarfishes  of  the  genus  .-)«r;,?o^i/a  of  the  western  At- 
lantic. Bull.  Ha.  State  Mus.  Biol.  Sci.  4(7):205-250. 

.  1959b.  Young  jack  crevalles  (Caraw.v  species)  off  the  south- 
eastern Atlantic  coast  of  the  United  States.  U.S.  Fish  Wildl.  .Serv. 
Fish.  Bull.  59:417-535. 

.    1960.  Scaleandscutedevelopment  of  the  carangid  fish  Cara^i.v 

crysos  (Mitchell).  Q.  J.  Ha.  Acad.  Sci.  23:59-66. 

.    1 964a.   Aspects  of  the  development  of  the  upper  jaw  bones  in 

teleosts.  Copeia  1964:375-384. 

.    1964b.   Review  and  emendation  of  Family  Clupeidae.  Copeia 

1964:720-730. 

.    1969.   f/agato  Wp(/j/7Mtoa  (Pisces:  Carangidae):  morphology 


of  the  fins  and  other  characters.  Copeia  1969:454-463. 

— ,  and  W.  W.  Anderson.  1961.  Stargazer  fishes  from  the  western 
North  Atlantic  (family  Uranoscopidae).  Proc.  U.S.  Natl.  Mus.  1 12 
(3448):563-586. 

— ,  and  W.  J.  Baldwin.  1966.  Triggerfish  (Balistidae)  of  the  east- 
em  Pacific.  Proc.  Calif  Acad.  Sci.  34:429-474. 

— ,  and  I.  Barrett.  1963.  Gillraker  analysis  and  speciation  in 
the  thread  herring  genus  Opisthonema.  Inter-Am.  Trop.  Tuna 
Comm.  Bull.  7:112-190. 

— ,  AND  H.  C.  Perkins.  1966.  Survey  of  pelagic  fishes  of  the 
California  Current  area.  U.S.  Fish  Wildl.  Serv.  Fish.  Bull.  65:625- 
682. 

— ,  AND  W.  J.  Richards.  1973.  Characters  useful  to  the  study  of 
larval  fishes.  Mid.  Atl.  Coast.  Fish.  Cent.  Tech.  Pap.  1:48-65. 

— ,  and  C.  R.  Robins.  1967.  Macnstiella  perlucens.  a  new  clu- 
peiform  fish  from  the  Gulf  of  Mexico.  Copeia  1967:46-50. 

— ,  AND  W.  F.  Smith-Vaniz.  1978.  Carangidae. //!.  FAO  species 
identification  sheets  for  fisheries  purposes.  Western  Central  Atlan- 
tic, Fishing  Area  31.  Vol.  I.  W.  Fischer  (ed.).  FAO,  Rome. 

— ,  AND  L.  E.  VoGELE.    1961.    Filefishes  (Monacanthidae)  of  the 


western  north  Atlantic.  U.S.  Fish  Wildl.  Serv.  Fish.  Bull.  61:61- 
109. 

Bertelsen,  E.  1951.  The  ceratioid  fishes.  Ontogeny,  taxonomy,  dis- 
tribution and  biology.  Dana-Rep.  Carlsberg  Found.  39. 

.  1958.  The  argenlinoid  fish  Xenoplhalmichthys  danae.  Dana- 
Rep.  Carlsberg  Found.  45. 

.    1980.    Notes  on  Linophrynids.  V.  A  revision  of  the  deep-sea 

anglerfishes  (Ceratioidci)  of  the  Lituyphryne  arborifera-group. 
Steenstrupia  6(6):29-70. 


LITERATURE  CITED 


693 


— .    1982.    Notes  on  Linophnnidae.  VIII.  A  review  of  the  genus 

Lmophryne.  with  new  records  and  descriptions  of  two  new  species. 

Steenstrupia  8(3):49-104. 
— .    1983.    First  records  of  metamorphosed  males  of  the  families 

Diceratiidae  and  Centrophrvnidae.  Steenstrupia  8(  16):309-316. 
— ,  G.  Krefft  and  N.  B.  M,'\rsh.\ll.     1976.    The  fishes  of  the 

family  Notosudidae.  Dana-Rep.  Carlsberg  Found.  86. 
— ,  .AND  N.  B.  Marshall.    1956.    The  Mirapinnati.  a  new  order 

of  teleost  fishes.  Dana-Rep.  Carlsberg  Found.  42. 
— ,  AND .     1958.    Notes  on  Minpinnati  (an  addendum  to 

Dana-Rep.  42).  Dana-Rep.  Carlsberg  Found.  45:9-10. 
— ,  AND  O.  Mlink.    1964.    Rectal  light  organs  in  the  argentinoid 

fishes  Opislhoproctus  and  IVinleria.  Dana-Rep.  Carlsberg  Found. 

62. 
— ,  AND  T.  W.  PiETSCH.    1975.    Results  of  research  cruises  of  FRV 


"Walther  Herwig"  to  South  Amenca.  XXXVIIl.  Osteology  and 

relationships  of  the  ceratioid  anglerfish  genus  Spiniphryne  (family 

Oneirodidae).  .Arch.  Fischereiwiss.  26(1):  1-1  1. 
— , AND  R.  J.  Lavenberg.    1981.   Ceratioid  anglerfishes  of 

the  family  Giganlactinidae:  morphology,  systematics,  and  distn- 

bution.  Nat.  Hist.  Mus.  Los  Ang.  Cty.  Contnb.  Sci.  332. 
— .  AND  P.  J.  Struhsaker.    1977.   The  ceratioid  fishes  of  the  genus 

Thaumalichlhys:  osteology,  relationships,  distnbution  and  biology. 

Galathea  Rep.  14:7-40. 
-.  B.  Theisen  AND  O.  MuNK.    1965.   On  a  post-larval  specimen. 


anal  light  organ,  and  tubular  eyes  of  the  argentinoid  fish  Rhyn- 

chohyalus  natalensis  (Gilchnst  and  von  Bonde).  Vidensk.  Medd. 

Dan.  Naturhist.  Foren.  128:357-371. 
Bertin,  L.     1937.    Les  poissons  abyssau.x  du  genre  Cycma  Gunther 

(anatomie,  embryologie,  bionomie).  Dana-Rep.  Carlsberg  Found. 

10:1-30. 
Bertmar,  G.    1959.   On  the  ontogeny  of  the  chondral  skull  in  Charac- 

idae,  with  a  discussion  on  the  chondrocranial  base  and  the  visceral 

chondrocranium  in  fishes.  Acta  Zool.  (Stockh.)  40:203-364. 
Bertolini,  F.    1933a.   Famiglia  I:  Apogonidae,  p.  306-309.  In:  Uova, 

larve  e  sladi  giovanili  di  Teleostei.  Fauna  Rora  Golfo  Napoli 

Monogr.  38. 
.    1933b.    Famiglia  2:  Serranidae.  p.  310-331.  In:  Uova.  larve 

e  stadi  giovanili  di  Teleostei.  Fauna  Flora  Golfo  Napoli  Monogr. 

38. 
Beyer,  J.  E.,  AND  G.  C.  Laurence.    1981.    Aspects  of  stochasticity  in 

modelling  growth  and  survival  of  clupeoid  fish  larvae.  Rapp.  P.  V. 

Reun.  Cons.  Int.  Explor.  Mer  178:17-23. 
Bhargava,  H.  N.    1958.   The  development  of  the  chondrocranium  of 

Maslacemhelus  armalus  (Cuv.  et  Val.).  J.  Morphol.  102:401-426. 
Bhimachar,  B.  S.,  and  A.  David.    1945.  On  the  occurrence  of  "external 

gills"  in  the  \oach  —  Lepidocephalus  ihermalis  (C.  V.).  Curr.  Sci. 

14(3):74, 
BiGELOW,  H.  B.,  AND  W.  C.  Schroeder.    1953.    Fishes  of  the  Gulf  of 

Maine.  U.S.  Fish  Wildl.  Serv.  Fish.  Bull.  53. 
,  AND .    1963.   Family  Osmeridae,  p.  553-597. /«■  Fishes 

of  the  western  North  Atlantic.  H.  B.  Bigelow  (ed.).  Mem.  Sears 

Found.  Mar.  Res.  No.  1  (Pt.  3). 
,  AND  W.  W.  Welsh.    1925.    Fishes  of  the  Gulf  of  Maine.  Bull. 


Bur.  Fish.  Wash.  40:1-567. 
BiNi,  G.    1971.    Atlante  dei  pesci  delle  Coste  italiane.  Vol.  3.  Mondo 

Sommerso,  Milano. 
BiRDSONG,  R.  S.    1975.    The  osteology  of  Microgobius  signalus  Poey 

(Pisces:  Gobiidae),  with  comments  on  other  gobiid  fishes.  Bull.  Fl. 

State  Mus.  Biol.  Sci.  19(3):  135-1 87. 
Blaber.S.  J.,  D.  P.  CiRL  SAND  A.  K.Whitfield.    1981.   The  influence 

of  zooplankton  food  resources  on  the  morphology  of  the  estuanne 

clupeid  Gikhnstella  aestuanus  (Gilchnst.   1914).  Environ.  Biol. 

Fishes  6:351-355. 
Blache,  J.     1964.    Sur  la  presence  de  Luvarus  impenales  Raf   1810 

dans  I'Atlantique  onental  sud  revelee  par  la  decouverte  de  deux 

larves  au  Stade  Hystncinella  de  L.  Roule  (1924)  (Pisces,  Teleostei, 

Perciformi.  Luvareidei,  Luvariidae)  Cah.  ORSTOM  Ser.  Oceanogr. 

5:57-59. 


.    1977.   Leptocephales  des  poissons  Anguilliformes  dans  la  zone 

sud  du  Golfe  de  Guinee.  Faune  Trop.  20:1-381. 
Blackburn,  J.  E.    1973.    A  survey  of  the  abundance  and  distribution 

of  ichthyoplankton  in  Skagit  Bay,  Unpubl.  MS  thesis.  Univ.  Wash- 
ington, Seattle. 
Blackburn,  M.    1950.  The  Tasmanian  whitebait,  Loveltia  seali  (io\\i\- 

ston)  and  the  whitebait  fishery.  Aust.  J.  Mar.  Freshwater  Res. 

1:155-198. 
Blackler,  R.  W.    1974.    Recent  advances  in  otolith  studies,  p.  67-90. 

In:  Sea  fisheries  research.  F.  R.  Harden  Jones  (ed.).  John  Wiley  & 

Sons,  New  York. 
Blaxter,  J.  H.  S.    1965.  The  feeding  of  herring  larvae  and  their  ecology 

in  relation  to  feeding.  Calif.  Coop.  Oceanic  Fish.  Invest.  Rep.  10: 

78-88. 
.    1966.    The  effect  of  light  intensity  on  the  feeding  ecology  of 

herring,  p.  393-409.  In:  Light  as  an  ecological  factor.  R.  Bainbndge. 

G.  C.  Evans  and  O.  Rackman  (eds).  Blackwell,  Oxford. 
.    1968.    Rearing  herring  larvae  to  metamorphosis  and  beyond. 

J.  Mar.  Biol.  Assoc.  U.K.  43:17-28. 
.    1969a.    Development:  eggs  and  larvae,  p.  177-252.  In:  Fish 

physiology.  Vol.  3.  W.  S.  Hoar  and  D.  J.  Randall  (eds.).  Academic 

Press,  New  York. 
.    1969b.    Expenmental  reanng  of  pilchard  larvae,  Sardina  pil- 

chardus.  J.  Mar.  Biol.  Assoc.  U.K.  49:557-575. 
.    1971.    Feeding  and  condition  of  Clyde  herring  larvae.  Rapp. 

P.  V.  Reun.  Cons.  Int.  Explor.  Mer  160:129-136. 
.  (ED.).     1974.    The  early  life  history  of  fish.  Springer- Verlag, 

Berlin. 
.    1976.    Reared  and  wild  fish  — how  do  they  compare?  Tenth 

Europ.  Symp.  Mar.  Biol.,  Ostend,  Belgium.  Sept.  17-23.  1975,  Vol. 

1:11-26. 
,  J.  A.  Gray  AND  A.  C.  Best.    1983.   Structure  and  development 


of  the  free  neuromasts  and  lateral  line  system  of  the  herring.  J,  Mar. 

Biol.  Assoc.  U.K.  63:247-260. 
,  AND  G.  Hempel.    1963.    The  influence  of  egg  size  on  herring 

larvae.  J.  Cons.  Int.  Explor.  Mer  28:21 1-240. 
,  AND  F.  G.  T.  HoLLiDAY.    1963.   The  behaviour  and  physiology 

of  herring  and  other  clupeids.  Adv.  Mar.  Biol.  1:261-393. 
,  AND  J.  R.  Hunter.    1982.    The  biology  of  the  clupeoid  fishes. 

Adv.  Mar.  Biol.  20:1-223. 
,  AND  M.  E.  Staines.    1971.   Food  searching  potential  in  manne 

fish  larvae,  p.  467-485.  In:  Fourth  European  manne  biology  sym- 
posium. D.  J.  Crisp  (ed.).  University  Press,  Cambndge. 
Bleeker,  P.    1854-1857.   Bidrage  tot  de  kennis  der  Sphyraenoiden  van 

den  Indischen  Archipel.  Verh.  Baravi  Genoot.  26:1-22. 
.    1 874.  Equisse  d'un  systeme  naturel  des  Gobioides.  Arch.  Neerl. 

Sci.  9:289-331. 
Block,  B.  A.    1983.   Brain  heaters  in  the  billfish.  Am.  Zool.  23(4):936. 

(Abstract.) 
Bock,  W.  J.    1969.    The  concept  of  homology.  Ann.  N.Y.  Acad.  Sci. 

167:71-73. 
BoHLKE,  J.  E.    1957.    A  review  of  the  blenny  genus  Chaenopsis.  and 

the  description  of  a  related  new  genus  from  the  Bahamas.  Proc. 

Acad.  Nat.  Sci.  Phila.  109:81-122. 
.     1960a.    Comments  on  serranoid  fishes  with  disjunct  lateral 

lines,  with  the  descnption  of  a  new  one  from  the  Bahamas.  Not. 

Nat.  (Phila.)  330:1-1 1. 
.     1960b.    Two  new  Bahaman  species  of  the  clinid  fish  genus 

Paraclinus.  Not.  Nat.  (Phila.)  337:1-8. 
.    1 966.   Order  Lyomeri.  Deep-sea  gulpers,  p.  603-628.  In:  Fish- 
es of  the  western  North  Atlantic.  G.  W.  Mead  (ed.).  Mem.  Sears 

Found.  Mar.  Res.  No.  l(Pt.  5). 
.    1978.    Muraenidae,  Ophichthidae.  In:  FAO  species  identifi- 


cation sheets  for  fishery  purposes.  Western  Central  Atlantic,  Fishing 
Area  31.  Vol.  3.  W.  Fischer  (ed.).  FAO,  Rome. 
— ,  AND  D.  K.  Caldwell.    1961.    On  the  occurrence  of  the  dac- 
tyloscopid  fish  genus  Daclylagnus  in  the  west  Atlantic.  Bull.  Mar. 
Sci.  Gulf  Caribb.  11:537-542. 


694 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


,  AND  C.  C.  G.  Chaplin.     1968.    Fishes  of  the  Bahamas  and 

adjacent  tropical  waters.  Livingston  Publ.  Co..  Wynnewood,  PA. 

,  AND  C.  R.  Robins.    1 959.   Studies  on  fishes  of  the  family  Ophi- 

diidae.  II.  Three  new  species  from  the  Bahamas.  Proc.  Acad.  Nat. 
Sci.  Phila.  111:37-52. 

,  and  C.  H.  Robins.    1974.    Description  of  a  new  genus  and 

species  of  clinid  fish  from  the  western  Caribbean,  with  comments 
on  the  families  of  the  Blennioidea.  Proc.  Acad.  Nat.  Sci.  Phila. 
126(1):  1-8. 

,  AND  D.  G.  Smith.     1968.    A  new  xenocongrid  eel  from  the 

Bahamas,  with  notes  on  other  species  of  the  family.  Proc.  Acad. 
Nat.  Sci.  Phila.  120:25-43. 

,  AND  V.  G.  Springer.    1961.    A  review  of  the  Atlantic  species 

of  the  clinid  fish  genus  Slarksia.  Proc.  Acad.  Nat.  Sci.  Phila.  1 13(3): 
29-60. 

,AND .   1975.  A  new  genus  and  species  offish  (A'e/>!ac//>7i« 

aleleslos)  from  the  western  Atlantic  (Perciformes:  Clinidae).  Proc. 
Acad.  Nat.  Sci.  Phila.  127(7):57-61. 

BOLIN,  R.  1930.  Embryonic  development  of  the  labrid  fish  0.vi7w//i 
ca/(/ormcM5  Giinther.  Copeia  1930(4):I22-I28. 

.  1934.  Studies  on  California  Cottidae:  an  analysis  of  the  prin- 
ciples of  systematic  ichthyology.  Unpubl.  PhD  dissertation.  Stan- 
ford Univ.,  Stanford,  CA. 

.    1936a.    Embryonic  and  early  larval  stages  of  the  California 

anchovy.  Calif  Fish  Game  22:314-321. 

.     1936b.    The  systematic  position  of  Indoslomus  paradoxus 

Prashad  and  Mukerji,  a  new  fresh  water  fish  from  Burma.  J.  Wash. 
Acad.  Sci.  26:420-423. 

.    1939a.  A  new  stomiatoid  fish  from  California.  Copeia  1939(1): 


39-41. 
.    1939b.   A  review  of  the  myctophid  fishes  of  the  Pacific  Coast 

of  the  United  States  and  of  lower  California.  Stanford  Ichthyol. 

Bull.  1(4):89-156. 
.     1941.    Embryonic  and  early  larval  stages  of  the  cottid  fish 

Orthonopias  Iriacis  Starks  and  Mann.  Stanford  Ichthyol.  Bull.  2(3): 

73-82. 
Bond,  C.  E.,  and  T.  Uyeno.    1981.    Remarkable  changes  in  the  ver- 
tebrae of  perciform  fish  Scombrolabrax  with  notes  on  its  anatomy 

and  systematics.  Jpn.  J.  Ichthyol.  28:259-262. 
BoNHAM,  K.,  and  W.  H.  Bavliff.    1953.   Radiography  of  small  fishes 

for  meristic  studies.  Copeia  1953:150-151. 
Booth,  R.  A.    1967.   A  description  of  the  larval  stages  of  the  tomcod, 

Microgadus  tomcod.  with  comments  on  its  spawning  ecology.  Un- 
publ. PhD  dissertation.  Univ.  Connecticut,  Storrs. 
Bortone,  S.  a.    1977.    Osteological  notes  on  the  genus  Centroprislis 

(Pisces:  Serranidae).  Northeast  Gulf  Sci.  1:23-33. 
Boulenger,  G.  a.     1902.    Notes  on  the  classification  of  teleostean 

fishes.  IV.  On  the  systematic  position  of  the  Pleuronectidae.  Ann. 

Mag.  Nat.  Hist.,  Ser.  7.  10:295-304. 

.    1904.   Teleostei.  Cambndge  Nat.  Hist.  (Lond.)  7:539-727. 

.    1915.   Catalogueofthefreshwaterfishesof  Africa  in  the  British 

Museum  (Natural  History).  Vol.  3. 
Bowman,  A.    1920.    The  eggs  and  larvae  of  the  angler  (Lophius  pis- 

calorius  L.)  in  Scottish  waters.  Sci.  Invest.  Fish.  Board  Scotl.  2:1- 

42. 
Boyd,  J.  F.,  and  R.  C.  SiMMONDS.    1974.   Continuous  laboratory  pro- 
duction of  fertile  Fundulus  heieroditus  (Walbaum)  eggs  lacking 

chorionic  fibrils.  J.  Fish  Biol.  6:389-394. 
BoYDE,  A.,  AND  C.  Wood.     1969.    Preparation  of  animal  tissues  for 

surface  scanning  electron  microscopy.  J.  Microsc.  90:221. 
Bradbury,  M.  G.    1967.   The  genera  of  batfishes  (family  Ogcocephal- 

idae).  Copeia  1967:399-422. 
Branson,  B.  A.,  and  G.  A.  Moore.    1962.   The  lateralis  components 

of  the  acoustico-lateralis  system  in  the  sunfish  family  Centrarchi- 

dae.  Copeia  1962:1-108. 
Brauer,  a.    1902.    ijber  einige  von  der  Valdivia-Expedition  gesam- 

melte  Tiefseefische  und  ihre  Augen.  Sitzungsber.  Ges.  Beford.  Na- 

turwiss.  Marburg  (1901)  nr.  8:115-130. 


.    1906.   DieTiefsee-Fische.  I.SystematischerTeil.  Wiss.  Ergebn. 

Dtsch.  Tiefsee-Exped.  Valdivia  15(1). 
.    1908.   DieTiefsee-Fische.  II.  AnatomischerTeil.  Wiss.  Ergebn. 

Dtsch.  Tiefsee-Exped.  Valdivia  15(2). 
Bray,  D.    1983.   A  revision  of  the  family  Zeidae.  Unpubl.  MSc  thesis. 

Australia.  (Complete  reference  not  available.) 
Breder,  C.  M.,  Jr.    1923.    Some  embryonic  and  larval  stages  of  the 

winter  flounder.  Bull.  Bur.  Fish.  Wash.  38:31  1-315. 
.    1927.    Scientific  results  of  the  first  oceanographic  expedition 

of  the  "Pawnee"  1925.  Fishes.  Bull.  Bingham  Oceanogr.  Collect. 

Yale  Univ.  l(l):l-90. 
.     1932.    On  the  habits  and  development  of  certain  Atlantic 

Synentognathi.  Carnegie  Inst.  Wash.  Publ.  435,  Pap.  Tortugas  Lab. 

28(l):l-35. 
.     1938.    A  contribution  to  the  life  histories  of  Atlantic  flying 

fishes.  Bull.  Bingham  Oceanogr.  Collect.  Yale  Univ.  6(5):1-121. 
.     1939.    On  the  life  history  and  development  of  the  sponge 

blenny,  Paraclmus  marmoratus  (Steindachner).  Zoologica  (N.Y.) 

24:487-496. 
.     1941.    On  the  reproductive  behavior  of  the  sponge  blenny, 

Paraclinus  marmoratus  (Steindachner).  Zoologica  (N.Y.)  26:233- 

236. 
.    1 944.    Materials  for  study  of  the  life  history  of  Tarpon  atlan- 


ticus.  Zoologica  (N.Y.)  29:217-252. 
.     1959,    Observations  on  the  spawning  behavior  and  egg  de- 
velopment of  Strongylura  notaia  (Poey).  Zoologica  (N.Y.)  44: 141- 

147. 
.    1962.    On  the  significance  of  transparency  in  osteichthid  fish 

eggs  and  larvae.  Copeia  1962:561-567. 
,  and  E.  Clark.    1947.   A  contribution  to  the  visceral  anatomy, 

development,  and  relationships  of  the  Plectognathi.  Bull.  Am.  Mus. 

Nat.  Hist.  88:287-320. 
,  and  P.  Rasquin.    1952.    The  sloughing  of  the  melanic  area  of 

the  dorsal  fin,  an  ontogenic  process  in  Tylosurus  raphidoma.  Bull. 

Am.  Mus.  Nat.  Hist.  99(1):  1-24. 
,  AND .    1954.    The  nature  of  post-larval  transformation 

in  Tylosurus  acus  (Lacepede).  Zoologica  (N.Y.)  39: 1  7-30. 
,  AND  D.  E.  Rosen.    1966.    Modes  of  reproduction  in  fishes.  T. 

F.  H.  Publ.,  Inc.,  Neptune  City,  New  Jersey. 
Brembach,  M.    1976.   Anatomische  Beitrage  zur  Systematik  lebendge- 

barender  Halbschnabler  (Hemirhamphidae,  Pisces).  Z.  Zool.  Syst. 

Evolutionsforsch.  14(3):  169-1 77. 
Briggs,  J.C.    1952.   Systematic  notes  on  the  oceanic  fishes  of  the  genus 

Lophotus.  Copeia  1952:206-207. 
.    1955.    A  monograph  of  the  clingfishes  (order  Xenopterygii). 

Stanford  Ichthyol.  Bull.  6:1-224. 
Brill,  J.    1981.    Cremchthys  baileyi  (GxXhen.  1983).  J.  Am.  Killifish 

Assoc.  14(5):172-184. 
Brinley,  F.  J.    1939.   Spawning  habits  and  development  of  beau-greg- 

ory  (Pomacenlrus  leucosliclus).  Copeia  1939:185-188. 
Brook,  G.    1890.    Notes  on  larval  stages  of  Afotella.  Proc.  R.  Phys. 

Soc.  Edinb.  10:156-161. 
Brothers,  E.  B.    1981.    What  can  otolith  microstructure  tell  us  about 

daily  and  subdaily  events  in  the  early  life  history  of  fish?  Rapp. 

P.V.  Reun.  Cons.  Int.  Explor.  Mer  178:393-394. 
,  C.  P.  Mathews  and  R.  Lasker.    1976.    Daily  growth  incre- 
ments in  otoliths  from  larval  and  adult  fishes.  Fish.  Bull.  U.S.  74: 

1-8. 
,  and  W.  N.  McFarland.    1981.    Correlations  between  otolith 


microstructure,  growth,  and  life  history  transitions  in  newly  re- 
cruited French  grunts,  [ffaemulon  /lavolincalum  (Desmarest),  Hae- 
mulidae].  Rapp.  P.V.  Reun.  Cons.  Int.  Explor.  Mer  178:369-374. 

— ,  E.  D.  Prince  AND  D.  W.  Lee.  1984.  Age  and  growth  of  young- 
of-the-year  bluefin  tuna  (Thunmis  thynnus)  from  otolith  micro- 
structure.  NOAA  Tech.  Rep.  8.  In  press. 

,  AND  R.  E.  Thresher.  MS.  Interspecific  differences  in  the  du- 
ration of  the  pelagic  larval  stage  of  Western  Atlantic  angelfishes 
(Pomacanthidae):  possible  biogeographic  and  systematic  applica- 
tions. 


LITERATURE  CITED 


695 


,  D.  M.  Williams  AND  P.  F.  Sale.    1983.    Length  of  larval  life 

in  12  families  of  fishes  at  One  Tree  Lagoon.  Great  Barrier  Reef, 

Australia.  Mar.  Biol.  (Berl.)  76:319-324. 
Brownell,  C.  L.    1979.   Stages  in  the  early  development  of  40  marine 

fish  species  with  pelagic  eggs  from  the  Cape  of  Good  Hope.  Rhodes 

Univ.  J.  L.  B.  Smith  Inst.  Ichthyol.  Bull.  40. 
.    1980a.   Water  quality  requirements  for  first-feeding  in  marine 

fish  larvae.  1.  Ammonia,  nitrite,  and  nitrate.  J.  Exp.  Mar.  Biol.  Ecol. 

44:269-283. 

1 980b.   Water  quality  requirements  for  first-feeding  in  marine 


fish  larvae.  II.  pH,  oxygen,  and  carbon  dioxide.  J.  Exp.  Mar.  Biol. 
Ecol.  44:285-298. 

Bruggeman.  E.,  P.  Sorgeloos  AND  P.  Van  Haecke.  1980.  Improve- 
ments in  the  decapsulation  techniqueof.-Jr/e?«;a  cysts,  p.  261-269. 
In:  The  brine  shrimp  Artemia.  Vol.  3.  Ecology,  culturing,  use  in 
aquaculture.  G.  Persoone,  P.  Sorgeloos,  O.  Roels,  and  E.  Jaspers 
(eds.).  Universa  Press,  Wetteren,  Belgium. 

Brummett.  a.  R.,  and  J.  N.  Dumont.  1981.  A  comparison  of  cho- 
rions from  eggs  from  northern  and  southern  populations  of  Fun- 
dulus  helerocUtus.  Copeia  1981:607-614, 

Brundin,  L.  1966.  Transantarctic  relationships  and  their  significance 
as  evidenced  by  chironomid  midges.  K.  Sven.  Vetenskapsakad. 
Handl.  4(11)1-472. 

Bri'Ton,  M,  N.  1979.  The  breeding  biology  and  early  development 
of  Clanas  gariepinus  (Pisces:  Clariidae)  in  Lake  Sibaya,  South  Af- 
rica, with  a  review  of  breeding  in  species  of  the  subgenus  Clarias 
(Clarias).  Trans.  Zool.  Soc.  Lond.  35:1-45. 

Bruun,  A.  F.  1935.  Flying-fishes  (Exocoetidae)  of  the  Atlantic.  Dana- 
Rap.  Carlsberg  Found.  6. 

.  1937.  Chascanopselta  in  the  Atlantic;  a  bathypelagic  occur- 
rence of  flatfish  with  remarks  on  distribution  and  development  of 
certain  other  forms.  Vidensk.  Medd.  Dan.  Naturhist.  Foren.  101: 
125-135. 

.    1 940.  A  study  of  a  collection  of  the  fish  Schindlena  from  south 

Pacific  waters.  Dana-Rep.  Carlsberg  Found.  21. 

Bryan,  P.  G.,  and  B.  B.  Madraisall  1977.  Larval  rearing  and  de- 
velopment of  Siganus  Imeatus  (Pisces:  Siganidae)  from  hatching 
through  metamorphosis.  Aquaculture  10:243-252. 

BiDD,  P.  L.  1940.  Development  of  the  eggs  and  early  larvae  of  six 
California  fishes.  Calif  Dep.  Fish  Game  Fish.  Bull.  56. 

Budgett,  J.  S.  1902.  The  habits  and  development  of  some  West 
African  fishes.  Proc.  Camb.  Philos.  Soc.  1 1:102-104. 

.    1903.    On  the  breeding-habits  of  some  West-African  fishes. 

with  an  account  of  the  external  features  in  development  o{  Prolop- 
lerus  annectens.  and  a  description  of  the  larva  of  Prowpterus  la- 
pradei.  Trans.  Zool.  Soc.  Lond.  16:1  15-136. 

Burdak,  V.  D.  1969.  Ontogenesis  of  squamation  in  the  mullet  (Mugil 
saliens  Risso).  Zool.  Zh.  48:241-247. 

Burgess.  W.  E.  1965.  Larvae  of  the  surgeonfish  genus  .-lfa«//n«-«i  of 
the  tropical  western  Atlantic.  Unpubl.  MS  thesis.  Univ.  Miami, 
Coral  Gables,  FL. 

.    1974.    Evidence  for  the  elevation  to  family  status  of  the  an- 

gelfishes  (Pomacanthidae),  previously  considered  to  be  a  subfamily 
of  the  butterfly-fish  family,  Chaetodontidae.  Pac.  Sci.  28:57-71. 

.     1978.    Butterflyfishes  of  the  worid.  TFH  Publications,  Ltd. 

Neptune  City,  NJ. 

.    1982.    The  first  aquarium  spawning  of  the  woodcat,  Trachy- 

corystes  insignis.  Trop.  Fish  Hobby.  1982  (August):84-89. 

Burke,  V.  1930.  Revision  of  the  fishes  of  the  family  Liparidae.  U.S. 
Natl.  Mus.  Bull.  150:1-204. 

Bussing,  W.  A.  1965.  Studiesof  the  midwater  fishes  of  the  Peru-Chile 
Trench.  Antarct.  Res.  Ser.  5:185-227. 

Bitler,  J.  L.  1979.  The  nomeid  genus  Cubiceps  (Pisces)  with  a  de- 
scription of  a  new  species.  Bull.  Mar.  Sci.  29:226-241. 

,  and  E.  H.  Ahl-STROM.    1976.   Review  of  the  deep-sea  fish  genus 

Scopetengys  (Neoscopelidae)  with  a  description  of  a  new  species. 
Scopelengys  clarkei,  from  the  central  Pacific.  Fish.  Bull.  U.S.  74: 
142-150. 

,  H.  G.  Moser,  G.  S.  Hageman  and  L.  L.  Nordgren.    1982. 


Developmental  stages  of  three  California  sea  basses  (Paralabrax. 
Pisces,  Serranidae).  Calif  Coop.  Oceanic  Fish.  Invest.  Rep.  23:252- 
268. 

Byrne,  D.  M.  1978.  Life  history  of  the  spotfin  killifish.  Fundutus  luciae 
(Pisces:  Cypnnodontidae),  in  Fox  Creek  Marsh,  Virginia.  Estuaries 
1:211-227. 

Cadwallader,  p.  L.  1976.  Breeding  biology  of  a  non-diadromous 
galaxiid  Galaxias  vulgaris  Stokell  in  a  New  Zealand  river.  J.  Fish 
Biol.  8:157-177. 

Cailliet,  G.  M.,  and  R.  N.  Lea.  1977.  Abundance  of  the  "rare" 
zoarcid,  Maynea  californica  Gilbert.  1915,  in  the  Monterey  Can- 
yon. Monterey  Bay,  California.  Calif  Fish  Game  63:253-261. 

Caldwell,  D.  K.  1962.  Development  and  distribution  of  the  short 
bigeye,  Pseudopriacanlhus  alius  (Gill)  in  the  western  North  Atlan- 
tic. U.S.  Fish  Wildl.  Serv.  Fish.  Bull.  62(203):  103-150. 

Caldwell,  M.  C.  1962.  Development  and  distnbution  of  larval  and 
juvenile  fishes  of  the  family  Mullidae  of  the  western  North  Atlantic. 
U.S.  Fish  Wildl.  Serv.  Fish.  Bull.  62(213):403-457. 

Callegrini,  C,  and  F.  Basaglia.  1978.  Biochemical  characteristics 
of  Mugilids  in  the  lagoons  of  the  Po  Delta.  Boll.  Zool.  45(1):35- 
40. 

Cameron,  J.  1959.  The  larval  and  post-larval  stages  of  Gymnam- 
modytes  semisquamatus  (Jourdain).  J.  Mar.  Biol.  Assoc.  U.K.  38: 
17-25. 

Campana,  S.  E.  1983.  Calcium  deposition  and  otolith  check  formation 
during  periods  of  stress  in  coho  salmon.  Oncorhynchus  kisutch. 
Comp.  Biochem.  Physiol.  75A:2 15-220. 

Campos,  H.  1 969.  Reproduccion  del  Aplochiton  laematus  Jenyns.  Bol. 
Mus.  Nac.  Hist.  Nat.  Chile  29:207-222. 

.    1972.    Breeding  season  and  early  development  of  i?rac/n'^a- 

laxias  bullock:  (Osteichthyes:  Galaxiidae).  Tex.  J.  Sci.  23:53 1-544. 

.    1979.   Multivariate  analysis  of  the  taxonomy  of  the  fish  family 

Galaxiidae.  Zool.  Anz.  202:280-288. 

Cantwell,  G.  E.  1964.  A  revision  of  the  genus  Parapercis.  family 
Mugiloididae.  Pac.  Sci.  18:239-280. 

Carbine.  W.  F.  1 944.  Egg  production  of  the  northern  pike.  Esox  lucius 
L.,  and  the  percentage  survival  of  eggs  and  young  on  the  spawning 
grounds.  Pap.  Mich.  Acad,  of  Sci.  Arts  Lett.  29:123-137. 

Cardeilhac,  p.  T.  1976.  Induced  maturation  and  development  of 
pinfish  eggs.  Aquaculture  8:389-393. 

Carlstrom,  D.  1963.  A  crystallographic  study  of  vertebrate  otoliths. 
Biol.  Bull.  (Woods  Hole)  125:441-463. 

Carr,  J.  1982.  Aphyosemion  ainieli  Radda.  1976.  J.  Am.  Killifish 
Assoc.  15(3):  102- 104. 

Carr,  W.  E.  S.,  and  C.  A.  Adams.  1972.  Food  habits  of  juvenile 
marine  fishes:  evidence  of  the  cleaning  habit  in  the  leatherjacket, 
Oligoplites saunis.  and  the  spottail  pinfish,  Diplodus  holbronki.  Fish 
Bull.  U.S.  70:1111-1120. 

Caruso,  J.  H.  1981.  The  systematics  and  distribution  of  the  lophiid 
anglerfishes:  I.  A  revision  of  the  genus  Lophioides  with  the  descrip- 
tion of  two  new  species.  Copeia  1981:522-549. 

.  1983.  The  systematics  and  distribution  of  the  lophiid  angler- 
fishes:  II.  Revisions  of  the  genera  Lophiomus  and  Lophius.  Copeia 
1983:11-30. 

,  AND  H.  R.  BiiLLis,  Jr.    1976.    A  review  of  the  lophiid  angler 


fish  genus  Sladenia  with  a  description  of  a  new  species  from  the 
Caribbean  Sea.  Bull.  Mar.  Sci.  26:59-64. 

Cassel,  a.  1 98 1 .  Garmanella pulchra.  a  fascinating  killifish  from  Mex- 
ico. J.  Am.  Killifish  Assoc.  1 4(5):  1 67- 1 68. 

Casteel,  R.  W.  1974.  Identification  of  the  species  of  Pacific  salmon 
(Genus  Oncorhynchus)  native  to  North  .America  based  on  otoliths. 
Copeia  1974:305-311. 

Castle,  P.  H.  J.  1968.  Larvaldevelopment  ofthecongrid  eel  G«ar/io- 
phis  capensis  (Kaup),  ofl^ southern  Africa,  with  notes  on  the  identity 
of  Congermuraena  auslralis  Barnard.  Zool.  Afr.  3:139-154. 

.    1969.   An  index  and  bibliography  of  eel  larvae.  Rhodes  Univ. 

J.  L.  B.  Smith  Inst.  Ichthyol.  Spec.  Publ.  7:1-121. 

.    1973.    A  giant  notacanthiform  leptocephalus  from  the  Chat- 


696 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


ham  Islands,  New  Zealand.  Rec.  Dom.  Mus.  Wellington  8(8):  121- 

124. 
.    1976.    Criteria  of  satisfactory  fixation  and  preservation  of  eel 

larvae  (Leptocephali),  p.  319-321.  In:  Zooplankton  fixation  and 

preservation.  H.  F.  Steedman  (ed.).  Unesco  Press,  Pans. 
.    1979.    Early  life-history  of  the  eel  Moringua  edwardsi  (Pisces, 

Moringuidae)  in  the  western  North  Atlantic.  Bull.  Mar.  Sci.  29:1- 

18. 
.    1980.    Larvae  of  the  ophichthid  eel  genus  Necnchelys  in  the 

Indo-Pacific.  Pac.  Sci.  34:165-171. 
,  .AND  J.  E.  BoHLKE.    1976.   Sexual  dimorphism  in  sizc  and  ver- 
tebral number  in  the  western  Atlantic  eel  Moringua  edwardsi  (An- 

guilliformes,  Moringuidae).  Bull.  Mar.  Sci.  26:615-619. 
,  AND  N.  S.  Raju.    1975.    Some  rare  leptocephali  from  the  At- 
lantic and  Indo-Pacific  oceans.  Dana-Rep.  Carlsberg  Found.  85:1- 

25. 
,  AND  D.  A.  Robertson.    1974.    Early  life  history  of  the  congrid 

eels  Gnalhophis  hahenatus  and  G.  incogniliis  in  New  Zealand  waters. 

N.Z.  J.  Mar.  Freshwater  Res.  8:95-1  10. 
Cavender,T.  M.    1969.  An  Oligocenemudminnow  (family  Umbridae) 

from  Oregon  with  remarks  on  relationships  within  the  Esocoidei. 

Occas.  Pap.  Mus.  Zool.  Univ.  Mich.  No.  660. 
.    1970.   A  comparison  of  coregonines  and  other  salmonids  with 

the  earliest  known  teleost  fishes,  p.  1-32.  In:  Biology  of  coregonid 

fishes.  C.  C.  Lindsey  and  C.  S.  Woods  (eds.).  Univ.  Manitoba  Press, 

Winnipeg. 
Cerny,  K.    1977.   The  early  development  of  chub  LeMmcM.5  cep/ia/iu 

(L.,  n5S),rudd-Scardiniuserythrophthalmus{L.,  1758)and  roach- 

RutilusruulusCL.,  1758).  Acta  Univ.  Carol.  Biol.  1977(1-2):  1-149. 
Cervigon  M.,  F.    1966.    Los  peces  marinos  de  Venezuela.  Fund.  La 

Salle  Cienc.  Nat.  (Caracas),  Vol.  I  and  2. 
,  and  E.  Velazquez.     1978.    Revision  systematica  y  estudio 

osteologico  de  las  especies  del  genero  Jenkinsia  de  las  costas  de 

Venezuela  (Osteichthyes:  Clupeiformes).  Cuad.  Oceanogr.  Univ. 

Onente  7:3-54. 
Chabanaud,  p.    1934.    Contribution  a  la  morphologic  des  poissons 

heterosomes.  Bull.  Soc.  Zool.  Fr.  59:123-129. 
.    1937.    L'extension  prorse  peripherique,  la  contraction  axiale 

post-hypophysaire  et  I'anisoconie  rhachidienne  des  teleosteens  dys- 

symetriques.  Bull.  Soc.  Zool.  Fr.  62:368-385. 
.    1940.    Notules  ichtyologiques.  Bull.  Mus.  Natl.  Hist.  Nat.  (2) 

12:149-156. 
.    1941.   Sur  la  classification  et  la  geonemie  de  soleides  du  genre 


— .    1948.   The  osteology  and  relationships  of  the  Microstomidae, 

a  family  of  oceanic  fishes.  Proc.  Calif  Acad.  Sci.,  Ser.  4,  26:1-22. 

— ,  andL.  D.  TowNSEND.    1938.  1\te  o^leoXogy  of  Zaprora  silenus. 


Aesopia.  Copeia  1941:31-32. 

.    1949.   Le  probleme  de  la  phylogenese  des  Heterosomata.  Bull. 

Inst.  Oceanogr.  (Monaco)  950:1-24. 

.     1950.    Contribution  a  I'anatomie  et  a  la  systematique  de  la 

famine  des  Bothidae,  s.  str.  Bull.  Soc.  Zool.  Fr.  74:246-253. 

.     1969.    Sur  la  morphologie  de  Samans  cristalus  Gray  et  de 

Samaris  cristalus  erythraeus  P.  Chabanaud  natio  nova  (Pleuronec- 
toidea  Samaridae).  Bull.  Mus.  Natl.  Hist.  Nat.  (2)  40:874-890. 

Chacko,  p.  I.  1944.  On  the  bionomics  of  the  big-jawed  jumper,  Lac- 
tarius  lactarius  (Cuv.  and  Val.).  Curr.  Sci.  (Bangalore)  13(4):  108. 

Chambers,  J.  R.,  J.  A.  MusicK  AND  J.  Davis.  1976.  Methods  of  dis- 
tinguishing larval  alewife  from  larval  blueback  herring.  Chesapeake 
Sci.  17:93-100. 

Chang,  H.,  X.Sha,Z.  Chen,  H.  Rl  an  and  G.  He.  1981.  Observations 
on  the  egg  and  larval  development  of  a  clupeid  Harengula  zunasi 
Bleeker.  Trans.  Chin.  Ichthyol.  Soc.  1:57-64. 

, ,  G.  He  AND  L.  Song.  1980.  A  description  of  the  mor- 
phological characters  of  the  eggs  and  larvae  of  the  flathcad  fish, 
Ptatycephalus  indicus.  Oceanol.  Limnol.  Sin.  1 1(2):I61-168. 

Changjiang,  Y.  1976.  Fishes  of  the  Yangtze  River.  Hunan  Province 
Inst.  Hydrobiol.,  Science  Press,  Peking.  (In  Chinese.) 

Chapman,  W.  M.  1939.  Eleven  new  species  and  three  new  genera  of 
oceanic  fishes  collected  by  the  International  Fisheries  Commission 
from  the  northeastern  Pacific.  Proc.  U.S.  Natl.  Mus.  86:501-542. 

.    1942.   The  osteology  and  relationships  of  the  Argentinidae,  a 

family  of  oceanic  fishes.  J.  Wash.  Acad.  Sci.  32:104-1 17. 


with  notes  on  its  distribution  and  early  life  history.  Ann.  Mag.  Nat. 
Hist.,  Ser.  11,  2:89-117. 

Chaudhuri,  H.  1962.  Induced  breeding  and  development  of  a  com- 
mon catfish,  Ornpok  bimaculatus  (Bloch).  Proc.  Indian  Sci.  Congr. 
49:390-391. 

,  J.  V.  JuARio,  J.  H.  Primavera,  R.  Samson  and  R.  Mateo. 

1978.  Observations  on  artificial  fertilization  of  eggs  and  the  em- 
bryonic and  larval  development  of  milkfish,  Chanos  chanos  (For- 
skal).  Aquaculture  13:95-1  13. 

Chen,  F.  Y.,  M.  Chow,  T.  M.  Cha^)  and  R.  Lim.  1977.  Artificial 
spawning  and  larval  rearing  of  the  grouper,  Epinephelus  tauvina 
(Forskal)  in  Singapore.  Singapore  J.  Primary  Ind.  5:1-21. 

Chenoweth,  S.  B.  1970.  Seasonal  variations  in  condition  of  larval 
herring  in  Boothbay  Harbor  area  of  the  Maine  coast.  J.  Fish.  Res. 
Board  Can.  27:1875-1879. 

Chereshnev,  I.  A.,  AND  A.  V.  Balushkin.  1980.  A  new  species  of 
blackfish,  Dallia  asmirabilis  sp.  n.  (Umbridae,  Esociformes)  from 
the  Amguema  River  basin  (Arctic  Chukotka).  J.  Ichthyol.  (Engl. 
Transl.  Vopr.  Ikhtiol.)  20(6):25-30. 

Chernyayev,  Zh.  a.  1971.  Some  data  on  the  reproduction  and  de- 
velopment of  the  cottoid  fish  (Comephorus  dyhowskii  Korotneff). 
J.  Ichthyol.  (Engl.  Transl.  Vopr.  Ikhtiol.)  1  1:706-716. 

.  1975.  Morphological  and  ecological  features  of  the  reproduc- 
tion and  development  of  the  "Big  Golomyanka"  or  Baikal  oil-fish 
(Comephorus  baicalensis).  J.  Ichthyol.  (Engl.  Transl.  Vopr.  Ikhtiol.) 
14:856-868. 

.    1978.    Morpho-ecological  characteristics  of  the  reproduction 

and  development  of  the  Baikal  sand  sculpin,  Paracoltus  {Leocoltus) 
kessleri.  J.  Ichthyol.  (Engl.  Transl.  Vopr.  Ikhtiol.)  17:890-903. 

.  1981.  Reproduction  and  development  ofthe  "Big-head"  scul- 
pin, Balrachocollus  baicalensis.  from  Lake  Baikal.  J.  Ichthyol.  (Engl. 
Transl.  Vopr.  Ikhtiol.)  l9(6):83-94. 

Chidambaram,  K.  1942.  Observations  on  the  development  of  Anus 
jella  (Day).  Proc.  Indian  Acad.  Sci.  Sect.  B  14:502-508. 

Chkhrinskiv,  a.  I.  1970.  The  nature  of  oogenesis  and  fecundity  in 
the  Japanese  horsemackerel  [Trachurus  japonicus  (Temminck  et 
Schlegel)].  J.  Ichthyol.  (Engl.  Transl.  Vopr.  Ikhtiol.)  10:755-760. 

Chirichigno,  N.  1973.  Nuevas  especies  de  peces  de  los  generos  A/ui- 
lelus  (fam.  Triakidae),  Raja  (fam.  Rajidae)  y  Schedophilus  (fam. 
Centrolophidae).  Inst.  Mar.  Peru  (Callao)  Inf  42. 

Chrapkova-Kovalevskaya,  N.  V.  1963.  Data  on  reproduction,  de- 
velopment and  distribution  of  larvae  and  young  fish  oi  Oxyporhani- 
phus  micropterus  Val.  (Pisces,  Oxyporhamphidae).  Tr.  Inst.  Okean- 
ol.  Acad.  Nauk  SSSR  62:49-61. 

Chung,  M.  1961.  Illustrated  encyclopedia.  The  fauna  of  Korea  (2). 
Fishes. 

.    1977.   The  fishes  of  Korea.  II  Ji  Sa  Publishing  Co.,  Seoul. 

CiPRiA.  G.  1927.  Contributo  alia  conoscenza  dello  sviluppo  postem- 
brionale  di  Hypsirhynchus  hepalicus  Facciola.  Mem.  R.  Com.  Tal- 
assogr.  Ital.  130. 

.    1934.    Uova,  stadi  embrionali  e  post-embrionali  di  Blennius 

palmicornis  C\i\ .  Mem.  R.  Com.  Talassogr.  Ital.  218. 

.    1936.   Uova,  stadi  embrionali  e  post -embnonali,  nei  Blennidi. 

I.  Blennius pavo  Risso.  II.  Blennius  maequalis  C.  V.  Mem.  R.  Com. 
Talassogr.  Ital.  231. 

Clancey,  J.  F.  1956.  A  contribution  to  the  life  history  of  the  fish, 
Bregmaceros  atlanticus  Goode  and  Bean,  from  the  Florida  Current. 
Bull.  Mar.  Sci.  Gulf  Camb.  6:233-260. 

Clark,  E.  1950.  Notes  on  the  behavior  and  morphology  of  some  West 
Indian  plectognath  fishes.  Zoologica  (N.Y.)  35:159-168. 

Clark,  F.  N.  1925.  The  life  history  of  Leurest ties  tenuis,  an  atherine 
fish  with  tide  controlled  spawning  habits.  Calif  Dep.  Fish  Game 
Fish  Bull.  10:1-51. 

.    1929.  The  life  history  of  the  California  jack  smelt,  Atherinopsis 

californiensis.  Calif.  Dep.  Fish  Game  Fish  Bull.  16:1-22. 

Clark,  J.,  W.  G.  Smith,  A.  W.  Kendall,  Jr.  and  M.  P.  Fahav.    1 969. 


LITERATURE  CITED 


697 


Studies  of  estuarine  dependence  of  Atlantic  coastal  fishes.  U.S.  Bur. 

Sport  Fish.  Wild!.  Tech.  Pap.  28. 
Clark,  R.  S.    1914.    General  report  on  the  larval  and  post-larval  te- 

leosteans  in  Plymouth  waters.  J.  Mar.  Biol.  Assoc.  U.K.  10:327. 
.    1920.  The  pelagic  young  and  early  bottom  stages  of  teleosteans. 

J.  Mar.  Biol.  Assoc.  U.K.  12:159-240. 
Clakke,  T.  a.     1973.    Some  aspects  of  the  ecology  of  lantemfishes 

(Myctophidae)  in  the  Pacific  Ocean  near  Hawaii.  Fish.  Bull.  U.S. 

71:401-434. 
,  AND  P.  J.  Wagner.     1976.    Vertical  distribution  and  other 

aspects  of  the  ecology  of  certain  mesopelagic  fishes  taken  near  Ha- 
waii. Fish  Bull.  U.S.  74:635-645. 
Clausen,  H.  S.    1959.    Denticipitidae,  a  new  family  of  isospondylous 

teleosls  from  west  African  freshwater.  Vidensk.  Medd.  Dan.  Nat- 

urhist.  Foren.  121:141-156. 
Cohen,  D.  M.    1958.   A  revision  of  the  fishes  of  the  subfamily  Argen- 

tininae.  Bull.  Fl.  State  Mus.  Biol.  Ser.  3:93-172. 
.    1 960.   New  records  of  the  opisthoproctid  genus  Balhylychnops. 

with  a  notice  of  neoteny  in  the  related  genus  DoUchopleryx.  Copeia 

1960:147-149. 
.     1964a.    Bioluminescence  in  the  Gulf  of  Mexico  anacanthine 

fish  Steindachnena  argentea.  Copeia  1964:406-409. 
.     1964b.    Suborder  Argentmoidea.  p.   1-70.  In:  Fishes  of  the 

western  North  Atlantic.  H.  B.  Bigelow  (ed.).  Mem.  Sears  Found. 

Mar.  Res.  No.  1  (Pt.  4). 
.    1973.   Eretmophoridae,  p.  322-326. /«.■  Check-list  of  the  fishes 

of  the  north-eastern  Atlantic  and  of  the  Mediterranean  (Clofnam). 

Vol.  I.  J.  C.  Hureau  and  T.  Monod  (eds.).  UNESCO,  Pans. 
.    1974.    A  review  of  the  pelagic  ophidioid  fish  genus  Brotula- 

taenia  with  descriptions  of  two  new  species.  Zool.  J.  Linn.  Soc.  55: 

119-149. 
.    1979.    Notes  on  the  mond  fish  genera  L()?e//a  and  P/!.vi;cM/j« 

in  Japanese  waters.  Jpn.  J.  Ichthyol.  26:225-230. 
,  ANn  J.  G.  Nielsen.    1978.   Guide  to  the  identification  of  genera 

of  the  fish  order  Ophidiiformes  with  a  tentative  classification  of  the 

order.  NOAA  Tech.  Rep.  NMFS  Circ.  147. 
,  and  J.  L.  Russo.    1979.    Variations  in  the  fourbeard  rockling. 


Enchelyopus  cimhrius.  a  North  Atlantic  gadid  fish,  with  comments 
on  the  genera  of  rocklings.  Fish.  Bull.  U.S.  77:91-104. 

Cole,  F.  J.,  and  J.  Johnstone.  1902.  Pleuronecles  (the  plaice).  Mem. 
Liverpool  Mar.  Biol.  Comm.  8. 

Coleman,  L.  R.,  and  B.  G.  Nafpaktitis.  1972.  Dorsadcna  yaqiunae. 
a  new  genus  and  species  of  myctophid  fish  from  the  eastern  nonh 
Pacific  Ocean.  Nat.  Hist.  Mus.  Los  Ang.  Cty.  Contrib.  Sci.  225:1- 
11. 

Colin,  P.  L.  1975.  The  neon  gobies.  The  comparative  biology  of  the 
gobies  of  the  genus  Gobwsoma.  subgenus  Elacatmus.  (Pisces:  Go- 
biidae)  in  the  tropical  western  north  Atlantic  Ocean.  T.F.H.  Pub- 
lications, Neptune  City,  NJ. 

.    1982.   Spawning  and  larval  development  of  the  hogfish,  Lflc/;- 

nolaimus  maximus.  (Pisces,  Labridae).  Fish.  Bull.  U.S.  80:853- 
862. 

CoLLETT,  R.  1905.  Fiske  indsamlede  under  "Michael  Sars"  togter  i 
nordhavet  1900-1902.  Rep.  Norw.  Fish.  Mar.  Invest.  2(3):1-I47. 

Coi  LETTE,  B.  B.  1962.  Hemiramphus  bermudensis.  a  new  halfbeak 
from  Bermuda,  with  a  survey  of  endemism  in  Bermudian  shore 
fishes.  Bull.  Mar.  Sci.  Gulf  Canbb.  12:432-449. 

.    1966.    Betomon.  a  new  genus  of  fresh-water  needlefishes  from 

South  America.  .Am.  Mus.  Novit.  No.  2274. 

.     1974a.    South  American  freshwater  needlefishes  (Belonidae) 

of  the  genus  Pseudolylosunis.  Zool.  Meded.  (Leiden)  48(16):I69- 
186. 

.    1974b.   The  garfishes  (Hemiramphidae)  of  Australia  and  New 

Zealand.  Rec.  Aust.  Mus.  29(2):  1 1-105. 

.  1974c.  Sfrowgi'/wra /iififci.  a  new  species  of  freshwater  needle- 
fish from  the  Usumacinta  Province  of  Guatemala  and  Mexico. 
Copeia  1974:611-619. 

.    1979.  Adaptations  and  systematics  of  the  mackerels  and  tunas. 


p.  7-39.  In:  The  physiological  ecology  of  tunas.  G.  Sharp  and  A. 

Dizon  (eds.).  Academic  Press,  NY. 
— .    1982a.    South  American  freshwater  needlefishes  of  the  genus 

Polamorrhaphis  (BelomfOTmes:  Belonidae).  Proc.  Biol.  Soc.  Wash. 

95:714-747. 
— .    1982b.    Rediscovery  of  Hyporhamphus  xanthoptenis.  a  half- 
beak  endemic  to  Vembanad  Lake,  Kerala,  southern  India.  Matsya 

7:1-12. 
— .    1982c.   Two  new  species  of  freshwater  halfbeaks  (Pisces:  Hem- 
iramphidae) of  the  genus  Zenarchoplerus  from  New  Guinea.  Copeia 

1982:265-276. 
— .    1983.    Recognition  of  two  species  of  double-lined  mackerels 

(Grammatorcynus:  Scombridae).  Proc.  Biol.  Soc.  Wash.  96:715- 

718. 
— ,  and  F.  H.  Berry.    1965.    Recent  studies  on  the  needlefishes 

(Belonidae):  An  evaluation.  Copeia  1965:386-392. 
— ,  AND  L.  N.  Chao.    1975.    Systematics  and  morphology  of  the 

bonilos  (Sarda)  and  their  relatives  (Scombridae.  Sardini).  Fish. 

Bull.  U.S.  73:516-625. 
— ,  andN.  V.  Parin.   1970.  Needlefishes  (Belonidae)  of  the  eastern 

Atlantic  Ocean.  Atl.  Rep.  1  1:7-60. 
— ,  AND  J.  L.  Rlsso.    1979.   An  introduction  to  the  Spanish  mack- 


erels, genus  Scomberomonis.  p.  3-16.  In:  Proc.  Mackerel  Collo- 
quium. E.  L.  Nakamura  and  H.  R.  Bullis,  Jr.  (eds.).  Gulf  States 
Mar.  Fish.  Comm. 

,  AND .    In  press.    Morphology,  systematics.  and  biology 

of  the  Spanish  mackeTeh(,Scomberomorus.  Scombridae).  Fish.  Bull. 
U.S. 

Collins,  L.  A.,  J.  H.  Finucane  and  L.  E.  Barger.  1980.  Description 
of  larval  and  juvenile  red  snapper,  Lutjanus  campechanus.  Fish. 
Bull.  U.S.  77:965-974. 

Collins,  S.  P.,  and  M.  P.  Barron.  1981.  Demersal  and  pelagic  trawl- 
ing survey  of  the  M.T.  "Denebola"  in  southern  Australian  waters, 
1979-80  summer.  Tasmanian  Fish.  Res.  24:2-48. 

Colton,  J.  B.,  AND  K.  A.  Honey.  1963.  The  eggs  and  larval  stages  of 
the  butterfish,  Poronotus  triacanlfius.  Copeia  1963:447-450. 

,  AND  R.  R.  Marak.    1969.    Guide  for  identifying  the  common 

planktonic  fish  eggs  and  larvae  of  continental  shelf  waters.  Cape 
Sable  to  Block  Island.  NMFS  Woods  Hole  Biol.  Lab.  Ref.  Doc. 
69-9. 

CoNAND,  F.  1978,  Systematique  des  larves  de  Clupeides  de  PAtlan- 
tique  oriental  entre  20°N  et  15°S  (eaux  marine  et  saumatres).  Cah. 
ORSTOM  Ser.  Oceanogr.  16:3-8. 

,  AND  E.  Fagetti.    1971.   Description  et  distribution  saisonniere 

des  larves  de  sardinelles  des  cotes  du  Senegal  et  de  la  Gamble  en 
1968  et  1969.  Cah.  ORSTOM  Ser.  Oceanogr.  9:293-318. 

,  AND  C.  Franqlieville.     1973.    Identification  et  distribution 

saisonniere  de  larves  de  Carangides  au  large  du  Senegal  et  de  la 
Gamble.  Bull.  Inst.  Fondam.  Afr.  Noire,  Ser.  A  Sci.  Nat.  35:951- 
978. 

CoNSTANTZ,  G.  D.  1981.  Life  history  patterns  of  desert  fishes,  p.  237- 
290.  In:  Fishes  in  North  American  deserts.  R.  J.  Naiman  and  D. 
L.  Soltz  (eds.).  John  Wiley  &  Sons.  N.Y. 

Coombs,  S.  H.,  and  A.  R.  Hiby.  1979.  The  development  of  the  eggs 
and  early  larvae  of  blue  whiting,  Micromesistius  poiassou  and  the 
effect  of  temperature  on  development.  J.  Fish  Biol.  14:1 1 1-123. 

Cooper,  J.  E.  1978.  Identification  of  eggs,  larvae,  and  juveniles  of  the 
rainbow  smelt,  Osmerus  morda.x.  with  comparisons  to  larval  ale- 
wife,  Alosa  pseudoharengus,  and  gizzard  shad.  Durosoma  cepe- 
dianum.  Trans.  Am.  Fish.  Soc.  107:56-62. 

Costa,  A.  1869.  Osservacioni  sul  Krohnius  filamentosus  e  sullo  svi- 
luppodella  pinna caudale del  Trachyptenis.  .Ann.  Mus.  Zool.  Napoli 
5:1-41. 

Courtenay,  W.  R.,  Jr.  1971.  Sexual  dimorphism  of  the  sound  pro- 
ducing mechanism  of  the  stnped  cusk-eel,  Rissola  marginala  (Pisces: 
Ophidiidae).  Copeia  1971:259-268. 

,  and  F.  A.  McKiTTRicK.    1970.   Sound-producing  mechanisms 

in  carapid  fishes,  with  notes  on  phylogenetic  implications.  Mar. 
Biol.  (Beri.)  7:131-137. 


698 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Cressey,  R.  1981.  Revision  of  Indo-Pacific  lizardfishes  of  the  genus 
Synodus  (Pisces:  Synodontidae).  Smithson,  Contrib.  Zool.  342. 

Cressey,  R.  P.,  B.  B.  CoLLETTE  AND  J.  L.  Rlisso.  1983.  Copepodsand 
scombrid  fishes:  A  study  in  host-parasite  relationships.  Fish.  Bull. 
U.S.  81:227-265. 

Crossland,  J.  1981.  Fish  eggs  and  larvae  of  the  Hauraki  Gulf  New 
Zealand.  N.Z.  Minist.  Agric.  Fish.  Res.  Bull.  23. 

• .    1982.   Distribution  and  abundance  of  fish  eggs  and  larvae  from 

the  spring  and  summer  plankton  of  north-east  New  Zealand,  1976- 
78.  N.Z.  Minist.  Agric.  Fish.  Res.  Bull.  24. 

Grossman,  E.  J.  1980.  Esox  amencanus.  Esox  lucius,  Esox  inasqui- 
nongy.  Esox  niger.  pp.  1 31-138.  In:  Atlas  of  North  American  fresh- 
water fishes.  D.  S.  Lee,  C.  R.  Gilbert,  C.  H.  Hocutt,  R.  E.  Jenkins, 
D.  E.  McAllister  and  Jay  R.  StaufTer,  Jr.  (eds.).  North  Carolina  State 
Mus.  Nat.  Hist.,  Raleigh,  NC. 

Cunningham,  J.  T.  1887.  The  eggs  and  larvae  of  teleosteans.  Trans. 
R.  Soc.  Edinb.  33:97-136. 

.    1889.    Studies  of  the  reproduction  and  development  of  te- 

leostean  fishes  occurring  in  the  neighbourhood  of  Plymouth.  J.  Mar. 
Biol.  Assoc.  U.K.  1:10-54. 

.  1890.  A  treatise  on  the  common  sole  {Solea  vulgaris),  con- 
sidered both  as  an  organism  and  as  a  commodity.  Mar.  Biol.  Assoc. 
U.K.,  Plymouth. 

.    1891.    On  some  larval  stages  of  fishes.  J.  Mar.  Biol.  Assoc. 

U.K.  2:68-74. 

Curless,  W.  W.  1979.  Early  development  and  larval  biology  of  the 
California  topsmelt,  Atherinops  affinis.  Amer.  Zool.  19:949.  (Ab- 
stract.) 

CusHiNG,  D.  H.  1975.  Marineecology  and  fisheries.  Cambridge  Univ. 
Press,  Cambridge. 

.  1977.  Science  and  the  fisheries.  Camelot  Press  Ltd.,  South- 
ampton. 

Cuvier,G.  1829.  DerOreosome(Oreo50wa,  nov.jetparticulierement 
de  rOreosome  de  I'Atlantique  (Oreosoma  atlanlicum.  nov.),  p.  515- 
518.  In:  Histoire  naturelledes  poissons.  4.  Des  Acanthopterygiens. 
G.  Cuvier  and  A.  Valenciennes  (eds.).  F.  G.  Levrault,  Paris. 

CzoLowsK.'V,  R.  1962.  Morphogenesis  of  skin  scuta  in  S/p/!o«osfo/«a 
typhle  L.  Zool.  Pol.  12(3):269-303. 

Daan,  N.  1981.  Comparison  of  estimates  of  egg  production  of  the 
Southern  Bight  cod  stock  from  plankton  surveys  and  market  sta- 
tistics. Rapp.  P.V.  Reun.  Cons.  Int.  Explor.  Mer  1  78:242-243. 

Dakin,  W.  J.,  AND  G,  L.  Kesteven.  1938.  The  Murray  cod  (Ma«-»- 
lochella  macquarensis  (Cuv.  et  Val.)).  Some  experiments  on  breed- 
ing with  notes  on  the  early  stages  and  a  reference  to  the  problems 
of  depletion  and  restocking.  State  Fish.  N.S.W.  Res.  Bull.  I. 

Dale,  T.  1984.  Embryogenesis  and  growth  of  otoliths  in  the  cod  (Gat/- 
us  morhua  L.),  p.  231-250.  In:  The  propagation  of  cod.  E.  Dahl, 
D.  S.  Danielssen,  E.  Moksness  and  P.  Solemdal  (eds.).  Flodevigen 
Rapp.  1. 

D'Ancona,  U.  1929.  Note  di  ittiologia  mediterranea.  1.  Stadi,  gio- 
vanili  e  uova  di  Exocoetus.  Pubbl.  Staz.  Zool.  Napoli  9:203-21 1. 

.    1931a.    Clupeidae,  p.  1-16.  In:  Uova,  larve  e  stadi  giovanili 

di  Teleostei.  Fauna  Flora  Golfo  Napoli  Monogr.  38. 

.    1931b.   Ordine:  Apodes  (Muraenoidei),  p.  94-156.  In:  Uova. 

larva  e  stadi  giovanili  di  Teleostei.  Fauna  Flora  Golfo  Napoli 
Monogr.  38. 

• .    1933a.    Famiglia  1:  Gadidae,  p.  178-255.  In:  Uova,  larve  e 

stadi  giovanili  di  Teleostei.  Fauna  Flora  Golfo  Napoli  Monogr.  38. 

.    1933b.    Famiglia:  Lampridae.  p.  280.  In:  Uova,  larve  e  stadi 


giovanili  di  Teleostei.  Fauna  Flora  Golfo  Napoli  Monogr.  38. 
— .    1933c.    Famiglia:  Syngnathidae,  p.  281-298.  In:  Uova,  larve 

e  stadi  giovanili  di  Teleostei.  Fauna  Flora  Golfo  Napoli  Monogr. 

38. 
— .    1933d.  Famiglia:  Macrorhamphosidae,  p.  299-306. /«.•  Uova, 

larve  e  stadi  giovanili  di  Teleostei.  Fauna  Flora  Golfo  Napoli 

Monogr.  38. 
— ,  AND  G.  Cavinato.    1965.    The  fishes  of  the  family  Bregma- 

cerotidae.  Dana-Rep.  Carlsberg  Found.  64. 


Dando,  p.  R.    1975.   On  the  northern  rockling  Ciliata  septentrionalis 

(Collett)  in  the  Plymouth  area.  J.  Mar.  Biol.  Assoc.  U.K.  55:925- 

931. 
Daniels,  R.  A.     1978.    Nesting  behaviour  of  Harpagifer  bispinis  in 

Arthur  Harbour,  Antarctic  Peninsula.  J.  Fish  Biol.  12:465-474. 
.    1981.    Cryoihenia  peninsulae,  a  new  genus  and  species  of  no- 

totheniid  fish  from  the  Antarctic  Peninsula.  Copeia  1981:558-562. 
Danil'chenko,  p.  G.    1947.   Phylogenelic  relations  between  the  genera 

Palaeogadus and  Merluccius.  Dokl.  Akad.  Nauk  SSSR  58:659-662. 

(In  Russian;  transl.  by  J.  H.  Slep,  1967,  U.S.  Bur.  Commer.  Fish., 

Washington,  D.C.,  Transl.  46.) 
.     1950.    The  genus  Palaeogadus  and  its  development.  Tr.  Pa- 

leontol.  Inst.  Akad.  Nauk  SSSR  25:3-25.  (In  Russian,  transl.  by  J. 

H.  Slep,  1968,  U.S.  Bur.  Commer.  Fish.,  Washington.  DC,  Transl. 

53.) 
.    1960.    Bony  fishes  of  the  Maikop  Deposits  of  the  Caucasus. 

Tr.  Paleontol.  Inst.  Akad.  Nauk  SSSR  78:1-247. 
Dannevig,  a.     1919.    Fish  eggs  and  larvae,  p.   1-80.   In:  Canadian 

fisheries  expedition  1914-15  in  the  Gulf  of  St.  Lawrence  and  At- 
lantic waters  of  Canada.  J.  Hjort  (ed.).  Can.  Nav.  Serv.,  King's 

Printer,  Ottawa. 
David,  A.    1961.    Notes  on  the  embryonic  and  larval  development  of 

the  "GooT\c\\."  —  Baganus  baganus  (Hamilton).  J.  Zool.  Soc.  India 

13:194-205. 
David,  L.  R.    1939.   Embryonic  and  early  larval  stages  of  the  grunion, 

Leuresthes  tenuis,  and  the  sculpin,  Scorpaena  guttata.  Copeia  1939: 

79-8 1 . 
Davis,  J.  R.,  and  D.  E.  Louder.    1969.    Life  history  of  ecology  of 

Menidia  extensa.  Trans.  Am.  Fish.  Soc.  98:466-472. 
Davis,  W.  P.    1966.   A  review  of  the  dragonets  (Pisces:  Callionymidae) 

of  the  western  Atlantic.  Bull.  Mar.  Sci.  16:834-862. 
Davy,  B.     1972.    A  review  of  the  lantemfish  genus  Taaningichthys 

(family  Myctophidae)  with  the  description  of  a  new  species.  Fish. 

Bull.  U.S.  70:67-78. 
Dawson,  C.  E.    1971a.   Occurrence  and  description  of  prejuvenile  and 

early  juvenile  Gulf  of  Mexico  cobia,  Rachycentron  canadum.  Co- 
peia 1971:65-71. 
.    1971b.    Records  of  the  pearlfish,  Carapus  bennudensis.  in  the 

Northern  Gulf  of  Mexico  and  of  a  new  host  species.  Copeia  1971: 

730-731. 
.    1973.    Indo-Pacific  distribution  of  microdesmid  fishes  (Go- 

bioidea).  J.  Mar.  Biol.  Assoc.  India  15(l):3l8-322. 
.    1 974a.   Studies  on  eastern  Pacific  sand  stargazers  (Pisces:  Dac- 

tyloscopidae).  1 .  Platygillelhis  new  genus,  with  descriptions  of  new 

species.  Copeia  1974:39-55. 
.    1974b.    A  review  of  the  Microdesmidae  (Pisces:  Gobioidea). 

1 .  Cerdale  and  Clarkichthys  with  descriptions  of  three  new  species. 

Copeia  1974:409-448. 
.    1975.   Studies  on  eastern  Pacific  sand  stargazers  (Pisces:  Dac- 

tyloscopidae).  2.  Genus  Dactyloscopus.  with  descriptions  of  new 

species  and  subspecies.  Nat.  Hist.  Mus.  Los  Ang.  Cty.  Sci.  Bull.  22. 
.    1976.    Studies  on  eastern  Pacific  sand  stargazers.  3.  Dactylag- 


nus  and  Myxodagnus.  with  description  of  a  new  species  and  sub- 
species. Copeia  1976:13-43. 
.    1982.   Atlantic  sand  stargazers  (Pisces:  Dactyloscopidae),  with 

descriptions  of  one  new  genus  and  seven  new  species.  Bull.  Mar. 

Sci.  32:14-85. 
,  F.  Yasuda  and  C.  Imai.    1979.    Elongate  dermal  appendages 

in  species  of  Yozia  (Syngnathidae)  with  remarks  on  Trachyrham- 

phus.  Jpn.  J.  Ichthyol.  25:244-250. 
Dean,  D.  M.    1972.    Osteology  and  systematics  of  the  echeline  worm 

eels  of  the  Atlantic  Ocean  (Pisces,  Anguilliformes).  Unpubl.  PhD 

dissertation.  Univ.  Miami,  Coral  Gables,  FL. 
Dearborn,  J.  H.     1965.    Reproduction  in  the  nototheniid  fish  Tre- 

matoinus  bernacchii  Boulenger  at  McMurdo  Sound,  Antarctica. 

Copeia  1965:302-307. 
de  AzEVEDO,  P.,  and  A.  L.  Gomes.    1942.   Contribuifao  ao  estudo  da 

biologia  da  traira  Hoplias  malabarica  (Bloch,  1 749).  Bol.  Ind.  Anim. 

5:15-64. 


LITERATURE  CITED 


699 


,  M.  ViANNA  DiAS  AND  B.  BoRGES  ViEiRA.    1938.    Biologia  do 

saguiru.  Mem.  Inst.  Oswaldo  Cruz  Rio  de  J.  33(4):48 1-554. 
DE  Beaufort,  L.  F.,  and  W.  M.  Chapman.    1951.    The  fishes  of  the 

IndoAustralian  Archipelago.  E.  J.  Brill,  Ltd.,  Leiden. 
DeBeer,  G.  R.    1937.  Thedevelopment  of  the  vertebrate  skull.  Oxford 

Univ.  Press,  London. 
.    1951.   Embryos  and  ancestors  (revised  edition).  Oxford  at  the 

Clarendon  Press,  London. 
DE  BuEN,  F.    1940.    Huevos,  crias,  larvas  yjovenes  de  C/!/ro.sro/«a  del 

Lago  de  Patzcuaro.  Trab.  Estac.  Limnol.  Patzcuaro  3:1-14. 
DE  CiECHOMSKi,  J.  D.     1965.    Observaciones  sobre  la  reproduccion, 

desarrollo  embrionario  y  larval  de  la  anchoita  argentina  (£'igraw//.s 

anchoila).  Bol.  Inst.  Biol.  Mar.  9:1-29. 
.     1968.    Huevos  y  larvas  de  tres  especies  de  peces  marinos, 

Anchoa  marinii,  Brevoortia  aurea  y  Prionolus  nudigula  de  la  zona 

de  Mar  del  Plata.  Bol.  Inst.  Biol.  Mar.  17:1-28. 
.     1972,     Desarrollo  embrionario  y  larval  del  comalito  Aus- 

troatherina  incisa  (Jenyns,  1942)  Marrero,  1950.  Pisces  Atherini- 

dae.  An.  Soc.  Cient.  Argent.  193:273-281. 
.    1981.    Ictioplancton.  In:  Atlas  del  zooplancton  del  Atlantico 

suboccidental  y  metodos  de  trabajo  con  zooplancton  marine.  D. 

Botovskoy  (ed.).  Institute  Nacional  de  Investigacion  y  Desarollo 

Pesquero  (INIDEP),  Ministeno  de  Comercio  e  Intereses  Maritimos, 

Republica  Argentina. 
,  andC.  I.  BooMAN.    1981.   Descripcion  deembrionesy  de  areas 

de  reproduccion  de  los  granaderos  Macrourus  whitsoni  y  Coelo- 

rhynchus  fascialus.  de  la  polaca  Micromesistius  aiisl rails  y  del  ba- 

calao  austral  Salilota  australis  en  la  zona  patagonica  y  fueguina  del 

Atlantico  sudoccidental.  Physis  Secc.  A  Oceanos  Org.  40  (98):5- 

14. 
,  AND  G.  Weiss.    1973.    Reproduccion,  desarrollo  embrionario 

y  larval  del  surel,  Trachurus  picluratus  australis  (Pisces),  en  el  area 

del  Mar  del  Plata  y  zonas  adyacentes.  Physis  Secc.  A  Oceanos  Org. 

32:85-93. 
,  AND .    1974.   Distribucion  de  huevos  y  larvas  de  la  mer- 

luza,  Merlucclus  merluccius  hubhsi.  en  las  aguas  de  la  plataforma 

de  la  Argentina  y  Uruguay  en  relacion  con  la  anchoita,  Engraulis 

anchoita.  y  las  condiciones  ambientales.  Physis  Secc.  A  Oceanos 

Org.  33(86):  1 85-1 98. 
,  AND .    1976.   Desarrollo  y  distribucion  de  postlarvas  del 

robalo  Eleglnops  maclovinus  (Valenciennes,  1830)  Dollo  1904;  de 

la  merluza  negra  Dissostlchus  eleglnoides  Smitt,  1899  y  de  las  no- 

totenias  Nolothenia  spp.  Pisces.  Nototheniidae.  Physis  Secc.  A 

Oceanos  Org.  35(9 1 ):  1  1  5-1 25. 
DeGaetani,  D.    1926.  Contribute  alia  conoscenzadello  sviluppopost- 

larvale  di  Mora  medllerranea  Risso.  Mem.  R.  Com.  Talassogr.  Ital. 

129. 
.     1928.    Uova  y  larve  de  Physiculus^ dalwlgkii  Kp.  Mem.  R. 

Com.  Tallasogr.  Ital.  137. 
.     1937.    Contribute  alia  cenoscenza  delle  sviluppo  post-em- 


brionale  in  Apogon  Imberbis  L.  Mem.  R.  Com.  Talassogr.  Ital.  243. 
— .    1940.    Stadi  larvali  e  giovanili  di  L(c/»a  g/a!(ca  Risso.  Mem. 


R.  Com.  Talassogr.  Ital.  270:1-15. 

Degens,  E.  T,  W.  G.  Deuser  and  R.  L.  Haedrich.  1969.  Molecular 
structure  and  composition  of  fish  otoliths.  Mar.  Biol.  (Berl.)  2:105- 
113. 

DE  Jager,  B.  v.  D.  1955.  The  South  Afncan  pilchard  (Sardinops  ocel- 
toa)— The  development  of  the  snoek  (  Thyrsnes  alun).  a  fish  pred- 
ator of  the  pilchard.  Invest.  Rep.  Div.  Fish.  S.  Afr.  19:1-16. 

Dekhnik,  T.  V.  1959.  Material  on  spawning  and  development  of 
certain  far  eastern  flatfishes.  Akad.  Nauk.  SSSR,  Zool.  Inst.  Issled. 
Dafnevost.  Morei  SSSR  6:109-131.  (In  Russian.) 

.    1973.    Iktioplankton  chemogo  moria.  (Ichthyoplankton  of  the 

Black  Sea).  Naukova  Dumka,  Kiev. 

,  M.  HiiARES  AND  D.  Salabariya.    1966.   Distribution  of  pelagic 

eggs  and  larvae  of  fishes  in  the  coastal  waters  of  Cuba.  Inst.  Biol. 
Southern  Seas,  Acad.  Sci.  USSR  1  ( 1 ):  1 38-1 6 1 .  (English  translation 
1973,  U.S.  Dep.  Commerce,  TT70-57762  Investigations  of  the 
Central  American  Seas). 


,  AND  V.  I.  SiNYUKovA.    1966.    Distribution  of  pelagic  fish  eggs 

and  larvae  in  the  Mediterranean  Sea.  Part  II.  On  the  reproduction 
and  ecology  of  larvae  in  Mediterranean  Myclophidae.  In:  Studies 
on  the  plankton  of  southern  seas.  Nauka,  Moskow.  (In  Russian; 
English  transl.  No.  55.  Bur.  Commer.  Fish.,  Syst.  Lab.,  Washing- 
ton.) 

DeLacy,  a.  C,  and  B.  S.  Batts.  1963.  Possible  heterogeneity  in  the 
Columbia  River  smelt.  Univ.  Wash.  Fish.  Res.  Inst.  Circ,  No.  198. 

,  C.  R.  HiTZ,  and  R.  L.  Dryfoos.    1964.    Maturation,  gestation 

and  birth  of  rockfish  (Sebastodes)  from  Washington  and  adjacent 
waters.  Wash.  Dep.  Fish.  Fish.  Res.  Pap.  2(3):51-67. 

DeLamater,  E,  D..  and  W.  R.  Courtenay,  Jr.  1973a.  Studies  on 
the  scale  structure  of  flat-fishes  I.  the  genus  Trinectes.  with  notes 
on  related  forms.  Proc.  Southeast  Game  Fish  Comm.  1973:591- 
608. 

,  AND .    1973b.    Variations  in  the  structure  of  the  lateral 

line  canal  on  scales  of  teleostean  fishes.  Z.  Morphol.  Tiere  75:259- 
326. 

,  AND .    1974.    Fish  scales  as  seen  by  scanning  electron 

microscopy.  Fla.  Sci.  37:141-149. 

Delsman,  H.  C.  1921.  Fish  eggs  and  larvae  from  the  Java  Sea.  1. 
Fistularia  serrata  Cuv.  Treubia  2:92-108. 

.    1923.   Fish  eggs  and  larvae  from  the  Java  Sea.  2.  C/;;rofe«/rw.s 

dorab.  Treubia  3:38-46. 

.    1925.   Fish  eggs  and  larvae  from  the  Java  Sea.  4.  Dussumieria 

haseltii.  Treubia  6:297-307. 

.    1926a.    Fish  eggs  and  larvae  from  the  Java  Sea.  5.  Caran.x 

kurra.  macrosorna.  and  crumenophthalmus.  Treubia  8: 1 99-2 1 1 . 

.    1 926b.   Fish  eggs  and  larvae  from  the  Java  Sea.  7.  The  genus 

Clupea.  Treubia  8:218-239. 

.    1926c.    Fish  eggs  and  larvae  from  the  Java  Sea.  8.  Dorosoma 

chacunda  (H.B.).  Treubia  8:389-394. 

.    1926d.   Fish  eggs  and  larvae  from  the  Java  Sea.  10.  On  a  few 


larvae  of  empang  fishes.  Treubia  8:400-412. 
.    1927.    Fish  eggs  and  larvae  from  the  Java  Sea.  1 1.  The  genus 

Trichiurus.  Treubia  9:338-351. 
— .    1929a.   Fish  eggs  and  larvae  from  the  Java  Sea.  12.  The  genus 

Engraulis.  Treubia  11:275-281. 
— .    1929b.    Fish  eggs  and  larvae  from  the  Java  Sea.  13.  Chanos 

chanos  (Forsk).  Treubia  1 1:281-286. 
— .    1929c.   The  study  of  pelagic  fish  eggs.  Proc.  Pac.  Sci.  Congr.. 

4th  Ses.  3:61-67. 
.    1930a.    Fish  eggs  and  larvae  from  the  Java  Sea.  14.  The  genus 

Pellona.  Treubia  12:37-46. 
.    1930b.    Fish  eggs  and  larvae  from  the  Java  Sea.  15.  On  Chl- 

rocentrus  hypselosoma  and  dorab.  Treubia  1 2:46-50. 
.    1930c.   Fisheggsandlarvaefrom  the  JavaSea.  16.  .-t/jpp/^pr/o/! 

perculac.\.  Treubia  12:367-370. 
— .    1930d.   The  study  of  pelagic  fish-eggs.  Proc.  Fourth  Pac.  Sci. 


Congr.  (1929)  3:61-67. 
— .    1931.    Fish  eggs  and  larvae  from  the  Java  Sea.  1 7.  The  genus 

Stolephorus.  Treubia  13:217-243. 
— .    1 932a.   Fish  eggs  and  larvae  from  the  Java  Sea.  1 9.  The  genus 

Selipmna.  Treubia  14:109-114. 
— .    1932b.   Fish  eggs  and  larvae  from  the  Java  Sea.  20.  The  genus 

Coilia.  Treubia  14:1 14-1  16. 
— .    1933a.   Fish  eggs  and  larvae  from  the  Java  Sea.  22.  C/!*^^^);^^^ 

lile(C.V.).  Treubia  14:247-249. 
— .    1933b.    Fish  eggs  and  larvae  of  the  Java  Sea.  23.  A  few  more 

Clupea-eggs.  Treubia  14:249-254. 
— .     1938.    Fish  eggs  and  larvae  from  the  Java  Sea.  24.  Mycto- 

phoidea.  Treubia  16:415-420. 
,  AND  J.  D.  G.  Hardenberg.    1934.    De  Indische  Zeevisschen 


en  Zeevischeriji,  p.  330-333.  N.  V.  Boekhandel  en  Drukkerij  Visser 
Co.,  Batavia. 
DeMartini,  E.  E.  1976.  The  adaptive  significance  of  territoriality  and 
egg  cannibalism  in  the  painted  greenling,  Oxylebius  pictus  Gill,  a 
northeastern  Pacific  marine  fish.  Unpubl.  PhD  thesis.  Univ.  Wash- 
ington, Seattle. 


700 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


.    1978.    Apparent  paternal  care  in  Lipans  fucensis  (Pisces:  Cy- 

clopteridae).  Copeia  1978:537-539. 
Demir,  M.    1961.    On  the  eggs  and  larvae  of  rrac/iwnM  ?rac/n<n«  (L.) 

and  Trachurus  mediterranus  (Stohrr.)  from  the  Sea  of  Marmara 

and  the  Black  Sea.  Rapp.  P.  V.  Reun.  Comm.  Int.  Explor.  Sci.  Mer 

Mediterr.  Monaco  16:317-320. 
Demir,N.   1972.  The  abundance  and  distribution  ofthe  eggs  and  larvae 

of  some  teleost  fishes  off  Plymouth  in   1969  and   1970.  II.  The 

postlarvae  of  Callionymus.  J.  Mar.  Biol.  Assoc.  U.K.  52:997-1010. 
.    1976.  Callionymidaeof  the  northeastern  North  Atlantic.  Fich. 

Ident.  Zooplanclon  148:1-5. 
.    1982.   On  postlarvae  and  pelagic  juveniles  of  rocklings,  Ga;(y- 

ropsarus  mediterranem  and  G.  hiscayensis.  J.  Mar.  Biol.  Assoc. 

U.K.  62:647-665. 
Dessier,  a.     1969.    Note  sur  les  stades  larvaires  et  post-larvaires  d' 

Ilisha  africana  (Bloch,  1795)  (Pisces,  Clupeidae).  Cah.  ORSTOM 

Ser.  Oceanogr.  8(4):21-27. 
DE  Sylva,  D.  P.    1963.    Systematics  and  life  history  of  the  great  bar- 
racuda, Sphyraena  barracuda  (Walbaum).  Stud.  Trop.  Oceanogr. 

(Miami). 
.    1975.    Barracudas  (Pisces:  Sphyraenidae)  of  the  Indian  Ocean 

and  adjacent  seas  — a  preliminary  review  of  their  systematics  and 

ecology.  J.  Mar.  Biol.  Assoc.  India  15(l):74-94. 
,  AND  W.  N.  EscHMEYER.    1977.   Systematics  and  biology  of  the 

deep-sea  fish  family  Gibberichthyidae,  a  senior  synonym  of  the 

family  Kasidoroidae.  Proc.  Calif  Acad.  Sci.,  Ser.  4,  41:215-231. 
AND  L.  N.  ScoTTON.     1972.    Larvae  of  deep-sea  fishes  (Sto- 


miatoidea)  from  Biscayne  Bay,  Florida,  USA,  and  their  ecological 

significance.  Mar.  Biol.  (Bed.)  12:122-128. 
Detwyler,  R.     1963.    Some  aspects  of  the  biology  of  the  seasnail, 

Liparis  atlanticus  (Jordan  and  Evermann).  Unpubl.  PhD  disser- 
tation. Univ.  of  New  Hampshire. 
Deuel,  D.  G.,  J.  R.  Clark  and  A.  J.  Mansueti.    1966.    Description 

of  embryonic  and  early  larval  stages  of  the  bluefish,  Pomatomus 

saltatnx.  Trans.  Am.  Fish.  Soc.  95:264-271. 
Devanesan,  D.  W.    1937.    On  the  spawning  habits  and  the  hatching 

experiments  with  the  eggs  of  Hemirhamphus  georgii  Cuv.  et  Val. 

C.  R.  12th  Congr.  Intematl.  Zool.  (Lisbon)  3:2175-2181. 
Devaraj,  K.  v.,  T.  J.  Varchese  and  G.  P.  S.  Rao.     1972.    Induced 

breeding  ofthe  freshwater  catfish,  Clarias  balrachus  (,Lmn.)by  using 

pituitary  glands  from  manne  catfish.  Curr.  Sci.  41:868-870. 
Devi,  C.  B.  L.    1969.   Occurrence  o{  larvae  of  Psctidorlionihiis  elevatus 

Ogilby  (Heterosomata-Pisces)  along  the  south-west  coast  of  India. 

Proc.  Indian  Acad.  Sci.  Sect.  B:  178-1 86. 
DeWitt,  H.  H.    1962a.   A  new  genus  and  species  of  zoarcid  fish  from 

McMurdo  Sound.  Antarctica.  Copeia  1962:819-826. 
.    1962b.    A  new  Antarctic  nototheniid  fish  with  notes  on  two 

recently  described  nototheniiformes.  Copeia  1962:826-833. 
.    1964a.    A  revision  of  the  Antarctic  genus /^acYnv/rM  (Pisces, 

Bathydracomdae).  Copeia  1964:496-506. 
.    1 964b.   A  redescription  of  Pathogenis  anlarctica.  with  remarks 

on  the  genus  Treinatomus  (Pisces.  Nototheniidae).  Copeia  1964: 

683-686. 
.    1970.    A  revision  of  the  fishes  of  the  genus  Notothenia  from 

the  New  Zealand  region,  including  Macquarie  Island.  Proc.  Calif 

Acad.  Sci.  38(16):299-340. 
.    1971.   Coastal  and  deep-water  benthic  fishes  of  the  Antarctic. 

Am.  Geogr.  Soc,  Antarct.  Map  Folio  Ser.,  Folio  15. 
.    1977.   A  new  genus  and  species  of  eelpout  (Pisces,  Zoarcidae) 

from  the  Gulf  of  Mexico.  Fish.  Bull.  U.S.  75:789-793. 
,  AND  J.  C.  Hureau.     1979.    Fishes  collected  during  "Hero" 

Cruise  72-2  in  the  Palmer  Archipelago  Antarctica  with  the  descrip- 
tions of  two  new  genera  and  three  new  species.  Bull.  Mus.  Natl. 

Hist.  Nat.  4e  Ser.  I,  1979,  Sec.  A  (3):775-820. 
,  AND  J.  C.  Tyler.    I960.    Fishes  from  the  Stanford  Antarctic 


da  lumbil  (Bloch)  off  southwest  coast  of  India,  p.  460-473.  In: 
Proc.  Symp.  Warm  Water  Zooplankton.  Natl.  Inst.  Oceanogr. 

DiMkhele,  L.,  AND  M.  H.  Taylor.  1980.  The  environmental  control 
of  hatching  in  Fundidus  heteroclitus.  J.  Exp.  Zool.  214:181-187. 

DiNGERKUS,  G.,  AND  L.  D.  Uhler.  1977.  Enzyme  clearing  of  alcian 
blue  stained  whole  small  vertebrates  for  demonstration  of  cartilage. 
Stain  Technol.  52:229-232. 

DiTTY,  J.  G.,  AND  F.  M.  Truesdale.  1983.  Comparative  larval  de- 
velopment of  Peprilus  burti.  P.  triacanlhus  and  P.  paru  (Pisces: 
Stromateidae)  from  the  western  North  Atlantic.  Copeia  1983:397- 
406. 

DoBBS,  G.  H.  1974.  Soft  tissues  of  marine  teleosts,  p.  72-82.  In: 
Principles  and  techniques  of  scanning  electron  microscopy.  M.  A. 
Hayat  (ed.).  Van  Nostrand  Reinhold  Co.,  New  York. 

.    1975.    Scanning  electron  microscopy  of  inlraovarian  embryos 

ofthe  viviparous  teleost.  Micrometrus  minimus  (Gibbons),  (Per- 
ciformes:  Embiotocidae).  J.  Fish  Biol.  7:209-214. 

Doi,  M.,  AND  O.  Okamura.  1983.  The  chlorophthalmid  fishes  col- 
lected from  the  western  Pacific.  Abstract  of  paper  presented  at  the 
15th  Ann.  Meet,  ofthe  Ichthyol.  Soc.  Jpn.,  1983. 

DooLEY,  J.  S.  1978.  Systematics  and  biology  ofthe  tilcfishes  (Perci- 
formes:  Branchiostegidae  and  Malacanthidae)  with  descriptions  of 
two  new  species.  NOAA  Tech.  Rep.  NMFS  Circ.  411. 

DoROFE^EVA,  Y.  A.,  Y.  A.  Zinov'yev,  v.  a.  Klyukanov,  Y.  S. 
Reshetnikov,  K.  a.  Savvaitova  and  G.  K.  Shaposhnikova. 
1 980.  The  present  state  of  research  into  the  phylogeny  and  classi- 
fication of  Salmonoidei.  J.  Ichthyol.  (Engl.  Transl.  Vopr.  Ikhtiol.) 
20(5):  1-20. 

DoROSHEv  S.  1 .  1 970.  Biological  features  ofthe  eggs,  larvae,  and  young 
ofthe  striped  bass  (Roccus  saxatilis  (Walbaum))  in  connection  with 
the  problem  of  its  acclimitization  in  the  U.S.S.R.  J.  Ichthyol.  (Engl. 
Transl.  Vopr.  Ikhtiol.)  10:235-248. 

,  AND  J.  W.  Cornacc  HiA.    1979.    Initial  swim  bladder  inflation 

in  the  larvae  of  TUapia  mossambica  (Peters)  and  Morone  saxatdis 
(Walbaum).  Aquaculture  16:57-66. 

, AND  K.  HotiAN.    1981.    Initial  swim  bladder  inflation 


in  the  larvae  of  physoclistous  fishes  and  its  importance  for  larval 

culture.  Rapp.  P  V.  Reun.  Cons.  Int.  Explor.  Mer  178:495-500. 
DoTSU,  Y.    1954.   On  the  life  history  of  a  goby  Chaenogobius  castanea 

O'Shaughnessy.  Contnb.  Biol.  Lab.  Sci.  Soc.  China  1  1(6):133-I38. 
.    1957.    On  the  bionomics  and  life  history  of  the  eel-like  goby, 

Odontamblyopusrubicundu5(YiZLVn\\lon).  Sci.  Bull.  Fac.  Agric.  Kyu- 
shu Univ.  16(1):101-1I0. 
.    1958.  The  bionomics  and  larvae  ofthe  goboid  fishes,  C/p/io(n'- 

pauchen  microcepholus  (Bleeker)  and  Taenioides  cirratus  (Blyth). 

Sci.  Bull.  Fac.  Agric.  Kyushu  Univ.  16(3):37 1-380. 
.     1979.    Eggs,  larvae,  and  juveniles  of  gobioid  fishes.  Kaiyo 

Kaguku  (Mar.  Sci.)  11:111-1 16.  (In  Japanese.) 
.    1982.   The  early  life  history  of  the  combtooth  blenny,  £)aMo/i 

Irossulus  and  spawnings  of  the  laboratory-reared  fish  about  three 

months  after  hatching.  Bull.  Fac.  Fish.  Nagasaki  Univ.  52:19-27. 

(In  Japanese,  English  summary.) 
,  AND  S.  FujiTA.    1959.    The  bionomics  and  life  history  ofthe 

eleotrid  fish.  Eleolris  oxycephata  Temminck  and  Schlegel.  Bull.  Fac. 

Fish.  Nagasaki  Univ.  8:191-195. 
,  andS.  MiTO.    1955.   On  the  breeding  habits,  larvae,  and  young 

of  a  goby,  Acanthogohius  flavimanus  (Temminck  et  Schlegel).  Jpn. 

J.  Ichthyol.  4:152-161. 
,  AND  .     1963.    The  bionomics  and  life  history  of  the 

eleotrid  fish,  Asterropteryx  semipunclalus  Ruppell.  Bull.  Fac.  Fish. 

Nagasaki  Univ.  15:1-3. 
,  AND  S.  MoRRicHi.    1980.    The  life  history  of  the  blcnniid  fish. 


Biological  Research  Program  1958-1959.  Stanford  Ichthyol.  Bull. 
7(4):162-I99. 
DiLEEP,M.P.    1977.  The  larval  development  and  distribution  of  S'lJi/n- 


BtenniusyalabeiOorAanelSnyAer).  Bull.  Fac.  Fish.  Nagasaki  Univ. 

49:17-24.  (In  Japanese,  English  summary.) 
— ,  AND  T.  OoTA.     1973.    The  life  history  of  the  blenniid  fish, 

Omobranchus  loxozonus.  Bull.  Fac.  Fish.  Nagasaki  Univ.  36:13- 

22.  (In  Japanese,  English  summary.) 
— ,  AND  M.  SHiociAK.1.    1971.    Larvae,  juvemlcs  and  young  of  the 

gobiid  fish,  Astrabe  lactisella.  Jpn.  J.  Ichthyol.  18:182-186. 


LITERATURE  CITED 


701 


DovEL,  W.  L.  1963.  Larval  development  of  clingfish.  GoWmo.v  i^rw- 
mosiis.  4.0  to  12.0  millimeters  in  total  length.  Chesapeake  Sci. 
4:161-166. 

Dryfoos,  R.  L.  1965.  The  life  histor>' and  ecology  of  the  longfin  smelt 
in  Lake  Washington.  Unpubl.  PhD  dissertation.  University  of 
Washington. 

DiiDNiK,  Y.  I.  1977.  A  contnbution  to  the  biology  of  larvae  and  fry 
of  morwongs  (Pisces,  Cheilodactylidae)  of  the  Atlantic  Ocean.  Zool. 
Zh.  56(11);1658-1667. 

DuMONT,  J.  N.,  AND  A.  R.  Brummett.  1980.  The  vitelline  envelope, 
chorion,  and  micropyle  of  Fundulus  heteroclitus  egg,s.  Gamete  Res. 
3:25-44. 

Dunbar,  M.  J.  1947.  Marine  young  fish  from  the  Canadian  eastern 
Arctic.  Bull.  Fish.  Res.  Board  Can.  73:1-1 1. 

DiiNCKER,  G.,  andE.  MoHR.  1939.  Revisionder  Ammodytidae.  Mitt. 
Zool.  Mus.  Berl.  24(I):3-3I. 

DuNKELBiiRGER,  D.  G.,  J.  M.  Dean  AND  N.  Watabe.  1980.  The  ul- 
trastructuie  of  the  otolithic  membrane  and  otolith  in  the  juvenile 
mummichog,  Fundulus  heterocliliis.  ].  Morphol.  163:367-377. 

Dunn.  J.  R.  1983a.  Development  and  distribution  of  the  young  of 
northern  smoothtongue,  Leuroglossus  schmidli  (Bathylagidae).  in 
the  northeast  Pacific,  with  comments  on  the  systemalics  of  the 
Genus  Leuroglossus  G\\hen.  Fish.  Bull.  U.S.  81:23-40. 

.    1983b.    The  utility  of  developmental  osteology  in  ta,\onomic 

and  systematic  studies  of  teleost  larvae:  a  review.  NOAA  Tech. 
Rep.  NMFS  Circ  450. 

,  AND  B.  M.  ViNTER.    1984.    Development  of  larvae  of  saffron 

cod,  Eleginus  gracilis,  with  comments  on  the  identification  of  gadid 
larvae  in  Pacific  and  Arctic  waters  contiguous  to  Canada  and  Alas- 
ka. Can.  J.  Fish.  Aquat.  Sci.  41:304-318. 

,  AND .    MS.   The  developmental  osteology  of  the  North 

Pacific  gadid  fish,  Theragra  chalcogramma.  Unpubl.  manuscr.. 
Northwest  and  Alaska  Fisheries  Center,  National  Marine  Fisheries 
Service,  NOAA,  2725  Montlake  Blvd.  E.,  Seattle,  WA  981  12. 

DuRAND.J,  R.,  andG.  LouBENS.  1971.  Developpementembryonnaire 
et  larvaire  A'Alestes  baremoze.  Cah.  ORSTOM.  Ser.  Hydrobiol. 
5:137-145. 

DuTT,  S.,  AND  V.  V.  Rao.  1960.  On  the  breeding  habits  and  eariy 
developmental  stages  of  Peirosciries  bipunclala  (Day).  ,1.  Zool.  Soc. 
India  12:158-161. 

D'ViNCENT,  S..  H.  G.  MosER  AND  E.  H.  Ahl-STROM.  1980.  Descnption 
of  the  larvae  and  early  juveniles  of  the  Pacific  butlerfish,  Peprilus 
simillimus  (Family  Stromateidae).  Calif  Coop.  Oceanic  Fish.  In- 
vest. Rep.  21:172-179. 

Dye,  J.  E.  1980.  The  production  and  eflicient  use  of  freshly  hatched 
brine  shrimp  nauplii  (Anemia)  in  the  larval  rearing  of  marine  fish 
at  the  hatcheries  of  the  Bntish  Whitcfish  Authority.  In:  The  Bnne 
Shrimp  Anemia.  Vol.  3.  G.  Persoone,  P.  ^orgeloos,  O.  Roels  and 
E.  Jaspers  (eds.).  Universa  Press,  Wetleren,  Belgium. 

Eat(in,  T.  H.  1953.  Pygmy  sunfishes  and  the  meaning  of  pygmyism. 
J.  Elisha  Mitchell  Sci.  Soc.  69:98-99. 

.    1956.    Notes  on  the  olfactory  organs  in  Centrarchidae.  Copeia 

1956:196-199. 

Ebeleng,  A.  W.  1957.  The  dentition  of  eastern  Pacific  mullets,  with 
special  reference  to  adaptation  and  taxonomy.  Copeia  1957:173- 
185. 

.    1961.    .Wi(?;7  ga/apagCT!i;.s,  a  new  mullet  from  the  Galapagos 

Islands,  with  notes  on  related  species  and  a  key  to  the  Mugilidae 
of  the  eastern  Pacific.  Copeia  1961:295-304. 

.    1962.   Melamphaidac.  I.  Systematics  and  zoogeography  of  the 

species  in  the  bathypelagic  fish  genus  .Melamphaes  Gunther.  Dana- 
Rep.  Carlsberg  Found.  58. 

,  AND  W.  Weed.     1963.    Melamphaidac.  111.  Systematics  and 

distribution  of  the  species  in  the  bathypelagic  fish  genus  Scopelo- 
gadus  Vaillant.  Dana-Rep.  Carlsberg  Found.  60. 

,    AND   .      1973.     Order   Xenoberyces   (Sephanoberyci- 


(Paralepididae,  Pisces).  J.  Ichthyol.  (Engl.  Transl.  Vopr.  Ikhtiol.) 
18:500-504. 

.    1983a.   Atlas  offish  larvae  from  the  Southern  Ocean.  Cybium, 

3e  Ser.  7(2):  1-74. 

.    1983b.   Description  ofeggs  and  larvae  of  A/«rae«o/ep/im/cTOpj 

Lonnberg  (Muraenolepidae)  from  the  Scotia  Sea.  J.  Ichthyol.  (Engl. 
Transl.  Vopr.  Ikhtiol.)  23(1):I39-I42. 

,  A.  N.  KoTLVAR  AND  L.  A.  LisovENKO.  1981.  Sex  cycles,  fe- 
cundity, and  distribution  of  "limbless"  soles  of  family  Bothidae 
from  the  southwestern  Atlantic.  .1.  Ichthyol.  (Engl.  Transl.  Vopr. 
Ikhtiol.)2l(6):89-99. 

,  AND  L.  A.  LrsovENKO.    1970.    Morphological  features  of  in- 

traovarian  and  pelagic  larvae  of  some  Sebastodes  species  inhabiting 
the  Gulf  of  Alaska,  p.  267-286.  In:  Soviet  fisheries  investigations 
in  the  northeast  Pacific.  Part  V.  P.  A.  Moiseev  (ed.).  (Transl.  from 
Russian;  available  from:  NTIS.  Springfield,  VA  TT7 1-50 127.) 

Ege,  V.  1918.  Stomiatidae.  Rep.  Dan.  Oceanogr.  Exped.  Mediterr. 
2(A.4). 

.  1930.  Sudidae(Para/e/)K).  Rep.  Dan.  Oceanogr.  Exped.  Med- 
iterr. 2(A.13). 

.    1953.    Paralepididae  I  (Para/epii  and  Le5/;V/;H/H).  Taxonomy, 

ontogeny,  phylogeny  and  distribution.  Dana-Rep.  Carlsberg  Found. 
40. 

.  1957.  Paralepididae  II  (Macroparalepis).  Taxonomy,  ontog- 
eny, phylogeny  and  distribution.  Dana-Rep.  Carlsberg  Found.  43. 
1958.    Omosudis  Gunther,  bathypelagic  genus  offish.  Dana- 


formes),  p.  397-478.  In:  Fishes  of  the  western  North  Atlantic.  D. 
M.  Cohen  (ed.).  Mem.  Sears  Found.  Mar.  Res.  No.  1  (Pt.  6). 
Efremenko,  V.  N.    1978.  A  description  of  the  larvae  of  A'ow/cp/sroari/ 


Rep.  Carlsberg  Found.  47. 

Eggert,  B,  1932.  Zur  Kenntnis  der  Biologic,  der  sekundaren  Ge- 
schlechtsmerkmale  und  des  Eies  von  Blennius pavo  (Risso).  Z.  Mor- 
phol. Oekol.  Tiere  24:682-703. 

Ehrenbaum,  E.  1 894.  Beitrage  zur  Naturgeschichte  einiger  Elbflsche. 
(Osmerus  epertanus  L.,  Clupea  finia  Cuv.,  .-icerina  cernua  L.,  Aci- 
penserslunoL.).  Wiss.  Meersunters.  (Helgol.)  l(Pt.I  ):35-82.  (Cited 
m:  W.  C.  Kendall.  1 927.  The  smelts.  Bull.  Bur.  Fish.  Wash.  42(  1 926): 
217-375.) 

.     1897.    Eier  und  Larven  von  Fischen  der  Deutschen  Buchl. 

Wiss.  Meeresunters.  (Helgol.)  2:253-324. 

.    1904.   Eier  und  Larven  von  Fischen  der  Deutschen  Bucht.  III. 

Fische  mit  festsitzenden  Eiem.  Wiss.  Meeresunters  (Helgol.)  N.F. 
6:127-200. 

.    1905-1909.    Eier  und  Larven  von  Fischen.  Nord.  Plankton  1 

and  2. 

,  AND  S.  Strodtmann.    1904.    Eier  und  Jugendformen  der  Ost- 

seefische.  Wiss.  Meeresunters.  (Helgol.)  6:57-126. 

Ehrich,  S.  1976.  Zur  Taxonomie,  Okologie  und  Wachstum  von  Ma- 
crorhamphosus  scnlopa.x  (Linnaeus,  1758)  (Pisces,  Syngnathi- 
formes)  aus  dem  subtropischen  Nordostatlantik,  Ber.  Dtsch.  Wiss. 
Komm.  Meeresforsch.  24:523-537. 

Ehrlich,  K.  F.,  J.  H.  S.  Blaxter  and  R.  Pemberton.  1976.  Mor- 
phological and  histological  changes  during  the  growth  and  starva- 
tion of  herring  and  plaice  larvae.  Mar.  Biol.  35:105-118. 

Eigenmann,  C.  H.  1902.  The  egg  and  development  of  the  conger  eel. 
Bull.  U.S.  Fish  Comm.  21:37-44. 

.    1928.    The  freshwater  fishes  of  Chile.  Mem.  Natl.  Acad.  Sci. 

22:1-80. 

EINARSSON,  H.  1951.  The  post-larval  stages  of  sandeels  in  Faroe,  Ice- 
land, and  W-Greenland  waters.  Acta  Nat.  Isl.  l(7):5-75. 

.    1955.  On  the  post-larval  stages  of  .■lm/)Joa'i7M/a/!fea(Cuvier). 

Acta  Nat.  Isl.  2(l):3-7. 

,  and  B.  Rojas  de  Mendioea.    1963.    Descripcion  de  huevos  y 

larvas  de  anchoveta  Peruana  {Engraiilis  ringens  }.).  Bol.  Inst.  Mar 
Peru(Callao)  1. 

Eldred,  B.  1967a.  Larval  bonefish,  .Ithula  vulpcs  (Linnaeus.  1758) 
(Albulidae)  in  Flonda  and  adjacent  waters.  Fla.  Board  Conserv. 
Mar.  Res.  Lab.  Lead.  Ser.  4  (Pt.  1,  No.  3). 

.     1967b.     Larval   tarpon,    Megalops  allanlicus  Valenciennes 

(Megalopidae)  in  Florida  waters.  Fla.  Board  Mar.  Res.  Lab.  Leafi. 
Ser.  4  (Pt.  1,  No.  4). 

.    1968.    The  larval  development  and  taxonomy  of  the  pygmy 


702 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


moray  eel,  Anarchias yoshiaeKanazcLVja  1952.  Fla.  Board  Conserv. 

Mar.  Res.  Lab.  Leafl.  Ser.  4  (Pt.  1.  No.  10). 
.    1969.   Embryology  and  development  of  the  Blackedge  moray 

Gymnolhorax  nigromarginatus  (Girard.  1859).  Fla.  Board  Conserv. 

Mar.  Res.  Lab.  Leafl.  Ser.  4  (Pt.  1,  No.  13). 
.    1972.   Note  on  larval  tarpon,  A/e.fa/opi  a//an;;cMi  (Megalopi- 

dae),  in  the  Florida  Straits.  Fla.  Dep.  Nat.  Resour.  Mar.  Res.  Lab. 

Leafl.  Ser.  4  (Pt.  1.  No.  22). 
,  AND  W.  G.  Lyons.    1966.    Larval  ladyfish,  Elops  saurus  Lm- 

naeus  1766,  (Elopidae)  in  Florida  and  adjacent  waters.  Fla.  Board 

Conserv.  Mar.  Res.  Lab.,  Leafl.  Ser.  4  (Pt.  1.  No.  2). 
Eldridge,  M.  B.    1970.    Larval  fish  survey  of  Humboldt  Bay.  Unpubl. 

MS  thesis.  Humboldt  State  Univ.,  Areata,  CA. 
.    1975.    Early  larvae  of  the  diamond  turbot.  Hypsopsetia  giil- 

tulata.  Calif  Fish  Game  61:26-34. 
Elston,  R.,  L.  Corazza  and  J.  G.  Nickhm.    1981.    Morphology  and 

development  of  the  olfactory  organ  in  larval  walleye,  Stizostedion 

vilreum.  Copeia  1981:890-893. 
Emery,  A.    1973.   Biology  survey  — summer e.xpedition. //i.  Arctic  Div- 
ing. Adv.  Comm.  Nor.  Dev.,  No.  60°,  Govt.  Canada,  Ottawa. 
Emery,  C.    1879.   Le  metamorfosi  del  Trachipterus  taenia.  Mitt.  Zool. 

Sta.  Neapel  1:58-588. 
.     1880.    Le  specie  del  genere  Fierasfer  nel  golfo  di  Napoli  e 

regione  limitrofe.  Fauna  Flora  Golf  Neapel  2:1-76. 
.    1886.    Contribuzioni  all'ittiologia.  XVIL  Larva  di  genere  ig- 

noto.  Mitt.  Zool.  Sta.  Neapel  6:149-164. 
Eng,  C.  T.    1969.   Early  development  of  the  glassfish  [Amhassis  kopsii 

Elk).  J.  Singapore  Natl.  Acad.  Sci.  l(2):37-44. 
Engen,  P.  C.    1968.    Organogenesis  in  the  walleye  surfperch,  Hyper- 

prosopon  argenleum  (Gibbons).  Calif  Fish  Game  54:156-169. 
Erdahl,  D.  A.,  AND  E.  F.  Grahm.    1980.    Preservation  of  gametes  of 

freshwater  fish,  p.  317-336.  In:  Proc.  9th  International  Congress 

on  Animal  Reproduction  and  A.L  Vol.  II,  Round  Tables,  Invited 

Guests.  Madrid,  Spain. 
EscHMEVER,  W.  N..  and  B.  B.  Collette.     1966.    The  scorpionfish 

subfamily  Setarchinae,  including  the  genus  Ectreposebastes.  Bull. 

Mar.  Sci.  16:349-375. 
,  E.  Herald  and  H.  Hammann.    1983.    A  field  guide  to  Pacific 

Coast  fishes  of  North  America.  Peterson  Field  Guide  Series  28. 

Houghton  Mifllin  Co.,  Boston. 
-,  and  J.  C.  HuREAU.    1971.    Sehasles  mouchezi.  a  senior  syn- 


onym of  Helicoleniis  tnstanensis.  with  comments  on  Sebastes  ca- 

pensis  and  zoogeographical  considerations.  Copeia  1971:576-579. 
EsPiNOSA,  F.  A.    1968.   Spawning  periodicity  and  fecundity  of  C/'e«;f/(- 

thys  baileyi,  a  fish  endemic  to  Nevada.  Unpubl.  MS  thesis.  Nevada 

South.  Univ.,  Las  Vegas. 
EsTABROOK.  G.  F.,  C.  S.  Johnson,  Jr.  and  F.  R.  McMoRRis.    1976.   A 

mathematical  foundation  for  the  analysisofcladisticcharacter  com- 
patibility. Math.  Biosci.  29:181-187. 
Evans,  J.  W.    1962.    Normal  stages  of  the  early  development  of  the 

flying  fish,  Hirundichlhys  ajfinis  (Gunther).  Bull.  Mar.  Sci.  Gulf 

Carib.  ll(4):483-502. 
Evans,  R.  R.     1973.    The  swimbladder  and  associated  structures  in 

western  .Atlantic  sea  robins  (Triglidae).  Copeia  1973:315-321. 
EvERMANN,  B.  W.,  AND  W.  C.  KENDALL.    1898.    Descriptions  of  new 

or  little  known  genera  and  species  of  fishes  from  the  United  States. 

Bull.  U.S.  Fish.  Comm.  17:125-133. 
EvERSON,  I.    1968.    Larval  stages  of  certain  Antarctic  fishes.  Bull.  Br. 

Antarct.  Surv.  16:65-70. 
.    1970.   Reproduction  in  No/oJ/ien/a  neg/ec(a  Nybelin.  Bull.  Br. 

Antarct.  Surv.  33:81-92. 
EvsEENKO.  S.  A.     1977.     Larval  Engyophrys  senlus  Ginsburg.   1933 

(Pisces,  Bolhidae)  from  the  American  Mediterranean  Sea.  Coop. 

Invest.  Carib.  Adjac.  Reg.-II.  FAO  Fish.  Rep.  200:171-185. 
.    1979.   Larvae  ofthe  flounder  CR7op5c»a  Gill.  1888  (Bothidae, 

Pisces)  from  the  north-western  Atlantic.  Biol.  Morva  (Vladivost.) 

2:67-75. 
.    1981.    On  the  sinistral  flatfish  larvae  (Scophthalmidae,  Both- 


idae,  Pisces)  from  the  west  Atlantic,  p.  593-594.  In:  The  early  life 

history  of  fish.  Recent  studies.  R.  Lasker  and  K.  Sherman  (eds.). 

Rapp.  P.  V.  Reun.  Cons.  Int.  Explor.  Mer  178. 
.    1982a.   Eco-morphological  peculanties  of  the  early  life  history 

of  flatfishes  of  the  western  North  Atlantic.  Proc.  Ichthyoplankton 

and  Nekton  of  the  World  Ocean.  P.  P.  Shirshov  Inst.  Oceanol. 

Acad.  Sci.  USSR.  1 18:43-84. 
.    1982b.    Ichthyoplankton  of  slope  and  Gulf  Stream  waters  off 

Nova  Scotia  in  late  autumn   1974.  J.  Northwest  Atl.  Fish.  Sci. 

3:127-139. 
Faber,  D.  J.     1970.    Ecology  of  larval  whitefishes  in  Lake  Huron,  p. 

481-500.  In:  Biology  of  coregonid  fishes.  C.  C.  Lindsey  and  D.  S. 

Woods  (eds.).  Univ.  Manitoba  Press,  Winnipeg. 
.    1976.    Identification  of  four  northern  blennioid  fish  larvae  in 

the  Canadian  Atlantic  Ocean  (Stichaeidae,  Lumpenidae).  J.  Fish. 

Res.  Board  Can.  33:1798-1802. 
.    1982.    Fish  larvae  caught  by  a  light-trap  at  littoral  sites  in  Lac 

Heney,  Quebec.  1979  and  1980,  p.  42-46.  //;.  Proc.  5th  Annual 

Larval  Fish  Conference.  La.  Coop.  Fish.  Res.  Unit,  Baton  Rouge, 

LA. 
Fabre-Domergue,  p.,  and  E.  BmxRLX.     1905.    Developpement  de  la 

Sole  (Solea  vulgaris).  Trav.  Lab.  Zool.  Mar.  Concameau. 
Facciola,  L.    1882.    La  forma  giovanile  del  Phycts  blennioides.  Nota. 

Nat.  Siciliano  (Palermo)  2:25-29. 
Fage,  L.    1910.    Recherches  sur  les  Stades  pelagiques  de  quelques  Te- 

leosteens  de  la  mer  de  Nice  (Parages  de  Monaco)  et  du  Golfe  du 

Lion.  Annls.  Inst.  Oceanogr.  (Monaco)  1,  Fasc.  7. 
.    1918.    Shore  Fishes.  Rep.  Dan.  Oceanogr.  Exped.  Mediterr. 

2(A.3). 
Fahay,  M.  p.     1975.    An  annotated  list  of  larval  and  juvenile  fishes 

captured  with  surface-towed  meter  net  in  the  south  Atlantic  Bight 

during  four  R/V  Dolphin  cruises  between  May  1967  and  February 

1968.  NOAA  Tech.  Rept.  NMFS  SSRF  685. 
.    1983.    Guide  to  the  early  stages  of  marine  fishes  occurring  in 

the  western  North  Atlantic  Ocean,  Cape  Hatteras  to  the  southern 

Scotian  Shelf  J.  Northwest.  .\X\.  Fish.  Sci.  4. 
,  AND  P.  Berrien.    1981.    Preliminary  description  of  larval  tile- 
fish  (Lophotatilus  chamaeleonticeps).  Rapp.  P.  V.  Cons.  Int.  Explor. 

Mer  178:600-602. 
,  AND  C.  L.  Obenchain.    1978.    Leptocephali  of  the  ophichthid 


genera  .4hlia.  Myrophis.   Ophichthus.   Pisodonophis.   Caltechctys. 

Letharchus.  and  .Iptenchtus  on  the  Atlantic  continental  shelf  of  the 

United  States.  Bull.  Mar.  Sci.  28:442-486. 
Fahv,  W.  E.    1982.    The  influence  of  temperature  change  on  number 

of  pectoral  fin  rays  developing  in  Fundubis  majalis  (Walbaum).  J. 

Cons.  Int.  Explor.  Mer  40:21-26. 
Fanara,  D.  M.    1964.    Notes  on  the  biology  of  a  salt  marsh  minnow. 

Proc.  N.J.  Mosq.  Exterm.  Assoc.  51:152-159. 
Fange,  R.     1958.    The  structure  and  function  of  the  gas-bladder  in 

Argentina  sihts.  Q.  J.  Microsc.  Sci.  99:95-102. 
F.A.O.    1983.   Catches  and  landings  1981.  Yearbook  of  fishery  statis- 
tics. 52.  FAO,  Rome. 
Farris,  J.  S.    1970.    Methods  for  computing  Wagner  trees.  Syst.  Zool. 

19:83-92. 
,  A.  G.  Kllige  AND  M.  J.  EchLARDT.    1970.  A  numerical  approach 

to  phylogenetic  systematics.  Syst.  Zool.  19:172-191. 
Farrugio,  H.   1977.  Clescommentees  pour  la  determination  desadultes 

et  des  alevins  de  Mugilidae  de  Tunisie.  Cybium.  3rd  Ser.  (2):57- 

73. 
Fast.  T.  N.    1 960.   Some  aspects  of  the  natural  history  ofStenobraciuus 

leucopsariis  Eigenmann  and  Eigenmann.  Unpubl.  PhD  dissertation. 

Stanford  Univ.,  Stanford,  CA. 
Fedorov,  V.  V.     1975.    Description  of  a  new  genus  and  species  of 

zoarcid  fish.  Puzanovia  rubra,  gen.  et  sp.  n.  (Pisces,  Zoarcidae)  from 

the  northern  Pacific  Ocean.  J.  Ichthyol.  (Engl.  Transl.  Vopr.  Ikhtiol.) 

15:527-531. 
.    1982.    A  new  ee\pout.  Hadropogonichthys  tindbergi  Fedorov. 

gen.  et  sp.  nov.  (Zoarcidae),  from  the  bathyal  depths  of  the  fourth 


LITERATURE  CITED 


703 


Kurile  Strait.  J.  Ichthyol.  (Engl.  Transl.  Vopr.  Ikhtiol.)  22(5):  16- 
23. 
-,  andA.  V.  Neyelov.   1978.  A  new  genus  and  species  of  eelpoul. 


Andriashevia  apiera  (Perciformes,  Zoarcidae).  J.  Ichthyol.  (Engl. 

Transl.  Vopr.  Ikhtiol.)  18:846-850. 
Fedotov,  V.  F.    1976.    Gadidae  of  the  Palaeogene-Neogene  from  the 

USSR.  Tr.  Paleonlol.  Inst.  Akad.  Nauk  SSSR  157. 
Fields,  H.  M.    1962.    Pompanos  (rrac/!;/io/;«  spp.)  of  South  Atlantic 

coast  of  the  United  States.  U.S.  Fish  Wildl.  Serv.  Fish.  Bull. 
FiERSTiNE,  H.  L.    1972.    The  paleontology  of  billfish  — the  state  of  the 

art,  p.  34-44.  In:  Proc.  Intematl.  Billfish  Symp.,  Kailua-Kona, 

Hawaii,  9- 1 2  August  1972,  Part  2.  Review  and  contributed  papers. 

R.  S.  Shomura  and  F.  Williams  (eds.).  NOAA  Tech.  Rep.  NMFS 

SSRF-675. 
Fink,  S.  V.,  and  W.  L.  Fink.     1981.    Interrelationships  of  the  ostar- 

iophysan  fishes  (Teleostei).  Zool.  J.  Linn.  Soc.  72:297-353. 
Fink,  W.  L.   1981.  Ontogeny  and  phylogenyoftooth  attachment  modes 

in  actinopterygian  fishes.  J.  Morphol.  167:167-184. 
.    1982.   The  conceptual  relationship  between  ontogeny  and  phy- 

logeny.  Paleobiology  8:254-264. 
,  K.  E.  Hartel,  W.  G.  Saul.  E.  M.  Moon  and  E.  O.  Wile'i  . 

MS.    A  report  on  current  supplies  and  practices  used  in  curation 

of  ichthyoplankton  collections.  Am.  Soc.  Ichthyol.  &  Herpetol. 

Mimeo. 
,  AND  S.  H.  Weitzman.    1982.    Relationships  of  the  stomiiform 


fishes  (Teleostei).  with  a  description  of  Diplophos.  Bull.  Mus.  Comp. 

Zool.  Harv.  Univ.  l50(2):31-93. 
FistHER,  W.  K.    1958a.    Primeras  fases  del  desarrollo  del  blanquillo 

{Prolatilus  jugidans)  Cuv.  et  Val.  (Pisces).  Rev.  Biol.  Mar.  8:3-24. 
.    1958b.    Huevos,  crias  y  primeras  prelarvas  de  la  "anchoveta" 

(Engraulis  ringens)  Jenyns.  Rev.  Biol.  Mar.  (Chile)  8:1  1  1-124. 
.    1959.   Huevos,  crias  y  prelarvas  de  la  merluza.  Rev.  Biol.  Mar. 

9(I,2,3):229-249. 
.    1963.   Die  Fische  des  Brackwassergebietes  Lenga  bei  Concep- 

cion  (Chile).  Int.  Rev.  Gesamten  Hydrobiol.  48:419-51 1. 
Fish,  M.  P.    1928.    Contributions  to  the  embryology  of  the  American 

eel  (Anguilta  rostrala  Lesueur).  Zoologica  (N.Y.)  8:289-324. 
.     1932.    Contributions  to  the  early  life  histories  of  sixty-two 

species  of  fishes  from  Lake  Erie  and  its  tributary  waters.  Bull.  Bur. 

Fish.  Wash.  47(10):293-398. 
FisHELSON,  L.    1963.  Observations  on  littoral  fishes  of  Israel.  II.  Larval 

development  and  metamorphosis  of  Blennius  pavo  (Risso)  (Te- 
leostei, Blenniidae).  Isr.  J.  Zool.  12:81-91. 
.    1975.    Ethology  and  reproduction  of  pteroid  fishes  found  in 

the  Gulf  of  Aqaba  (Red  Sea),  especially  Dendrochirus  brachyptenis 

(Cuvier),  (Pteroidae,  Teleostei).  Publ.  Staz.  Zool.  Napoli  39  (Suppl.); 

635-656. 
.    1976.   Spawning  and  larval  developrfient  of  the  blenniid  fish, 

Meiacanthus  nigrolineatus  from  the  Red  Sea.  Copeia  1976:798- 

800. 
FiTCH,  J.  E.     1964.    The  ribbonfishes  (Family  Trachipteridae)  of  the 

eastern  Pacific  Ocean,  with  a  description  of  a  new  species.  Calif 

Fish  Game  50:228-240. 
.     1966.    The  unicomfish,  Eumecichlhys  ftski  (Gunther)  in  the 

Eastern  Tropical  Pacific.  Calif  Fish  Game  52:208-210. 
.    1967.   The  tapertail  nbbon-fish  ( Trachipterus  fukuzakii  Fitch) 

added  to  the  marine  fauna  of  California.  Calif  Fish  Game  53:298- 

299. 
.    1982.    Revision  of  the  eastern  Nonh  Pacific  anthiine  basses 

(Pisces:  Serranidae).  Contrib.  Sci.  (Los  Ang.)  339. 
,  AND  L.  W.  Barker.    1972.    The  fish  family  Moridae  in  the 

eastern  North  Pacific  with  notes  on  morid  otoliths,  caudal  skeletons, 

and  the  fossil  record.  Fish.  Bull.  U.S.  70:565-584. 
,  AND  R.  J.  Lavenberg.    1968.    Deep-water  fishes  of  California. 

Univ.  Calif  Press,  Berkeley. 
,AND .    1975.  Tidepool  and  nearshore  fishes  of  California. 


Fives,  J.  M.  1970a.  BlenniidaeoftheNorth  Atlantic.  Cons.  Int.  Explor. 
Mer,  Fiches  Identification  Oeufs  Poissons  3. 

.     1970b.    Investigations  of  the  plankton  of  the  west  coast  of 

Ireland.  IV.  Larval  and  post-larval  stages  of  fishes  taken  from  the 
plankton  of  the  west  coast  in  surveys  during  the  years  1958-1966. 
Proc.  R.  Ir.  Acad.  Sect.  B  70:15-93. 

Flegler,  C.  1977.  Electron  microscopic  studies  on  the  development 
of  the  chorion  of  the  viviparous  teleost  Dennogenys  pusillus  (Hem- 
irhamphidae).  Cell  Tissue  Res.  179:255-270. 

Foersch,  W.  1966.  Spawning  Phractura  ansorgei.  Trop.  Fish  Hobby. 
1966  (April):4-6,  8-9. 

Follett,  W.  I.  1952.  Annotated  list  offishes  obtained  by  the  California 
Academy  of  Sciences  during  six  cruises  of  the  U.S.S.  Mulberry 
conducted  by  the  U.S.  Navy  of  Central  California  in  1949  and 
1950.  Proc.  Calif  Acad.  Sci.,  Ser.  4,  27:399-432. 

Ford,  E.  1922.  On  the  young  stages  of  B/enn/;«oce//am(L.),  Blennius 
pholis{L.).  and  Blennius  gallorugine(L.).  J.  Mar.  Biol.  Assoc.  U.K. 
12:688-692. 

FoREV,  P.  L.  1973a.  A  revision  of  elopiform  fishes,  fossil  and  recent. 
Bull.  Br.  Mus.  (Nat.  Hist.)  Geol.  Suppl.  10. 

.  1973b.  Relationships  of  elopomorphs.  p.  351-368. //;.  Inter- 
relationships of  fishes.  P.  H.  Greenwood,  R.  S.  Miles  and  C.  Pat- 
terson (eds.).  Zool.  J.  Linn.  Soc.  53(Suppl.  1.). 

FoRSTER,  R.  P.,  AND  S.  K.  HoNG.  1958.  In  vilro  transport  of  dyes  by 
isolated  renal  tubules  of  the  flounder  as  disclosed  by  direct  visu- 
alization: Intracellular  accumulation  and  transcellular  movement. 
J.  Cell.  Comp.  Physiol.  51:259-272. 

Foster,  N.  R.  1967.  Comparative  studies  on  the  biology  of  killifishes 
(Pisces:  Cyprinodontidae).  Unpubl.  PhD  dissertation.  Cornell  Uni- 
versity, NY. 

.    1968.    Utility  of  egg  and  larval  characters  in  the  classification 

of  atheriniform  fishes.  Am.  Zool.  8:338. 

.    1969.   Field  and  aquarium  observations  on  C/jr/opeoiWMpf «- 

gelleyi  Fowler,  1939.  the  Jamaican  killifish.  J.  Am.  Killifish  Assoc. 
6(1):  17-25. 

.    1973.    Behavior,  development,  and  early  life  history  of  the 

Asian  needlefish,  Xenenlodon  cancila.  Proc.  Acad.  Nat.  Sci.  Phila. 
I25(5):77-88. 

.  1974.  Cyprinodontidae,  p.  127-141.  In:  Manual  for  identi- 
fication of  early  developmental  stages  offishes  of  the  Potomac  River 
estuary.  A.  J.  Lippson  and  R.  L.  Moran  (eds.).  Power  Plant  Siting 
Program  of  Md.  Dep.  Nat.  Resour.  PPSP-MP-13. 

.    1982.    Literature  review.  J.  Am.  Killifish  Assoc.  15(1  ):4-9. 

,  AND  L.  FuiMAN.   MS.    Ecology,  behavior,  and  early  life  history 


of  the  lyre  goby,  Evorthodus  lyricus. 

Folrmanoir,P.  1970.  Notes  ichthyologiques(l).  Cah.  ORSTOM  Ser. 
Oceanogr.  8:19-33. 

.  1971a.  Notes  ichthyologiques  (3).  Cah.  ORSTOM  Ser.  Ocean- 
ogr. 9:267-278. 

.  1971b.  Notes  ichthyologiques  (4).  Cah.  ORSTOM  Ser.  Ocean- 
ogr. 9:491-500. 

.  1973.  Notes  ichthyologiques  (5).  Cah.  ORSTOM  Ser.  Ocean- 
ogr. 11:33-39. 

.    1976.    Formes  post-larvaires  et  juveniles  de  poissons  cotiers 


pris  au  chalut  pelagique  dans  le  sud-ouest  Pacifique.  Cah.  Pac.  19: 

47-48. 
Fowler,  H.  W.     1934.    Descriptions  of  new  fishes  obtained  1907  to 

1910,  chiefly  in  the  Philippine  islands  and  adjacent  seas.  Proc.  Acad. 

Nat.  Sci.  Phila.  85:233-267. 
.    1936.   Carangidae,  p.  675-756. /«;  The  marine  fishes  of  West 

Africa.  Bull.  Am.  Mus.  Nat.  Hist.  70(2). 
.    1944.    The  fishes,  p.  57-583.  In:  Results  of  the  Fifth  George 

Vanderbilt  expedition  (1941).  Monogr.  Acad.  Nat.  Sci.  Phila.  6. 
.     1945.    A  study  of  the  fishes  of  the  southern  Piedmont  and 


Univ.  Calif  Press,  Berkeley. 


coastal  plain.  Monogr.  Acad.  Nat.  Sci.  Phila.  7. 
Francis,  M.  P.     1981.    Meristic  and  morphometric  variation  in  the 
lancetfish,  Alepisaurus,  with  notes  on  the  distribution  of  A.  ferox 
and  A.  breviroslris.  N.Z.  J.  Zool.  8:403-408. 


704 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Frank,  K.  T.,  and  W.  C.  Legoett,    1983.    Multispecies  larval  fish 

associations:  Accident  or  adaptation?  Can.  J.  Fish.  Aquat.  Sci.  40: 

754-762. 
Frank,  S.    1972.    Entwicklung  der  Eier  des  Langflossensalmlers  Bry- 

conalestes longipinnis(GviiHheT.  1864)(Pisces:  Characinoidei).  Vestn. 

Cesk.  Spol.  Zool.  36:237-240. 
.    1978.    Beobachtung  der  Embryonalentwicklung.  Beispiel:  das 

Fischei.  Mikrokosmos  1978:251-255. 
Franke,  H.-J.    1979.    Algenbekampfung  ohne  Chemie,  Zucht  des  al- 

genfressenden  Antennenwelses  ,-l«a.s(n«  spec.  Aquarian  Mag.  1975: 

216-221. 
.    1981.    Cfe«o/i(aw.s  Aiyeto  (Valenciennes  in  Cuvier  and  Val- 
enciennes, 1849).  Aquarien  Terrarien  28:345-348. 
Fraser,  T.  H.    1971.   The  fish  Dinolestes  lewim  with  comments  on  its 

osteology  and  relationship.  Jpn.  J.  Ichthyol.  18:157-163. 
.    1972a.    Comparative  osteology  of  the  shallow  water  cardinal 

fishes  (PercifiDrmes:  Apogonidae)  wtih  reference  to  the  systematics 

and  evolution  of  the  family.  Rhodes  Univ.  J.  L.  B.  Smith  Inst. 

Ichthyol.  Bull.  No.  34. 
.    1972b.   Some  thoughts  about  the  teleostean  fish  concept  — the 

Paracanthopterygii.  Jpn.  J.  Ichthyol.  19:232-242. 
,  and  M.  M.  Smith.    1974.   An  exterilium  larval  fish  from  South 


Africa  with  comments  on  its  classification.  Copeia  1974:886-892. 

Fraser-Brunner,  A.  1931.  SomeinterestingWest  African  fishes,  with 
descriptions  of  a  new  genus  and  two  new  species.  Ann.  Mag.  Nat. 
Hist.,  Ser.  8(45):2 17-225. 

.    1949.   A  classification  of  the  fishes  of  the  family  Myctophidae. 

Proc.  Zool.  Soc.  Lend.  1  18(4):1019-1 106. 

.  1950.  Studies  in  plectognath  fishes  from  the  "Dana"  Expe- 
ditions. I.  An  interesting  new  genus  of  triacanthodid  fishes  from 
the  Celebes  Sea.  Dana-Rep.  Carlsberg  Found.  35:1-8. 

Freeman,  H.  W.  1951.  Contribution  to  the  evolution  and  classification 
of  the  fishes  of  the  family  Agonidae.  Unpubl.  PhD  dissertation. 
Stanford  Univ.,  Stanford,  CA. 

Freihofer,  W.  C.  1963.  Patterns  of  the  ramus  lateralis  accessorius 
and  their  systematic  significance  in  teleostean  fishes.  Stanford  Ich- 
thyol. Bull.  8:80-189. 

.    1970.    Some  nerve  patterns  and  their  systematic  significance 

in  paracanthopterygian,  salmoniform,  gobioid,  and  apogonid  fishes. 
Proc.  Calif  Acad.  Sci.,  Ser.  4.  38:215-264. 

.     1978.    Cranial  nerves  of  a  percoid  fish,  Polycentrus  schom- 


hrugkii  (family  Nandidae),  a  contribution  to  the  morphology  and 
classification  of  the  order  Perciformes.  Occas.  Pap.  Calif  Acad.  Sci. 
128. 

Fricke,  H.  W.  1980.  Mating  systems,  maternal  and  biparental  care 
in  triggerfish  (Balistidae).  Z.  Tierpsychol.  53(2):105-122. 

Fricke,  R.  1980.  Neue  Fundorte  und  noch  nicht  beschriebene  Ge- 
schlechtsunterschiede  einiger  Arten  der  Gattung  Calllonynnis  ( Pisces, 
Perciformes,  Callionymidae),  mil  Bemerkungen  zur  Systematik  in- 
nerhalb  dieser  Gattung  und  Beschreibung  einer  neuen  Untergattung 
und  einer  neuen  Art.  Ann.  Mus.  Civ.  Stor.  Nat.  Giacomo  Doria 
83:57-105. 

.  1981a.  Revision  of  the  genus  Synchirnpus  (Teleostei:  Cal- 
lionymidae). J.  Cramer,  Braunschweig,  F.R.G. 

.    1981b.   The  kaianus-gvouxt  of  the  genus  Callionymus  (Pisces: 

Callionymidae),  with  descriptions  of  six  new  species.  Proc.  Cal. 
Acad.  Sci.  42:349-377. 

Fridgeirsson,  E.  1978.  Embryonic  development  of  five  species  of 
gadoid  fishes  in  Icelandic  waters.  Rit  Fiskideildar  5. 

Fritzsche,  R.  a.  1978.  Development  of  fishes  of  the  Mid-Atlantic 
Bight.  An  atlas  of  egg,  larval  and  juvenile  stages.  Vol.  V.  Chaeto- 
donitidae  through  Ophidiidae.  U.S.  Fish  Wildl.  Ser.  Biol.  Serv. 
Prog.  FWS/OBS-78/12. 

.    1980.    Revision  of  the  eastern  Pacific  Syngnathidae  (Pisces: 

Syngnathiformes),  including  both  recent  and  fossil  forms.  Proc. 
Calif.  Acad.  Sci.  42(6):181-227. 

.  1982.  Osteichthyes,  p.  858-944.  In:  Synopsis  and  classifica- 
tion of  living  organisms.  S.  P.  Parker  (ed.).  McGraw-Hill.  Inc.,  New 
\ork. 


,  AND  G.  D.  Johnson.    1980.   Early  osteological  development  of 

white  perch  and  striped  bass  with  emphasis  on  identification  of 

their  larvae.  Trans.  Am.  Fish.  Soc.  109:387-406. 
Frost,  B.  W.,  and  L.  E.  McCrone.    1979.    Vertical  distribution,  diel 

vertical  migration  and  abundance  of  some  mesopelagic  fishes  in 

the  eastern  subarctic  Pacific  Ocean.  Fish.  Bull.  U.S.  76:751-770. 
FiiiMAN,  L.  A.    1976.   Notes  on  the  early  development  of  the  sea  raven, 

Hernitriplerus  amencanus.  Fish.  Bull.  U.S.  74:467-470. 
.    1979.   Descriptions  and  comparisons  of  catoslomid  fish  larvae: 

northern  Atlantic  drainage  species.  Trans.  Am.  Fish.  Soc.  108:560- 

603. 
.    1982a.   Correspondence  of  myomeres  and  vertebrae  and  their 

natural  variability  during  the  first  year  of  life  in  yellow  perch,  p. 

56-69.  In:  Proceedings  of  the  Fifth  Annual  Larval  Fish  Conference. 

C.  F.  Bryan,  J.  V.  Conner  and  F.  M.  Truesdale.  (eds).  Louisiana 

Cooperative  Fishery  Research  Unit,  Louisiana  State  L'niversily, 

Baton  Rouge,  LA. 
.    1982b.   Family  Esocidae,  pikes,  p.  155-193. //(.■  Identification 

of  larval  fishes  of  the  Great  Lakes  Basin  with  emphasis  on  the  Lake 

Michigan  drainage.  N.  A.  Auer  (ed.).  Ann  Arbor,  Michigan,  Great 

Lakes  Fish.  Comm.  Spec.  Publ.  82-3. 
.    1983a.    Analysis  of  egg  and  larval  characters  for  their  use  in 

the  systematics  of  ostariophysan  fishes.  Unpubl.  PhD  dissertation. 

University  of  Michigan.  Ann  Arbor,  Ml. 
.     1983b.    Growth  gradients  in  fish  larvae.  J.  Fish  Biol.  23(1): 

117-123. 
,  J.  V.  Conner,  B.  F.  Lathrop,  G.  L.  Buvnak,  D.  E.  Snyder 


AND  J.  J.  Loos.  1983.  Slate  of  the  art  of  identification  for  cyprinid 
fish  larvae  from  eastern  North  America.  Trans.  Am.  Fish.  Soc. 
112(2B):319-322. 

FijiTA,  K.  1971.  Early  development  of  the  squamation  in  Titapia 
sparrmanni.  Jpn.  J.  Ichthyol.  18:90-93. 

Fi^jiTA,  S.  1955.  On  the  development  and  prelarval  stages  of  the  file- 
fish,  Monacanlhus  cirrhifer  Temminck  et  Schlegel.  Sci.  Bull.  Fac. 
Agri.  Kyushu  Univ.  1 5(2):229-234.  (In  Japanese,  English  sum- 
mary.) 

.    1957a.    On  the  development  and  prelarval  stage  of  a  damsel 

fish,  Chromis  notatus  (Temminck  et  Schlegel).  Jpn.  J.  Ichthyol.  6: 
87-90. 

.     1957b.    On  the  larval  stages  of  a  scorpaenid  fish,  Sebastes 

pachycephalus  nigricans  (Schmidt).  Jap.  J.  Ichthyol.  6:91-93.  (In 
Japanese.  English  summary.) 

.    1958.  On  the  egg  development  and  larval  stages  of  a  viviparous 

scorpaenid  fish,  Sebastes  oblongns  Gunther.  Bull.  Jpn.  Soc.  Sci. 
Fish.  24:475-479.  (In  Japanese,  English  summary.) 

.    1959.   On  the  egg  development  of  a  viviparous  scorpaenid  fish 

Sebastes  pachycephalus  nigricans  (Schmidt).  J.  Jpn.  Assoc.  Zool. 
Gardens  and  Aquariums  l(2):42-43. 

.    1960.    Egg  development  and  prelarval  stages  of  a  silver-belly, 

Leiognathus  nuchalis  (Temminck  et  Schlegel).  Bull.  Jap.  Soc.  Sci. 
Fish.  26:1091-1094. 

.    1962.    Studies  on  the  life  history  of  common  pufferfishes  in 

Japan.  Rep.  Nagasaki  Prefect.  Fish.  Res.  Stn.  2.  (In  Japanese.) 

.  1965.  Early  development  and  reanng  of  two  common  flat- 
fishes, Eopsetta  gngorjewi  (Herzenstein)  and  Tanakius  kitaharai 
(Jordan  et  Starks).  Bull.  Jpn.  Soc.  Sci.  Fish.  31:258-262. 

.    1966.   Egg  development,  larval  stages  and  reanng  of  the  puffer. 


Lagocephalus  hinaris  spadiceus  (Richardson).  Jpn.  J.  Ichthyol.  13: 

162-168.  (In  Japanese,  English  summary.) 
,  AND  S.  MiTO.    1960.    Egg  development  and  hatched  larvae  of 

a  chaetodontid  fish,  Chaelodontoplus  septentnonalis  (Temminck  et 

Schlegel).  Bull.  Jap.  Soc.  Sci.  Fish.  26:227-29.  (In  Japanese,  English 

abstract.) 
— ,  AND  K.  Nakahara.    1955.  On  the  development  of  the  egg  and 

the  prelarval  stages  of  a  scorpaenid  fish,  Inimicusjapomcus  (Cuvier 

et  Valenciennes).  Sci.  Bull.  Fac.  Agric.  Kyushu  Univ.  15:223-228. 

(In  Japanese,  English  summary.) 
— ,  AND  T.  Takita.    1965.    Egg  development  and  prelarval  stages 


LITERATURE  CITED 


705 


of  a  sole,  Areliscus  trigrammus  (Gunthcr).  Bull.  Jpn.  Soc.  Sci.  Fish. 
31:488-492.  (In  Japanese,  English  abstract.) 

— ,  AND  K.  UcHiDA.  1957.  On  the  development  of  the  egg  and 
prelarval  stages  of  a  sole,  Cynoglossus  rohuslus  Gunther.  Sci.  Bull. 
Fac.  Agric.  Kyushu  Univ.  16:319-322.  (In  Japanese,  English  sum- 
mary.) 

— ,  AND  M.  Ueno.     1954.    On  the  development  of  the  egg  and 


prelarval  stages  of  Siganiis  fuscescens  (Houttuyn)  by  artificial  in- 
semination. Jpn.  J.  Ichthyol.  3:129-132. 
FuKL'HARA,  O.    1971.    Development  of  eggs  and  larvae  of  .-ljfra»;»w« 

agrammus.  Suisan  Yoshoka  19:159-165.  (In  Japanese.) 
,  andK.  Ito.    1978.  On  the  formation  of  the  fins  and  squamation 

in  the  Japanese  parrot  fish,  Opiegnathus  fascialus  reared  in  the 

laboratory.  Bull.  Nansei  Reg.  Fish.  Res.  Lab.  1  1:9-17. 
FnKUSHO,  K.     1975.    Development  of  color  pattern  in  Opiegnalhus 

fasciatus.  useful  indicator  in  rearing.  Jpn.  J.  Ichthyol.  22:23-30. 
Fulton,  T.  W.    1898.    The  ovaries  and  ovarian  eggs  of  the  angler  or 

frog-fish  (Lophius  piscalorius)  and  of  the  John  Dory  (Zeus  faher). 

Ann.  Rep.  Fish.  Board  Scotl.  16:125-134. 
Fi'RiiK.AWA,  I,    1951.    Clupeoid  fry  possibly  referable  to  Gonorynchiis 

ahbrcviatus  Temmmck  et  Schlegel.  Jpn.  J.  Ichthyol.  4:292-294.  (In 

Japanese,  English  summary.) 
FuSTER  DE  Plaza,  M.  L.,  and  J.  C.  Plaza.    1955.   Nuevos  ensayos  para 

obtener  la  reproduccion  artificial  de  las  percas  o  truchas  criollas 

(Percichthys  sp.).  Misc.  Publ.  Minist.  Agric.  Ganad.  (Buenos  Aires) 

407. 
FuTrH.  C.  R.    1971.    Ijarvae  of  Monolene  sessilicauda  Goodt,  1880 

(Bothidae).  Fla.  Dep.  Nat.  Resour.  Leafl.  Ser.  4  (Pi.  I,  No.  21). 
.    1977.   Larvae  of  Tnchopsetta  ventralis  (Pisces:  Bothidae),  with 

comments  on  intergeneric  relationships  within  the  Bothidae.  Bull. 

Mar.  Sci.  27:740-757. 
,  AND  F.  H.  HoFF,  Jr.    1971.    Larval  development  of  Syacn/m 

papillosum  (Bothidae)  with  notes  on  adult  morphology.  Fla.  Dep. 

Nat.  Resour.  Leafl.  Ser.  4  (Pt.  I,  No.  20). 
,  R.  W.  Topp  andE.  D.  HoLiDE.    1972.   Developmental  osteology 

of  the  lined  sole,  Achints  lineatus  (Pisces:  Soleidae).  Contrib.  Mar. 

Sci.  16:33-58. 
Gabriel.  M.  L.    1944.    Factors  affecting  the  number  and  form  of  ver- 
tebrae in  Fundiitus  heteroclUtis.  J.  Exp.  Zool.  95:105-143. 
Gamble,  J.  C,  AND  E.  D.  HouDE.    1984.  Growth,  mortality  and  feeding 

of  cod  (Gadus  morhua  L.)  larvae  in  enclosed  water  columns  and 

in  laboratory  tanks,  p.  123-143.  In:  The  propagation  of  cod.  E. 

Dahl,  D.  Danielssen,  E.  Moksness  and  P.  Solemdal  (eds.).  Flode- 

vigen  Rapp.  1. 
.  P.  Maclachlan.N.T.  Nicoll  AND  I.  G.Baxter.   1981.  Growth 

and  feeding  of  Atlantic  herring  reared  in  large  plastic  enclosures. 

Rapp.  P.  V.  Reun.  Cons.  Int.  Explor.  Mer  178:121-134. 
Garman,  S.    1892.    The  discoboli.  Mem.  Mus.  Comp.  Zool.  Harv. 

14(2):  1-96. 
Garnaud,  J.    I960.    La  ponte,  I'eclosion,  la  larve  du  baliste  Balistes 

caphscus  Linne  1758.  Bull.  Inst.  Oceanogr.  Monaco  1081:1-6. 
Garrett,  G.  P.    1982.   Variation  in  the  reproductive  traits  of  the  Pecos 

pupfish,  Cyprinodon  pecosensis.  Am.  Midi.  Nat.  180:355-363. 
Gatesoupe,  F.-J.,  AND  P.  LuQUET.   1981.   Practical  diet  for  mass  culture 

of  the  rotifer  Brachionus  plicatitis:  application  to  larval  rearing  of 

sea  bass,  Dicenlrarchus  labrax.  Aquaculture  22:149-163. 
,  AND  J.  Robin.     1981.    Commercial  single-cell  proteins  either 

as  sole  food  source  or  in  formulated  diets  for  intensive  and  con- 
tinuous production  of  rotifers  (Brachionus  plicalilis).  Aquaculture 

25:1-15. 
Galildie,  R.  W.,  E.  J.  Gravnoth  and  J.  Illinc^worth.     1980.    The 

relationship  of  the  iron  content  of  some  fish  otoliths  to  temperature. 

Comp.  Biochem.  Physiol.  66:19-24. 
Geffen,  A.    1982.   Otolith  nng  deposition  in  relation  to  growth  rate  in 

hemng  (Clupea  harengus)  and  turbot  (Scophlhalmus  maximus) 

larvae.  Mar.  Biol.  71:317-326. 
Gehringer,  J.  W.    1959a.    Early  development  and  metamorphosis  of 

the  ten-pounder  £/opiiaHrwi  Linnaeus.  U.S.  Fish  Wildl.  Serv.  Fish. 

Bull.  59(155):619-647. 


.     1959b.    Lcptocephalus  of  the  Atlantic  tarpon,  Megalops  al- 

lanticus  Valenciennes,  from  offshore  waters.  Q.  J.  Fla.  Acad.  Sci. 
2I(3):235-240. 

George,  A.,  and  V.  G.  Springer.  1980.  Revision  of  the  clinid  fish 
tribe  Ophiclinini,  including  five  new  species,  and  definition  of  the 
family  Clinidae.  Smithson.  Contrib.  Zool.  307:1-31. 

Gerking,  S.  D.,  and  R.  R.  Ralisch.  1979.  Relative  importance  of 
size  and  chronological  age  in  the  life  programme  of  fishes.  Arch. 
Hydrobiol.  13:181-194. 

Gery.  J.  1977.  Characoids  of  the  World.  T.F.H.  Publications,  Inc., 
Neptune  City,  NJ. 

GiBBS,  R.  H.,  Jr.  1959.  Asynopsisofthepost  larva  of  western  Atlantic 
lizard-fishes  (Synodontidae).  Copeia  1959:232-236. 

.  1964a.  Family  Astronesthidae,  p.  1-10. /ai.  Fishes  of  the  west- 
em  North  Atlantic.  H.  B.  Bigelow  (ed.).  Mem.  Sears  Found.  Mar. 
Res.  No.  1  (Pt.  4). 

.    1964b.    Family  Idiacanthidae,  p.  512-522.  In:  Fishes  of  the 

Western  North  Atlantic.  H.  B.  Bigelow  (ed.).  Mem.  Sears  Found. 
Mar.  Res.  No.  1  (Pt.  4). 

,  T.  A.  Clarke  and  J.  R.  Gomon.  1983.  Taxonomy  and  dis- 
tribution of  the  stomioid  fish  Genus  Eustomias  (Melanostomiidae), 
I:  Subgenus  Nominostomias.  Smithson.  Contrib.  Zool.  No.  380. 

,  and  B.  B.  Collette.    1 959.  On  the  identification,  distribution, 

and  biology  of  the  dolphins  Coryphaena  hippunis  and  C.  equiselis. 
Bull.  Mar.  Sci.  GulfCanbb.  9:117-152. 

,  and .    1967.    Comparative  anatomy  and  systematics  of 

the  tunas,  genus  Thunnus.  U.S.  Fish.  Wildl.  Serv.  Fish.  Bull.  66: 
65-130. 
-.  and  J.  C.  Staiger.    1970.   Eastern  tropical  Atlantic  flyingfishes 


of  the  genus  Cypseturus  (Exocoetidae).  Stud.  Trop.  Oceanogr. 
(Miami)  No.  4(Pt.  2):432-466. 

Gilbert,  C.  H.  1893.  Scientific  results  of  explorations  by  the  U.S. 
Fish  Commission  Steamer  Albatross.  No.  XXII.  Descriptions  of 
thirty-four  new  species  of  fishes  collected  in  1888  and  1889,  prin- 
cipally among  the  Santa  Barbara  Islands  and  in  the  Gulf  of  Cali- 
fornia': Proc.  U.S.  Natl.  Mus.  l4(880):539-566. 

.    1905.   The  deep  sea  fishes  of  the  Hawaiian  Islands.  Bull.  U.S. 

Fish  Comm.  23(2):577-713. 

,  AND  C.  V.  Burke.  1912.  Fishes  from  Bering  Sea  and  Kam- 
chatka. Bull.  Bur.  Fish.  Wash.  30(19101:31-96. 

,  AND  C.  L.  HuBBS.    1916.    Report  on  the  Japanese  macrourid 

fishes  collected  by  the  United  States  fisheries  steamer  "Albatross" 
in  1906,  with  a  synopsis  of  the  genera.  Proc.  U.S.  Natl.  Mus.  51(2149): 
135-214. 

,  AND  W.  F.  Thompson.     1916.    Family  Macrouridae,  p.  471- 

476.  In:  Fishes  collected  by  the  United  States  Bureau  of  Fisheries 
steamer  "Albatross"  during  1888,  between  Montevideo,  Uruguay, 
and  Tome,  Chile,  on  the  voyage  through  the  Straits  of  Magellan. 
W.  F.  Thompson  (ed.).  Proc.  U.S.  Natl.  Mus.  50(2133):40l-476. 

Gilbert,  C.  R.  1980.  Umbra  limi.  Umbra  pygmaea.  p.  129-130.  In: 
Atlas  of  North  Amencan  freshwater  fishes.  D.  S.  Lee,  C.  R.  Gilbert, 
C.  H.  Hocutt,  R.  E.  Jenkins,  D.  E.  McAllister  and  J.  R.  StaufTer, 
Jr.  (eds.).  North  Carolina  Mus.  Nat.  Hist.,  Raleigh,  NC. 

Gilchrist,  J.  D.  F.  1903.  The  development  of  South  -African  fishes. 
Part  1.  Mar.  Invest.  S.  Afr.  2:181-201. 

.    1904.   The  development  of  South  African  fishes.  Part  2.  Mar. 

Invest.  S.  Afr.  3:131-150. 

.    1918.  The  eggs  and  spawning-habits  ofthe  pilot  fish  (.Vawcrarw 

ductor).  Ann.  Mag.  Nat.  Hist.,  Ser.  9,  2(7):  1 14-1 18. 
-.  AND  H.  Hunter.    1919.    Reproduction  of  fishes  in  Table  Bay. 


Trans.  R.  Soc.  S.  Afr.  8(1):  1-1 6. 
Gill,  T.  N.    1862.    Synopsis  of  the  polynematoids.  Proc.  Acad.  Nat. 

Sci.  Phila.  13:271-282. 
.    1863.    Descriptive  enumeration  of  a  collection  of  fishes  from 

the  western  coast  of  Central  America.  Proc.  Acad.  Nat.  Sci.  Phila. 

15(4):  153-200. 
.    1865.  Synopsis  ofthecyclopteroidsofeastem  North  America. 

Proc.  Acad.  Nat.  Sci.  Phila.  16:189-194. 


706 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


— .  1872.  Arrangement  of  the  families  of  fishes,  or  classes  Pisces, 
Marsipobranchii  and  Leptocardii.  Smilhson.  Misc.  Collect.  1 1:247. 

— .  1888.  On  the  classification  of  the  mail-cheeked  fishes.  Proc. 
U.S.  Natl.  Mus.  11:567-592. 

— .  1891.  On  the  relations  of  Cyclopteroidea.  Proc.  U.S.  Natl. 
Mus.  13:361-376. 

— .    1893.    A  comparison  of  antipodal  faunas.  Mem.  Natl.  Acad. 


Sci.  6:91-124. 

GiNSBURG.  I.  1952.  Fishes  of  the  family  Carangidae  of  the  northern 
Gulf  of  Mexico  and  three  related  species.  Publ.  Inst.  Mar.  Sci.  Univ. 
Tex.  2:43-117. 

.    1955.    Fishes  of  the  family  Percophididae  from  the  coasts  of 

eastern  United  States  and  the  West  Indies,  with  descriptions  of  four 
new  species.  Proc.  U.S.  Natl.  Mus.  l04(3347):623-649. 

GiNZBURG,  A.  S.  1972.  Fertilization  in  fishes  and  the  problem  of 
polyspermy.  Acad.  Sci.  USSR,  Inst.  Develop.  Biol.  1968.  Trans- 
lated from  Russian  by  the  Isr.  Program  Sci.  Transl. 

GJ0SAETER,  J.  1973.  Age,  growth  and  mortality  of  the  myctophid  fish, 
Benthosema  glaciale  {Reinhardi),  from  western  Norway.  Sarsia  52: 
1-14. 

,  AND  K.  Kawaguchi.    1980.    A  review  of  the  world  resources 

of  mesopelagic  fish.  FAO  Fish.  Tech.  Pap.  193. 

GoiN,  C.  J.,  AND  D.  S.  Erdman.  195 1 .  The  crested  oarfish,  Lophotus 
lacepedei.  from  Florida;  first  record  from  the  western  North  Atlan- 
tic. Copeia  1951:285-288. 

Gold,  J.  R.  1979.  Cytogenetics,  p.  353-405.  //;.  Fish  Physiology.  W. 
S.  Hoar,  D.  J.  Randall  and  J.  R.  Brett  (eds.).  Academic  Press,  New 
York. 

Golvan,  Y.-J.  1962.  Catalogue  systematique  des  noms  de  genres  de 
poissons  actuels.  Ann.  Parasitol.  Hum.  Comp.  37(6):  1-227. 

Gomez  Gaspar,  A.  1981.  Observaciones  taxonomicas  y  ecologicas  de 
estados  larvales  Elopiformes  en  la  Laguna  de  la  Restinga  y  areas 
proximas  (Isla  Margarita,  Venezuela).  Bol.  Mus.  Mar.  Univ.  Bogota 
10:3-23. 

Gonzales,  T.  1946.  Contribute  alia  conoscenza  dello  sviluppo  post- 
embrionale  in  Microichlhys  coccoi.  Ruppell.  Boll.  Pesca  Pisci.  Id- 
robiol.  1:39-46. 

GooDE,  G.  B.,  AND  T.H.  Bean.  1896.  Oceanic  Ichthyology.  U.S.  Natl. 
Mus.  Spec.  Bull. 

Goody,  P.  C.  1969.  The  relationships  of  certain  Upper  Crealaceous 
teleosts  with  special  reference  to  the  myctophoids.  Bull.  Bnt.  Mus. 
(Nat.  Hist.)  Geol.  Suppl.  7. 

Goodyear,  R.  H.  1970.  A  new  species  of  .-l^a-vo/epw,  a  bathypelagic 
fish  from  the  Gulf  of  Panama  (Pisces,  Lampridiformes,  Megalo- 
mycteridae).  Steenstrupia  1(3):  17-20. 

GopiNATH,  K.  1946.  Notes  on  the  larval  and  post-larval  stages  of  fishes 
found  along  the  Trivandrum  coast.  Proc.  Indian  Natl.  Sci.  Acad. 
12(l):7-21. 

GoRBUNOVA,  N.  N.  1954.  The  reproduction  and  development  of  the 
walleye  pollock,  Theragra  chalcogramma  (Pallas).  Tr.  Inst.  Okean- 
ol.  Akad.  Nauk  SSSR  11:132-195. 

.    1 964a.   Breeding  and  development  of  hemilepidotine  sculpins 

(Cottidae,  Pisces),  p.  249-266.  In:  Fishes  of  the  Pacific  and  Indian 
Oceans.  Biology  and  distribution.  T.  S.  Rass  (ed.).  (Isr.  Program 
Sci.  Transl.  No.  1411.) 

.    1 964b.   Spawning  and  development  of  greenlings  (family  Hex- 

agrammidae),  p.  121-185.  In:  Greenlings:  taxonomy,  biology,  in- 
teroceanic  transplantation.  T.  S.  Rass  (ed.).  Springfield,  Va.  (Transl. 
available  from:  NTIS,  TT69-55097.) 

.    1981.   Larvae  of  the  genus  (7Hf/^/(>rna(Gonostomatidae)  with 

keys.  J.  Ichthyol.  (Engl.  Transl.  Vopr.  Ikhtiol.)  21(4):I38-141. 

.  1982.  Larvae  of  trichiuroid  fishes  from  collection  of  the  In- 
ternational Mexican  Biological  Center  (Gempylidac,  Trichiuridae, 
Pisces).  Proc.  P.  P.  Shirshov  Inst.  Oceanol.  1  18:85-104.  (In  Rus- 
sian.) 

— ,  AND  N.  V.  Parin.    1963.    Development  of  eggs  and  larvae  of 

the  flying  fish,  Cheilopogon  unicolor  (Cuv.  et  Val.).  Tr.  Inst.  Okean- 
ol.  Akad.  Nauk  SSSR  62:62-67.  (In  Russian,  English  abstract.) 

,  AND  O.  A.  Zvyagina.    1975.    Eggs  and  larvae  of  the  scaled 


sardine  (Harengiila  pensacolae)  (Pisces,  Clupeidae).  J.  Ichthyol. 

(Engl.  Transl.  Vopr.  Ikhtiol.)  15:829-833. 
Gordon,  D.  J.    1982.    Systematics  and  distribution  of  larval  fishes  of 

the  subfamily  Ophidiinae  (Pisces,  Ophidiidae)  in  the  eastern  Gulf 

of  Mexico.  Unpubl.  MS  thesis.  Univ.  Miami,  Coral  Gables,  FL. 
GosLiNE,  W.  A.    1948.    Some  possible  uses  of  X-rays  in  ichthyology 

and  fishery  research.  Copeia  1948:58-61. 
.    1951.    The  osteology  and  classification  of  the  ophichthid  eels 

of  the  Hawaiian  Islands.  Pac.  Sci.  5:298-320. 
.    1955.  Theosteology  and  relationships  of  certain  gobioid  fishes, 

with  particular  reference  to  the  genera  Kraemaria  and  Microdes- 

mus.  Pac.  Sci.  9:158-170. 
.    1960a.    Contributions  toward  a  classification  of  modem  iso- 

spondylous  fishes.  Bull.  Br.  Mus.  Nat.  Hist.  Zool.  6:325-365. 
.    1 960b.    Hawaiian  lava-flow  fishes.  Part  IV.  Snyderidia  canina 

Gilberi,  with  notes  on  the  osteology  of  ophidioid  families.  Pac.  Sci. 

14:373-381. 
.    1961.   Some  osteological  features  of  modem  lower  teleostean 

fishes.  Smithson.  Misc.  Collect.  142:1-42. 
.    1 962.   Systematic  position  and  relationships  of  the  percesocine 

fishes.  Pac.  Sci.  16:207-217. 
.    1963a.   Considerations  regarding  the  relationships  of  the  Per- 

copsiform,  Cyprinodontiform,  and  Gadiform  fishes.  Occas.  Pap. 

Mus.  Zool.  Univ.  Mich.  No.  629. 
.     1963b.    Notes  on  the  osteology  and  systematic  position  of 


Hypoptychus  dybowskii  Steindachner  and  other  elongate  perciform 

fishes.  Pac.  Sci.  17:90-101. 
.    1966.   The  limits  of  the  fish  family  Serranidae,  with  notes  on 

other  lower  percoids.  Proc.  Calif  Acad.  Sci.  33:91-1 1 1. 
— .    1968.  The  suborders  of  perciform  fishes.  Proc.  U.S.  Natl.  Mus. 

124(3647):  1-78. 
— .    1970.    A  reinterpretation  of  the  teleostean  fish  order  Gobie- 

sociformes.  Proc.  Cal.  Acad.  Sci.  38:363-382. 
.    1971.    Functional  morphology  and  classification  of  teleostean 

fishes.  Univ.  Press  Hawaii,  Honolulu. 
— .    1978.   Unbranched  dorsal-fin  rays  and  subfamily  classification 

in  the  fish  family  Cyprinidae.  Occas.  Pap.  Mus.  Zool.  Univ.  Mich. 

684. 
.    1980.   The  evolution  of  some  structural  systems  with  reference 

to  the  interrelationships  of  modem  lower  Teleostean  fish  groups. 

Jpn.  J.  Ichthyol.  27:1-28. 
,  N.  B.  Marshall  and  G.  W.  Mead.     1966.    Order  Iniomi: 


characters  and  synopsis  of  the  families.  Mem.  Sears  Found.  Mar. 
Res.  21(Pt.  5):l-29. 

GoszTON-i  I,  A.  E.  1977.  Results  of  the  research  cruises  of  FRV  "Wal- 
ther  Herwig"  to  South  America.  XLVIII.  Revision  of  the  South 
American  Zoarcidae  (Osteichthyes.  Blennioidei),  with  the  descrip- 
tion of  three  new  genera  and  five  new  species.  Arch.  Fischereiwiss. 
27:191-249. 

GoTSHALL,  D.  W.  1971.  Occurrences  of  the  bearded  eelpout,  Lyco- 
nema  barbatum.  off  northern  Califomia  and  southem  Oregon.  Cal- 
if Fish  Game  57:295-297. 

Gould,  W.  R.  1965.  The  biology  and  morphology  oi  Acyrtops  ber- 
ylUnus.  the  emerald  clingfish.  Bull.  Mar.  Sci.  Gulf  Canbb.  15:165- 
188. 

Gould,  S.  J.  1977.  Ontogeny  and  phylogeny.  Belknap  Press,  Cam- 
bridge, MA. 

GovoNi.J.  J.  1980.  Morphological,  histological  and  functional  aspects 
of  alimentary  canal  and  associated  organ  development  in  larval 
Leiostomus  xanthurus.  Rev.  Can.  Biol.  39:69-80. 

,  J.  E.  Olnev.  D.  F.  Markle  and  W.  R.  Curtsinger.    1984. 

Observations  on  stracture  and  evaluation  of  possible  functions  of 
the  vexillum  in  larval  Carapidae  (Ophidiiformes).  Bull.  Mar.  Sci. 
34:60-70. 

Graham,  D.  H.  1939.  Breeding  habits  of  fishes  of  Otago  Harbour  and 
adjacent  seas.  Trans.  R.  Soc.  N.Z.  69:361-372. 

Grassi,  G.  B.  1913.  Metamorfosi  dei  murenoidi.  Ricerche  sistema- 
tische  ed  ecologische.  Fischer,  Jena. 

,  and  S.  a.  Calandruccio.     1897.    Descrizione  d"un  Lepra- 


LITERATURE  CITED 


707 


cephalus  breviroslris  in  via  tranformarzi  in  Anguilta  vulgaris.  Atti 
Accad.  Naz.  Lincei,  Ser.  6.  5(Pt.  l):239-240. 

Graves,  J.  E.,  and  G.  N.  Somero.  1982.  Electrophoretic  and  func- 
tional enzymic  evolution  in  four  species  of  eastern  Pacific  barra- 
cudas from  different  thermal  environments.  Evolution  36:97-106. 

Greenfield,  D.W.  1965.  Systematics  and  zoogeography  of  .V/vnpm«i 
in  the  eastern  tropical  Pacific.  Calif  Fish  Game  51:229-247. 

.    1972.   Notes  on  the  biology  of  the  arrow  blenny  L!<cav'aWe«- 

nius  zingaro  (Bohlke)  from  Bntish  Honduras.  Copeia  1972:591- 
592. 
-.    1974.    A  revision  of  the  squirrelfish  genus  AAnprH/ii  Cuvier 


(Pisces:  Holocentridae).  Nat.  Hist.  Mus.  Los  Ang.  Cty.  Sci.  Bull. 
19. 

Greenwood,  P.  H.    1955.    Reproduction  in  the  catfish,  C/ariai  wo5- 
sambicus  Peters.  Nature  (Lond.)  176:516-519. 

.     1956.    The  reproduction  of  Clarias  mossambicus  Peters  in 

Uke  Victoria.  Sci.  Counc.  Afr.  South  Sahara  Publ.  25:77-78. 

.    1973.    Interrelationships  of  osteoglossomorphs,  p.  307-332. 

In:  P.  H.  Greenwood.  R.  S.  Miles  and  C.  Patterson  (eds.).  Inter- 
relationships of  fishes.  Academic  Press,  New  York. 

.    1976.    A  review  of  the  family  Centropomidae  (Pisces,  Perci- 

formes).  Bull.  Bnt.  Mus.  (Nat.  Hist.)  Zool.  29:1-81. 

.    1977.   Notes  on  the  anatomy  and  classification  of  elopomorph 

fishes.  Bull.  Br.  Mus.  (Nat.  Hist.)  Zool.  32(4):65-102. 

,G.S.  Myers,  D.E.Rosen  AND  S.  H.Weitzman.    1967.   Named 

main  divisions  of  teleostean  fishes.  Proc.  Biol.  Soc.  Wash.  80:227- 
228. 

,  AND  D.  E.  Rosen.    1971.    Notes  on  the  structure  and  relation- 
ships of  the  alepocephaloid  fishes.  Am.  Mus.  Novit.  2473. 

-,  S.  H.  Weitzman  and  G.  S.  Myers.    1966.    Phyletic 


studies  of  teleostean  fishes,  with  a  provisional  classification  of  living 
forms.  Bull.  Amer.  Mus.  Nat.  Hist.  131:339-456. 

Gregory,  W.  K.  1933.  Fish  skulls.  A  study  of  the  evolution  of  natural 
mechanisms.  Trans.  Am.  Philos.  Soc.  23:75-481. 

.  1959.  Fish  skulls.  A  study  of  the  evolution  of  natural  mech- 
anisms. Noble  Offset  Printers,  Inc.,  New  York.  (First  published  in 
1933.  Trans.  Am.  Philos.  Soc.  23:75^81.) 

,  and  G.  M.  Conrad.    1936.    Pictorial  phylogenies  of  deep-sea 

isopondyli  and  iniomi.  Copeia  1936:21-36. 

,AND .   1937.  The  comparative  osteology  of  the  swordfish 

(Xiphias)  and  the  sailfish  (Istiophorm).  Am.  Mus.  Novit.  952. 

,  AND  .     1943.    The  osteology  of  Liivarus  impenalis.  a 

scombroid  fish:  a  study  in  adaptive  evolution.  Bull.  Am.  Mus.  Nat. 
Hist.  31(2):225-283. 

Grey,  M.  1953.  Fishes  of  the  family  Gempylidae,  with  records  of 
Nesiarchus  and  Epirmula  from  the  western  Atlantic  and  descrip- 
tions of  two  new  subspecies  of  Epwnula  onenlalis.  Copeia  1953: 
135-141. 

.    1955a.   Notes  on  a  collection  of  Bermuda  deepsea  fishes.  Field- 

iana  Zool.  37:265-302. 

.  1955b.  The  fishes  of  the  genus  Telragonurus  Risso.  Dana- 
Rep.  Carlsberg  Found.  4 1 . 

.    1 960.    Description  of  a  western  Atlantic  specimen  of  Scom- 

brolabrax  helerolepis  Roule  and  notes  on  fishes  of  the  family  Gem- 
pylidae, Copeia  1960:210-215. 

1964.    Family  Gonostomatidae,  p.  78-240.  In:  Fishes  of  the 


western  North  Atlantic.  H.  B.  Bigelow  (ed).  Mem.  Sears  Found. 
Mar.  Res.  No.  1  (Pt.  4). 

Grioorev,  G.  v.,  AND  V.  P.  Serebri  AKOv.  1981.  Eggs  of  roundnose 
grenadier,  Coryphaenoides  rupestns  Gunnerus  1765.  J.  Northwest 
Atl.  Fish.  Sci.  2:73-74. 

Grimes,  P.  J.,  and  D.  A.  Robertson.  1981.  Egg  and  larval  develop- 
ment of  the  silver  warehou,  Senolella  punclala  (Pisces:  Centrolo- 
phidae).  N.Z.  J.  Mar.  Freshwater  Res.  15:261-266. 

Grimm,  H.  1979a.  Egg  size  and  fecundity  within  the  genus  .-Ip/wn/iw 
(Pisces,  Cyprinodontidae).  Meersforsch.  Rep.  Mar.  Res.  27(3):  177- 
185. 

.     1979b.    Investigations  of  the  evolution  of  differently  scaled 

populations  found  in  the  species  Aphanius  analoliac  (Leidenfrost, 


1912)  (Pisces,  Cyprinodontidae).  Genetics  of  the  egg  size.  Z.  Zool. 

Syst.  Evolutionsforsch.  17(l):66-77. 
Grindley,  J.  R..  AND  M.  J.  Penrith.    1965.   Notes  on  the  bathypelagic 

fauna  of  the  seas  around  South  Africa.  Zool.  Afr.  l(2):275-295. 
GiiDOER,  E.  W.    1905.    The  breeding  habits  and  the  segmentation  of 

the  egg  of  the  pipefish  {Siphosloma  Jloridae).  Proc.  U.S.  Natl.  Mus. 

29(1431):447-500. 
.   1912.  Oral  gestation  in  the  gaff-topsail  cat-fish,  Felichthysfeiis. 

Science  35:192. 
.     1916.    The  gaff-topsail  (Felichthys  felts):  A  sea  catfish  that 

carries  its  eggs  in  its  mouth.  Zoologica  (N.Y.)  2(5):  125-1 58. 
.    1918.   Oral  gestation  in  the  gaff-topsail  catfish,  Felichthysfeiis. 

Carnegie  Inst.  Wash.  Publ.  252:25-52. 
.    1926.    A  study  of  the  smallest  shark-suckers  (Echeneidae)  on 

record,  with  special  reference  to  metamorphosis.  Am.  Mus.  Novit. 

234:1-26. 
.    1928.    The  smallest  known  specimens  of  the  sucking-fishes, 

Reinora  brachyptera  and  Rhombochirus  osleochir.  Am.  Mus.  Nov- 
it. 294:1-5. 
.    1930.   The  opah  or  moonfish,  Lamprw /!<«a.  on  the  coasts  of 

California  and  of  Hawaii.  Am.  Nat.  65:531-541. 

1931.    The  tnple-tail,  Lobotes  surinamensis.  its  names,  oc- 


currence on  our  coasts  and  its  natural  history.  Am.  Nat.  65:49-69. 
Gltllard.  R.  R.  L.    1975.   Culture  of  phytoplankton  for  feeding  marine 

invertebrates,  p.  29-60.  /«.  Culture  of  marine  invertebrate  animals. 

W.  L.  Smith  and  M.  H.  Chanley  (eds.).  Plenum  Publ.  Corp.,  N.Y. 
Glitart  Manday,  D.,  and  D.  Juarez  Fernandez.    1966.    Desarrollo 

embrionario  primeros  estadios  larvales de  chema criolla,  Epinephe- 

lus  stnalus  (Bloch)  (Perciformes:  Serranidae).  Acad.  Cienc.  Cuba 

Estud.-Inst.  Oceanol.  1:35-45. 
Guitel,  R.    1888.   Recherches  sur  les  Ledapogasters.  Arch.  Zool.  Exp. 

Gen.,  Ser.  2.  6:423-647. 
GiiNTER,  G.,  L.  L.  Sulya  and  B.  E.  Box.    1961.    Some  evolutionary 

patterns  in  fishes'  blood.  Biol.  Bull.  (Woods  Hole)  121:302-306. 
GOnther,  A.    1861.    Catalogue  of  the  Acanthopterygian  fishes  in  the 

collection  of  the  British  Museum.  Vol.  3^.  Bntish  Museum,  Lon- 
don. 
.    1862.    Chromides.  p.  264-316.  //;.  Catalogue  of  the  fishes  in 

the  British  Museum.  Vol.  4. 
.    1887.   Report  on  the  deep-sea  fishes  collected  by  H. M.S.  Chal- 
lenger dunng  the  years  1873-76.  Rep.  Sci.  Results  Challenger,  Zool. 

22:1-335.  (Repnnt  1963.  Wheldon  and  Wesley,  Ltd.,  and  Hafner 

Publishing  Co.,  New  York.) 
.    1889.    Report  on  the  pelagic  fishes  collected  by  H. M.S.  Chal- 
lenger during  the  years  1873-1876.  Rep.  Sci.  Results  Challenger, 

Zool.  31:1-44. 
Gustato,  G.    1976.   Osservazioni  suUa  biologia  e  sul  comportamento 

di  Carapus  acus  (Ophidioidea,  Percomorphi).  Boll.  Soc.  Nat.  Na- 

poli  85:505-535. 
GuTHERZ,  E.  J.    1970.    Characteristics  of  some  larval  bothid  flatfish. 

and  development  and  distribution  of  larval  spotfin  flounder,  Cy- 

dopsetta  fimbnata  (Bothidae).  U.S.  Fish.  Wild.  Serv.  Fish.  Bull. 

68:261-283. 
Haak£,  P.  W.,  C.  A.  Wilson  and  J.  M.  Dean.    1982.   A  technique  for 

the  examination  of  otoliths  by  SEM  with  application  to  lars'al  fishes. 

In:  Proc.  5th  Annual  Larval  Fish  Conference,  Baton  Rouge,  LA. 
H^EDRiCH,  R.  L.     1967.    The  stromateoid  fishes:  systematics  and  a 

classification.  Bull.  Mus.  Comp.  Zool.  Harv.  Univ.  135:31-139. 
.    1969.   A  new  family  of  stromateoid  fishes  from  the  equatorial 

Indo-Pacific.  Dana-Rep.  Carlsberg  Found.  76. 
.    1970.    A  new  species  in  the  nomeid  fish  genus  Psenes  from 

the  equatorial  eastern  Pacific.  Breviora  351. 
.    1971.  Theponsmoultoni:asignificantcharacter.  Copeia  1971: 

167-169. 

.    1974.    A  rotating  eye  in  rraf/i/p/pnYi.  Copeia  1974:972-973. 

,  and  M.  H.  Horn.     1972.    A  key  to  the  stromateoid  fishes. 

Woods  Hole  Oceanogr.  Inst.  Tech.  Rep.  72-15. 
,  and  J.  G.  Nielsen.    1966.   Fishes  eaten  by  .-l/ep/iai/riu  (Pisces, 


708 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Iniomi)  in  the  southeastern  Pacific  Ocean.  Deep-Sea  Res.  13:909- 
919. 

Haegele,  C.  W.,  AND  J.  F.  ScHWEiGERT.  1984.  Distribution  and  char- 
acteristics of  herring  spawning  grounds  and  description  of  spawning 
behavior.  Can.  J.  Spec.  Publ.  Fish  Aquat.  Sci.  In  press. 

Haigh,  E.  H.  1972a.  Larval  development  of  three  species  of  econom- 
ically important  South  African  fishes.  Ann.  S.  Afr.  Mus.  59(3):47- 
70. 

.     1972b.    Development  of  Trachurus  Irac/ninis  (Carangidae), 

the  south  African  maasbanker.  Ann.  S.  Afr.  Mus.  59(8):  139- 149. 

Hallacher,  L.  E.  1974.  The  comparative  morphology  of  extrinsic 
gasbladder  musculature  in  the  scorpionfish  genus  Sehastes  (Pisces: 
Scorpaenidae).  Proc.  Calif  Acad.  Sci.  40:59-86. 

Halstead,  B.  W.,  D.  D.  Danielson.  W.  J.  Baldwin  and  P.  C.  Engen. 
1972.  Morphology  of  the  venom  apparatus  of  the  leatherback  fish 
Scomberoides  sanclipeln  (Cuvier).  Toxicon  10:249-258. 

Hamanda,  K.  1961.  Taxonomic  and  ecological  studies  of  the  genus 
Hypomesus  of  Japan.  Mem.  Fac.  Fish.  Hokkaido  Univ.  9(1):  1-56. 

Hansen,  E.,  AND  P.  J.  Herring.  1977.  Dual  bioluminescent  systems 
in  the  anglerfish  genus  Ltnophryne.  J.  Zool.  (Lond.)  182:103-124. 

Hara,  S.,  J.  P.  Canto,  Jr.,  and  J.  M,  E.  Almendras.  1982.  A  com- 
parative study  of  various  extenders  for  milkfish,  Chanos  chanos 
(Forsskal),  sperm  preservation.  Aquaculture  28:339-346. 

Harada,  T.  1962.  A  contribution  to  the  biology  of  the  black  rockfish, 
Sehastes  inermisCu\\er  ex  Valenciennes.  Publ.  Seto  Mar.  Biol.  Lab. 
10:307-361. 

,    K.   MiZlJNO,   O.    MURATA,   S.    MlVASHITA   AND   H.    HURLITANI. 

1971.  On  the  artificial  fertilization  and  rearing  of  larvae  in  yel- 
lowfin  tuna.  Mem.  Fac.  Agnc.  Kinki  Univ.  (4):1 45-151. 

,  O.  MuRATA  AND  H.  FuRHTANi.  1973.  On  the  artificial  fertil- 
ization and  rearing  of  larvae  in  Marusoda,  Auxis  tapeinosoma. 
Mem.  Fac.  Agric.  Kinki  Univ.  (6):1 13-1 16. 

, ANt5  S.  MlVASHITA.    1973.   On  the  artificial  fertilization 

and  rearing  of  larvae  in  Hirasoda,  Auxis  Ihazard.  Mem.  Fac.  Agric. 
Kinki  Univ.  (6):  109-1  12. 

, AND .    1974.    On  the  artificial  fertilization  and 


rearing  of  larvae  in  bonito.  Mem.  Fac.  Agric.  Kinki  Univ.  (7):l-4. 
Harbison,  G.  R.,  L.  P.  Madin  and  N.  R.  SwANBERCi.    1978.    On  the 

natural  history  and  distribution  of  oceanic  ctenophores.  Deep-Sea 

Res.  25:233-256. 
Hardy,  G.  S.    1982.    A  revision  of  the  fishes  of  the  family  Pentacer- 

otidae  (Perciformes).  N.Z.  J.  Zool.  10:177-220. 
Hardy,  J.  D.,  Jr.    1978a.   Development  of  fishes  of  mid-Atlantic  bight, 

an  atlas  of  egg,  larval,  and  juvenile  stages.  Vol.  II.  Anguillidae 

through  Syngnathidae.  U.S.  Fish  Wildl.  Serv.  Biol.  Serv.  Program 

FWS/OBS-78/12. 
.    1978b.    Development  of  fishes  of  the  mid-Atlantic  Bight.  An 

atlas  of  egg,  larval  and  juvenile  stages.  Vol.  III.  Aphredoderidae 

through  Rachycentridae.  U.S.  Fish  Wildl.  Ser.  Biol.  Serv.  Program 

FWS/OBS-78/12. 
Harrington,  R.  W.,  Jr.    1958.    Morphometry  and  ecology  of  small 

tarpon  Megalops  allanlica  Valenciennes  from  transitional  stage 

through  onset  of  scale  formation.  Copeia  1958:1-10. 
.     1959.    Delayed  hatching  in  stranded  eggs  of  marsh  killifish, 

Fundutus  confluenlus.  Ecology  40:430-437. 
.     1961.    Oviparous  hermaphroditic  fish  with  internal  self-fer- 
tilization. Science  135(3492):  1749-1 750. 
,  AND  J.  S.  Haeger.     1958.    Prolonged  natural  deferment  of 

hatching  in  killifish.  Science  1  28(3337):151 1. 
Harrison,  C.  M.  H.,  and  G.  Palmer.    1968.    On  the  neotype  of  Ra- 

diicephahis  elongalus  Osorio  with  remarks  on  its  biology.  Bull.  Bnt. 

Mus.  (Nat.  Hist.)  Zool.  16:187-211. 
Harry,  R.  R.     1 95 1.    A  new  cusk-eel  of  the  genus  Ophidian  from 

California  with  notes  on  the  genus.  Stanford  Ichthyol.  Bull.  4:30- 

35. 
Hart.J.L.    1973.   Pacific  fishes  of  Canada.  Bull.  Fish.  Res.  Board  Can. 

180. 
Hart,  N.  H.,  R.  Pietri  AND  M.  Donovan.    1984.   The  structure  of  the 


chorion  and  associated  surface  filaments  in  On'r/ai— evidence  for 

the  presence  of  extracellular  tubules.  J.  Exp.  Zool.  230:273-296. 
Haryli,  T.,  AND  T.  Nishiyama.    1981.    Larval  form  of  Zaprorid  fish, 

Zaprora  silenus  from  the  Bering  Sea  and  the  Northern  Pacific.  Jpn. 

J.  Ichthyol.  28:313-318. 
Hastings,  R.  W.,  AND  R.  W.  Yerger.    1971.    Ecology  and  life  history 

of  the  diamond  killifish,  Adinia  xenica  (Jordan  and  Gilbert).  Am. 

Midi.  Nat.  86:276-291. 
H ATTORi,  S.    1 964.   Studies  on  fish  larvae  in  the  Kuroshio  and  adjacent 

waters.  Bull.  Tokai  Reg.  Fish.  Res.  Lab.  (40):  158  pp. 
.    1970.    Reproductive  aspects  of  fish  resources,  p.  209-222. /?!. 

Ocean  development.  4.  Exploitation  of  fisheries  resources.  T.  Sasaki 

(ed.).  Ocean  Dev.  Cent.  Press,  Tokyo.  (In  Japanese.) 
Hay,  D.  E.    1981.    Effects  of  capture  and  fixation  on  gut  contents  and 

body  size  of  Pacific  herring  larvae.  Rapp.  P.  V.  Reun.  Cons.  Int. 

Explor.  Mer  178:395-400. 
.    1984.    Reproductive  biology  of  Pacific  herring.  Can.  J.  Spec. 

Publ.  Fish  Aquat.  Sci.  In  Press. 
Ha,YDOCK,  I.    1971.   Gonad  maturation  and  hormone-induced  spawn- 
ing of  the  Gulf  croacker,  Bairdwila  icistia.  Fish.  Bull.  U.S.  69: 1  57- 

180. 
Hearne,  M.     1983.    Identification  of  larval  and  juvenile  smelts  (Os- 

meridae)  from  Oregon  and  California  using  selected  morphomctric 

characters.  Unpubl.  MS  thesis.  San  Francisco  State  University,  San 

Francisco,  CA. 
Hecht,  T.    1978.   A  descriptive  systematic  study  of  the  otoliths  of  the 

neopterygian  marine  fishes  of  South  Africa.  Part  I.  Introduction. 

Trans.  R.  Soc.  S.  Afr.  43(2):  191. 
Heck,  K.  L.,  and  M.  P.  Weinstein.     1978.     Mimetic  relationships 

between  tropical  burrfishes  and  opisthobranchs.  Biotropica  10:78- 

79. 
Heemstra,  P.  C.    1974.  On  the  identity  of  certain  Pacific  and  Caribbean 

post-larval  fishes  (Perciformes)  described  by  Henry  Fowler.  Proc. 

Acad.  Nat.  Sci.  Phila.  126:21-26. 
.    1980.   A  revision  of  the  zeid  fishes  (Zeiformes:  Zeidae)  of  South 

Africa.  Rhodes  Univ.  J.  L.  B.  Smith  Inst.  Ichthyol.  Bull.  41. 
.   In  press.   Families  Regalecidae,  Lopholidae,  Veliferidae.  Lam- 

pridae.  In:  Sea  Fishes  of  South  Africa,  Grahamstown.  S.A. 
,  andS.  X.  Kannemeyer.   In  press.  The  families  Trachipteridae 

and  Radiicephalidae  (Pisces:  Lampriformcs)  and  a  new  species  of 

Zu  from  South  Africa.  Ann.  S.  Afr.  Mus. 
,  and  J.  E.  Randall.    1977.   A  revision  of  the  Emmelichthvidae 


(Pisces:  Perciformes).  Aust.  J.  Mar.  Freshwater  Res.  1977(28):36l- 

396. 
Hefford,  a.  E.    1910.    Notes  on  teleostean  ova  and  larvae  observed 

at  Plymouth  in  spring  and  summer,  1 909.  J.  Mar.  Biol.  Assoc.  U.K. 

9:2-58. 
Heincke,  F.,  AND  E.  Ehrenbaum.    1900.   Eier  und  Larven  von  Fischen 

der  Deutschen  Bucht.   II.  Die  Bestimmung  der  schwimmenden 

Fischeier  und  die  Methodik  der  Eimessungen.  Wiss.  Meeresuntcrs. 

(Helgol.)  3:239-243. 
Helfrk  H,  P.  H.    1958.  Theearly  life  history  and  reproductive  behavior 

of  the  maomao,  Ahudefduf  ahdoniinalis  (Quoy  and  Gainard).  Lln- 

publ.  PhD  dissertation.  Univ.  Hawaii,  Honolulu. 
Hempel,  G.    1979.   Early  life  history  of  marine  fish:  the  egg  stage.  Univ. 

Washington  Press,  Seattle. 
,  and  J.  H.  S.  Blaxter.    1961.    The  experimental  modification 

of  meristic  characters  in  herring  (.Clupea  harcngits  L.).  J.  Cons.  Int. 

Explor.  Mer  26:336-346. 
,  AND  H.  Weikert.    1972.    The  neuston  of  the  subtropical  and 


boreal  Northeastern  Atlantic  Ocean.  A  Review.  Mar.  Biol.  (Berl.) 

13:70-88. 
Hennig,  W.     1950.    Grundziige  einer  Theorie  der  phylogenetischen 

Systematik.  Deutscher  Zentralverlag,  Berlin. 

.    1966.   Phylogenetic  Systematics.  Univ.  Illinois  Press,  Urbana. 

Hensley,  D.  A.    1976.    Collection  of  postlarval  juvenile //op//ai  ma/- 

aharicus  (Characoidei:  Erythrinidae)  in  Florida.  Fla.  Sci.  39:236- 

238. 


LITERATURE  CITED 


709 


.    1977,    Larval  development  of  Engyophn's  senta  (Bothidae), 

with  comments  on  intermuscular  bones  m  flatfishes.  Bull.  Mar.  Sci. 

27:681-703. 
Henzelmann,  E.    1930.   Exvjasuhex  Jordanctlatlondue Goods  el  Bean. 

(Floridakarpfling).  Wochenschr.  Aquar.  Terranien.  27(  19):307-309. 
Herald,  E.  S.    1939.   The  opah  (La/xpra  re^/«i)  and  its  occurrence  off 

the  California  coast.  Calif  Fish  Game  25:228-232. 
Herre,  A.  W.    1942.    New  and  little  known  phallostelhids,  with  keys 

to  the  genera  and  Phillipine  species.  Stanford  Ichthyol.  Bull.  2: 1 37- 

156. 
Herring,  P.  J.,  and  J.  G.  Morin.    1978.    Bioluminescence  in  fishes, 

p.  273-329.  In:  Bioluminescence  in  action.  P.  J.  Herring  (ed.). 

Academic  Press,  London. 
Herwig,  N.,  and  D.  Dewev.    1982.    Congrogadus  subduscens:  A  new 

challenge  for  the  manne  aquarist.  Freshwater  Mar.  Aquarium  5:6- 

91. 
Hettler,  W.    1981.    Spawning  and  rearing  Atlantic  menhaden.  Prog. 

Fish  Cult.  43:80-84. 
Heufelder,  G.  R.    1982.    Family  Cottidae,  sculpins,  pp.  656-676. /«.- 

Identification  of  larval  fishes  of  the  Great  Lakes  Basin  with  em- 
phasis on  the  Lake  Michigan  drainage.  N.  A.  Auer  (ed.).  Great 

Lakes  Fish.  Comm.  Spec.  Publ.  82-3. 
He>  wood,  V.  H.  (ED.).    1971.    Scanning  electron  microscopy;  system- 
atic and  evolutionary  applications.  Academic  Press.  New  York. 
Hkkei,  G.  M.    1979.   Survival  of  fish  larvae  after  injury.  J.  Exp.  Mar. 

Biol.  Ecol.  37:1-17. 
.    1982.    Wound  healing  in  fish  larvae.  J.  Exp.  Mar.  Biol.  Ecol. 

57:149-168. 
Hk  K.LING,  C.  F.    1931.    The  structure  of  the  otolith  of  the  hake.  Q.  J. 

Microsc.  Sci.  74:547-561. 
Hickman,  C.  P.,  Jr.    1959.    The  larval  development  of  the  sand  sole 

(Psettichthys  melanosliclus).  Wash.  Dep.  Fish.  Res.  Pap.  2(2):38- 

47. 
HiKiTA,  T.    1 952.  On  the  development  o( Limanda  schreuki  (Schmidt). 

Sci.  Rep.  Hokkaido  Fish  Hatchery  7(  1,2):133-144. 
.    1958.    On  the  development  of  long  finned  smelt.  ,S'p;/-/«c7n« 

lanceolalus.  Sci.  Rep.  Hokkaido  Fish  Hatch.  13:1-4. 
HiLDEBRAND,  S.  F.    1922.    Notcs  on  habits  and  development  of  eggs 

and  larvae  of  the  silversides,  Memdia  menidia  and  Menidia  ber- 

yllina.  Bull.  Bur.  Fish.  Wash.  38:1  13-120. 
.    1946.   A  descriptive  catalogue  of  the  shore  fishes  of  Peru.  U.S. 

Natl.  Mus.  Bull.  189. 
.    1963a.   Family  Elopidae,  p.  1 1  1-1  13. /«.■  Fishes  of  the  Western 

North  Atlantic.  H.  B.  Bigelow  (ed).  Mem.  Sears  Found.  Mar.  Res. 

No.  l(Pt.  3). 
.    1963b.   Family  Albuhdae.p.  132-147. /n.Fishesofthe  West- 
em  North  Atlantic.  H.  B.  Bigelow  (ed).  Mem.  Sears  Found.  Mar. 

Res.  No.  l(Pt.  3). 
.     1963c.    Family  Engraulidae,  p.   152-249.   //;.■  Fishes  of  the 

western  North  Atlantic.  H.  B.  Bigelow  (ed.).  Mem.  Sears  Found. 

Mar.  Res.  No.  l(Pt.  3). 
.    1963d.   Family  Clupeidae,  p.  257-454. /«.■  Fishes  of  the  west- 
em  North  Atlantic.  H.  B.  Bigelow  (ed).  Mem.  Sears  Found.  Mar. 

Res.  No.  KPt.  3). 
,  AND  L.  E.  Cable.    1930.    Development  and  life  history  of  four- 
teen teleostean  fishes  at  Beaufort.  N.C.  Bull.  Bur.  Fish.  Wash.  46 

(l093):383-488. 
,  AND  L.  E.  Cable.     1938.    Further  notes  on  the  development 

and  life  history  of  some  teleosts  at  Beaufort,  N.C.  Bull.  Bur.  Fish. 

Wash.  48(24):505-642. 
-,  AND  W.  C.  ScHROEDER.    1928.   Fishes  of  Chesapeake  Bay.  Bull. 


Bur.  Fish.  Wash.  43(Pt.  1):  1-388. 
Hipeau-Jacquotte.  R.    1967.   Notes  de  faunistique  et  biologic  marine 

de  Madagascar.  IV  — Observations  sur  le  comportement  du  poisson 

Carapidae:  Carapushomei  (Kxchardson  1 844)  de  Madagascar.  Reel. 

Trav.  Stn.  Mar.  Endoume  Mars.  Ease.  Ser.  Suppl.  6:141-151. 
HoDA,  S.  M.  S.,  AND  H.  Tsukahara.    1971.    Studies  on  the  develop- 


ment and  relative  growth  in  the  carp,  Cypnnus  carpio  (Linne).  J. 

Fac.  Agric.  Kyushu  Univ.  16:387-509. 
Hoedeman,  J.  J.    1960a.    Studies  on  callichthyid  fishes.  3.  Notes  on 

the  development  oi  Callichthys  ( 1 )  ( Pisces-Siluriformes).  Bull.  Aquat. 

Biol.  1:53-72. 
.    1960b.    Studies  on  callichthyid  fishes.  4.  Development  of  the 

skull  of  Callichthys  and  Hoplostenmm  (1)  (Pisces-Siluriformes). 

Bull.  Aquat.  Biol.  1:73-84. 
.    1960c.    Studies  on  callichthyid  fishes.  5.  Development  of  the 

skull  of  Callichthys  and  Hoplosternum  (2)  (Pisces-Siluriformes). 

Bull.  Aquat.  Biol.  2:21-36. 
.    1960d.   Studies  on  callichthyid  fishes.  6.  The  axial  skeleton  of 

Callichthys  and  Hoplosternum  (Pisces-Siluriformes).  Bull.  Aquat. 

Biol.  2:37-44. 
HoESE,  D.  F.     1971.    A  revision  of  the  eastem  Pacific  species  of  the 

gobiid  fish  genus  Gohiosoma,  with  a  discussion  of  relationships  of 

the  genus.  Unpubl.  PhD  dissertation.  Univ.  Calif,  San  Diego. 
.     1976.    A  redescnption  of  HeterocUnus  adelaide  Costelnau 

(Pisces:  Clinidae)  with  description  of  related  new  species.  Aust. 

Zool.  19:51-67. 
Hc>ESE,  H.  D.,  and  R.  H.  Moore.    1977.    Fishes  of  the  Gulf  of  Mexico, 

Texas,  Louisiana,  and  adjacent  waters.  Texas  A&M  Univ.  Press, 

College  Station  and  London. 
HoFF,  F.  H.,Jr.   1970.  Artificial  spawning  ofblack  sea  bass,  Centropris- 

tes  slnatus  melanus  Ginsburg,  aided  by  chorionic  gonadotrophic 

hormones.  Fla.  Dep.  Nat.  Resour.  Mar.  Res.  Lab.  Spec.  Sci.  Rep. 

25. 
HoLCiK,  J.    1982.   Towards  the  characteristics  of  the  genera  Hiicho  and 

Brachymystax  {Pisces.  Salmonidae).  Folia  Zool.  31:369-380. 
HoLL,  E.  A.    1968.   Notes  of  spawning  behavior  of  barbel  Clarias  gar- 

iepinus  Burchell  in  Rhodesia.  Zool.  Afr.  3:185-188. 
Hollister.G.    1936.  Caudal  skeleton  of  Bermuda  shallow  water  fishes. 

I.  Order  Isospondyli:  Elopidae,  Megalopidae,  Albulidae,  Clupeidae, 

Dussumieriidae,  Engraulidae.  Zoologica  (N.Y.)  21:257-290. 
.    1937.    Caudal  skeleton  of  Bermuda  shallow  water  fishes.  II. 

Order  Percomorphi,  suborder  Percesoces:  Atherinidae,  Mugilidae. 

Sphyraenidae.  Zoologica  (N.Y.)  22(3):265-279. 
Holt,  E.  W.  L.    1891.   Survey  of  fishing  grounds,  west  coast  of  Ireland, 

1890.  On  the  eggs  and  larvae  of  teleosteans.  Sci.  Trans.  R.  Dubl. 

Soc,  Ser.  2,  4:435-474. 
,    1892.   Rhombus  maximus  Linn.  (The  turbot).  In:  "Notes  and 

Memoranda."  J.  Mar.  Biol.  Assoc.  U.K.  2:399-404. 
.    1893.   Survey  of  fishing  grounds,  west  coast  of  Ireland,  1 890- 

9 1 :  on  the  eggs  and  post-larval  stages  of  teleosteans.  Sci.  Trans.  R. 

Dubl.  Soc,  Ser.  2,  5. 
.     1897.    Preliminary  notes  on  the  reproduction  of  teleostean 

fishes  in  the  south-westem  district.  J.  Mar.  Biol.  Assoc.  U.K.  5:41- 

50. 
.    1898.    Contributions  to  our  knowledge  of  the  plankton  of  the 


Faeroe  channel.  No.  V.  Report  on  a  collection  of  very  young  fishes 

obtained  by  Dr.  G.  H.  Fowler  in  the  Faeroe  Channel.  Proc.  Zool. 

Soc.  Lond.  1898:550-557. 
HoNMA,  Y.    1956.   Short  note  on  the  breeding  habit  of  a  spiny  lumpftsh. 

Cyclolumpus  aspemmus.  Collect.  Breed.  18:342-343. 
.    1957.   Notes  on  a  smooth  lumpsucker,  Aplocyclus  ventricosiis 

(Pallas),  during  its  breeding  season.  Collect.  Breed.  19:235-236. 
Horn,  M.  H.     1970a.    The  swimbladder  as  a  juvenile  organ  in  stro- 

mateoid  fishes.  Breviora  359. 
.    1970b.    Systematics  and  biology  of  the  stromateoid  fishes  of 

the  genus  Pepnlus.  Bull.  Mus.  Comp.  Zool.  Harv.  Univ.  140:165- 

262. 
.    1973.    Systematic  companson  of  the  stromateoid  fishes  Stro- 

mateus  brasiliensis  Fowler  and  Slromateus  sicllatus  Cuvier  from 

coastal  South  America  with  a  review  of  the  genus.  Bull.  Br.  Mus. 

(Nat.  Hist.)  Zool.  24:317-339. 
.    1975.   Swim-bladder  state  and  structure  in  relation  to  behavior 

and  mode  of  life  in  stromateoid  fishes.  Fish.  Bull.  U.S.  73:95-109. 
,  AND  R.  L.  Haedrkh.     1973.    Systematic  and  distributional 


710 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


status  of  Psenes  sio  and  Psenes  pellucidus  (Pisces,  Stromateoidei) 
in  the  eastern  Pacific.  Copeia  1973:167-169. 

Hoss,  D.  E.,  AND  J.  H.  S.  Blaxter.  1982.  Development  and  function 
of  the  swimbladder-inner  ear-lateral  line  system  in  the  Atlantic 
menhaden  Brevoortia  lyrannus  {l^l^ohe).  J.  Fish  Biol.  20: 1 3 1-142. 

HoTTA,  H.  1955.  On  the  mature  mugilid  fish  from  Kabashima,  Na- 
gasaki Pref,  Japan,  with  additional  notes  on  the  intestmal  con- 
volution of  Mugilidae.  Jpn.  J.  Ichthyol.  4:162-169. 

,  AND  I.-S.  Tung.    1972.    Identification  of  fishes  of  the  family 

Mugilidae  based  on  the  pyloric  caeca  and  the  position  on  inserted 
first  intemeural  spine.  Jpn.  J.  Ichthyol.,  14:62-66.  (In  Japanese. 
English  summary.) 

HouDE.  E.  D.  1972a.  Some  recent  advances  and  unsolved  problems 
in  the  culture  of  marine  fish  larvae.  Proc.  World  Maricull.  Soc.  3: 
83-112. 

.     1972b.    Development  and  early  life  history  of  the  northern 

sennet,  Sphyraena  borealis  DeKay  (Pisces:  Sphyraenidae)  reared  in 
the  laboratory.  Fish.  Bull.  U.S.  70:185-195. 

.    1974.    Effects  of  temperature  and  delayed  feeding  on  growth 

and  survival  of  larvae  of  three  species  of  subtropical  marine  fishes. 
Mar.  Biol.  (Bed.)  26:271-285. 

.    1975.   Effects  of  stocking  density  and  food  density  on  survival, 

growth  and  yield  of  laboratory-reared  larvae  of  sea  bream  Archo- 
sargus  rhomboidalts  (L.)  (Sparidae).  J.  Fish  Biol.  7:1 15-127. 

.     1977.    Food  concentrations  and  stocking  density  effects  on 

survival  and  growth  of  laboratory-reared  larvae  of  bay  anchovy 
Anchoa  mitchilti  and  lined  sole  Achirus  lineatus.  Mar.  Biol.  43: 
333-341. 

.    1978.   Critical  food  concentrations  for  larvae  of  three  species 

of  subtropical  marine  fishes.  Bull.  Mar.  Sci.  28:395-41 1. 

.     1981.    Distribution  and  abundance  of  four  types  of  codlet 

(Pisces:  Bregmacerotidae)  larvae  from  the  eastern  Gulf  of  Mexico. 
Biol.  Oceanogr.  1:81-104. 

,  AND  J.  D.  Alpern-Lovdal.    MS.    Seasonality  of  occurrence, 

foods  and  food  preferences  of  ichthyoplankton  in  Biscayne  Bay, 
Florida. 

,  and  p.  L.  Fore.    1973.    Guide  to  identity  of  eggs  and  larvae 


of  some  Gulf  of  Mexico  clupeid  fishes.  Fla.  Dep.  Nat.  Resour.  Mar. 
Res.  Lab.  Leafl.  Ser.  4  (Pt.  1,  No.  23). 

— ,  C.  R.  FuTCH  and  R.  Detwyler.  1970.  Development  of  the 
lined  sole,  Achirus  lineatus.  described  from  laboratory-reared  and 
Tampa  Bay  specimens.  Fla.  Dep.  Nat.  Resour.  Mar.  Res.  Lab.  Tech. 
Ser.  62. 

— ,  J.  C.  Leak,  C.  E.  Dowd,  S.  A.  Berkeley  and  W.  J.  Richards. 
1979.  Ichthyoplankton  abundance  and  diversity  in  the  eastern  Gulf 
of  Mexico.  Final  Rep.  to  U.S.  Bur.  Land  Mgt.,  Contract  No.  AA550- 
CT7-28.  Rosensteil  School  of  Marine  and  Almosphenc  Science. 
Univ.  Miami,  FL. 

— ,  and  T.  PoTTHOFF.  1976.  Eggs  and  larval  development  of  the 
sea  bream  Archosargus  rhomboidalts  (Linnaeus):  Pisces,  Sparidae. 
Bull.  Mar.  Sci.  26:506-529. 

— ,  W.  J.  Richards  AND  V.  P.  Saksena.  1974.  Description  of  eggs 
and  larvae  of  scaled  sardine,  Harengula  jaguana.  Fish.  Bull.  U.S. 
72:1106-1122. 

,  AND  R.  C.  ScHEKTER.  1978.  Simulated  food  patches  and  Sur- 
vival of  larval  bay  anchovy  Anchoa  mitchilli  and  sea  bream  Ar- 
chosargus rhomboidalis.  Fish.  Bull.  U.S.  76:483-487. 

— ,  AND  L.  J.  SwANSON,  Jr.  1975.  Description  of  eggs  and  larvae 
of  yellowfin  menhaden,  Brevoortia  smithi.  Fish.  Bull.  U.S.  73:660- 
674. 

— ,  AND  A.  K.  Taniguchi.    1979.    Laboratory  culture  of  marine 


fish  larvae  and  their  role  in  marine  environmental  research,  p.  176- 
205.  In:  Advances  in  marine  environmental  research.  Proceedings 
of  a  Symposium.  F.  S.  Jacoff(ed.).  Environ.  Res.  Lab.,  U.S.  Environ. 
Prot.  Agency,  Narragansett,  RI.  Rep.  No.  EPA-600/9-79-035. 
Howe,  K.  M.,  AND  S.  L.  RICHARI5SON.  1978.  Taxonomic  review  and 
meristic  variation  in  marine  sculpins  (Osteichthyes:  Cottidae)  of  the 
northeast  Pacific  Ocean.  Fin.  Rep..  NOAA-NMFS  Contrib.  No. 
03-78-M02-120. 


Howell,  B.  R.     1973.    Marine  fish  culture  in  Britain  VIII.  A  marine 

rotifer  Brachionus  plicalilis  Muller  and  the  larvae  of  the  mussel 

Mylilus  edults  L.  as  food  for  larval  flatfish.  J.  Cons.  Int.  Explor. 

Mer  35:1-6. 
HuBBS,  C.    1952.    A  contribution  to  the  classification  of  the  blennioid 

fishes  of  the  family  Clinidae.  with  a  partial  revision  of  the  eastern 

Pacific  forms.  Stanford  Ichthyol.  Bull.  4(2):41-I65. 
.     1953a.    Revision  and  systematic  position  of  the  blennioid 

fishes  of  the  genus  NeocUnus.  Copeia  1953:1  1-23. 
.    1953b.   Revision  of  the  eastern  Pacific  fishes  of  the  clinid  genus 

Labrisomus.  Zoologica  (N.Y.)  38:1 13-136. 
.    1965.    Developmental  temperature  tolerance  and  rates  of  four 

southern  California  fishes,  Fundulus parvipmnis.  Athennops  affinis. 

Leureslhes  tenuis,  and  Hvpsoblennius  sp.  Calif  Fish  Game  51:113- 

122. 
,  andD.  F.  BuRNSiDE.    1972.   Developmental  sequences  of  Zrgo- 

nectes  notalus  at  several  temperatures.  Copeia  1972:862-865. 
HuBBS,  C.  L.    1918.  The  fishes  of  the  genus  Atherinops.  their  variation, 

distribution,  relationships,  and  history.  Bull.  Amer.  Mus.  Nat.  Hist. 

38:409^40. 
.    1925.   The  metamorphosis  of  the  California  ribbonfish  Tra- 

chipterus  rex-salmonorum.  Pap.  Mich.  Acad.  Sci.  Arts  Lett.  5:469- 

476. 
.    1934.    Coelorhynchus  marinii,  a  new  macrourid  fish  from  Ar- 
gentina and  South  Georgia.  Occas.  Pap.  Mus.  Zool.  LIniv.  Mich. 

298:1-9. 

.    1943.   Terminology  of  early  stages  of  fishes.  Copeia  1943:260. 

.     1945.    Phylogenetic  position  of  the  Cithandae.  a  family  of 

flatfishes.  Univ.  Mich.  Mus.  Zool.  Misc.  Publ.  63. 
.    1946.    Pertinence  of  the  East  Indian  heterosomate  fish  genus 


Lepidoblepharon  to  the  Citharidae.  Copeia  1946:97. 
.    1951.    Identification  as  osmerids  of  two  fishes  described  from 

the  North  Pacific  as  paralepidids.  Copeia  1951:296. 
.    1953.    Synonymy  of  the  bathypelagic  fish  genus  Rhynchohy- 

alus.  referred  to  the  expanded  family  Argentinidae.  Copeia  1953: 

96-97, 
.    1958.    Dikellorhynchus  and  Kanazawaichlhys:  nominal  fish 

genera  interpreted  as  based  on  prejuveniles  of  Malacanthus  and 

Antennarius,  respectively.  Copeia  1958:282-285. 
,  AND  Y.  T.  Chc.    1934.    Asiatic  fishes  (Diplopnon  and  Laeops) 

having  a  greatly  elongated  dorsal  ray  in  very  large  post-larvae. 

Occas.  Pap.  Mus.  Zool.  Univ.  Mich.  299:1-7. 
,  AND  L.  C.  HuBBS.     1945.    Bilateral  asymmetry  and  bilateral 

variation  in  fishes.  Pap.  Mich.  Acad.  Sci.  Arts  Lett.  30:229-310. 
,  AND  T.  IwAMOTO.    1977.    A  new  genus  {Mesobnis).  and  three 

new  bathypelagic  species  of  Macrouridae  (Pisces,  Gadiformes)  from 

the  Pacific  Ocean.  Proc.  Calif  Acad.  Sci.  Fourth  Ser.  41(7):233- 

251. 
,  AND  E.  M.  Kampa.    1946.   The  early  stages  (egg,  prolarva  and 

juvenile)  and  the  classification  of  the  California  flying  fish.  Copeia 

1946:188-218. 
,  AND  R.  L.  WiSNER.    1964.   fam/w.v.  a  new  genus  of  myctophid 


fishes  from  the  northeastern  Pacific,  with  two  new  species.  Zool. 
Meded.  (Leiden)  39:445-463. 
— ,  AND .    1980.    Revision  of  the  sauries  (Pisces,  Scomber- 


esocidae)  with  descriptions  of  two  new  genera  and  one  new  species. 

Fish  Bull.  U.S.  77:521-566. 
Hudson,  L.  L.,  and  J.  D.  Hardy,  Jr.    1975.    Eggs  and  larvae  of  the 

Atlantic  seahorse.  Hippocampus  hudsonnis.  Univ.  Md.,  CEES,  Ref 

175-I2CBL. 
Hlighes,  D.  R.    1981.    Developmeni  and  organization  of  the  posterior 

field  of  ctenoid  scales  in  the  Platycephalidae.  Copeia  1981:596- 

606. 
HuLET,  W.  H.    1978.   Structure  and  functional  development  of  the  eel 

leptocephalus  Ariosoma  balearicum  (De  La  Roche,  1809).  Philos. 

Trans.  R.  Soc.  Lond.  282(987):  107-1 38. 
Hulley,  p.  a.    1972.    A  report  on  the  mesopelagic  fishes  collected 


LITERATURE  CITED 


711 


during  the  deep-sea  cruises  on  R.S.  "Africana  II,"  1961-1966.  Ann. 

S.  Afr.  Mus.  60:197-236. 
.    1981.    Results  of  the  research  cruise  of  FRV  "Walter  Herwig" 

to  South  America.  LVIII.  Family  Myctophidae(Osteichthyes,  Myc- 

tophiformes).  Arch.  Fischereiwiss.  3 1(1):  1-300. 
,  AND  R.  E.  Rau.    1969.   A  female  Regalecus  glesne  from  Cape 

Province,  South  Africa.  Copeia  1969:835-839. 
HuMASON,  G.  L.    1979.  Animal  tissue  techniques.  W.  H.  Freeman  and 

Co.,  San  Francisco. 
Humphries,  J.  M.,  F.  L.  Bookstein,  B.  Chernoff,  C.  R.  Smith,  R.  L. 

Elder  AND  S.  G.  Poss.    1981.   Multivariate  discrimination  by  shape 

in  relation  to  size.  Syst.  Zool.  30:291-308. 
Hunter,  C.  J.    1969.   Confirmation  of  symbiotic  relationship  between 

liparid  fishes  and  male  king  crab.  Pac.  Sci.  23:546-547. 
Hunter,  J.  R.    1972.    Swimming  and  feeding  behavior  of  larval  an- 
chovy Engraulis  mordax.  Fish.  Bull.  U.S.  70:821-838. 
.     1976.    Culture  and  growth  of  northern  anchovy  Engraulis 

mordax  \aT\?ie.  Fish.  Bull.  U.S.  78:81-88. 
.    1981.   Feeding  ecology  and  predation  of  marine  fish  larvae,  p. 

a-ll.  In:  Marine  fish  larvae.  Morphology,  ecology  and  relation 

to  fisheries.  R.  Lasker  (ed.).  Univ.  Washington  Press,  Seattle.  WA. 
.    1984.    Inferences  regarding  predation  on  the  early  life  stages 

of  cod  and  other  fishes,  p.  533-562.  In:  The  propagation  of  cod. 

E.  Dahl,  D.  S.  Danielssen,  E.  Moksness  and  P.  Solemdal  (eds.). 

Flodevigen  Rapp.  1. 
,  AND  K.  M.  Coyne,    1982.   The  onset  of  schooling  in  northern 

anchovy  larvae  Engraulis  mordax.  Calif.  Coop.  Oceanic  Fish.  In- 
vest. Rep.  23:246-251. 
,  S.  E.  Kaupp  AND  J.  T.  Taylor.    1982.    Assessment  of  effects 

of  UV  radiation  on  marine  fish  larvae,  p,  459-497,  In:  The  role  of 

solar  ultraviolet  radiation  in  marine  ecosystems.  NATO  Conference 

Series  IV:  Marine  Sciences.  J.  Calkins  (ed.).  Plenum  Press,  New 

York. 

AND  C.  A.  Kimbrell.    1980.    Egg  cannibalism  in  the  northern 


anchovy  Engraulis  mordax.  Fish.  Bull.  U.S.  78:81  1-816. 

HuREAu,  J.  C.  1971.  Notes  sur  la  famille  des  Congiopodidae  (Te- 
leosteens,  Perciformes):  redecouverte  de  Zanctorhynchus  spinifer 
(Gunther,  1880)  aux  iles  Kerguelen  et  rehabilitation  de  Congio- 
podus  kienernSauvage,  1878).  Bull.  Mus.  Natl.  Hist.  Nat.  42: 1019- 
1026. 

,  AND  A.  ToMO.     1977.    Bovichthys  elongalus  n.  sp.,  Poisson 

Bovichthyidae,  famille  nouveau  pour  I'Antarctique,  Cybium  3e  Ser. 
1977(l):67-74. 

HussAiN,  N,,  S,  AiovTSU  AND  C.  El-Zahr.  1981.  Spawning,  egg  and 
early  development,  and  growth  oi Acanlhopagrus  cmieri  (Sparidae). 
Aquaculture  22:125-136. 

HussAiN,  N.  A.,  AND  M.  HioucHi,  1980.  Larval  rearing  and  devel- 
opment of  the  brown  spotted  grouper,  E^inephelus  tauvina  (For- 
skal).  Aquaculture  19:339-350. 

Huxley,  J.  S.  1932.  Problems  of  relative  growth.  Methucn  and  Co., 
London. 

HuYOEBAERT,  B.,  AND  D.  Nolf.  1979.  An  annotated  bibliography  of 
paleontological  and  systematic  papers  on  fish-otoliths,  published 
since  1968.  Meded.  Werkgr.  Tert.  Kwart.  Geol.  16(4):  139-170. 

Ida,  H.  1976.  Removal  of  the  family  Hypoptychidae  from  the  sub- 
order Ammodytoidei,  order  Perciformes,  to  the  suborder  Gaster- 
osteoidei,  order  Syngnathiformes.  Jpn.  J.  Ichthyol.  23:33-42. 

Iljin,  B.  S.  1930.  Le  systeme  des  Gobiides.  Trab.  Inst.  Esp.  Oceanogr. 
(2):  1-63. 

Imai,  S.  1 94 1 .  Seven  new  deep-sea  fishes  obtained  in  Sagami  Bay.  Jpn. 
J.  Zool.  9(2):233-250. 

.     1958.    Young  of  Myctophum  orienlale  (Gilbert),  p.  21.  In: 

Studies  on  the  eggs,  larvae  and  juveniles  of  Japanese  fishes.  Ser.  I. 
Uchida  el  al.  (eds.).  Fac.  .Agric.  Kyushu  Univ.,  Fukuoka,  Japan. 
(In  Japanese.) 

.    1959.   Studies  on  the  life  histories  of  the  flying-fishes  found  in 

theadjacent  waters  of  Japan.  — I.  Mem.  Fac.  Fish.  KagoshimaUniv. 
7:1-85.  (In  Japanese.) 

.    1960.   Studies  on  the  life  histories  of  flying-fishes  found  in  the 


adjacent  waters  of  Japan  — II.  Mem.  Fac.  Fish.  ICagoshima  Univ. 
8:8-45.  (In  Japanese.) 

,  AND  K.  Nakahar.a.    1957.    Coreoperca  kawamurainnd'WsWk 

history,  p.  591-601.  In:  Syehiro  et  al.  (eds.).  Suisangaku  Shusei. 
(In  Japanese.) 

Inaba,  D.,M.  Nomura,  AND  M.  Nakamura.  1957.  Preliminary  report 
on  the  spawning  of  grass-carp  and  silver-carp  in  the  Tone  River, 
Japan  and  the  development  of  their  eggs.  J.  Tokyo  Univ.  Fish. 
43(l):81-96. 

Inada,  T.  1981a.  Two  nominal  species  of  A/er/«cc/!tj  from  New  Zea- 
land and  southern  South  America.  Jpn.  J.  Ichthyol.  28:31-36. 

.    1981b.    Studies  on  the  merlucciid  fishes.  Bull.  Far  Seas  Fish. 

Res.  Ub.  (Shimizu)  18:1-172. 

,  and  I.  Nakamura.  1975.  A  comparative  study  of  two  pop- 
ulations of  the  gadoid  fish  Micromesislius  auslralis  from  the  New 
Zealand  and  Patagonian-Falkland  regions.  Bull.  Far.  Seas  Fish.  Res. 
Lab.  (Shimizu)  13:1-26. 

Inglis,  W.  G.  1966.  The  observational  basis  of  homology.  Syst.  Zool. 
15:219-228. 

Innes,  W.  T.  1956.  Exotic  aquarium  fishes.  19th  Ed.  G.  S.  Myers  (ed.). 
Innes  Publ.  Co.,  Phila. 

Irie,  T.  1960.  The  growth  of  the  fish  otolith.  J.  Fac.  Fish.  .'\nim.  Husb. 
Hiroshima  Univ.  3:203-221. 

Ishigaki,  T.  1957.  A  few  informations  concerning  "Shiwa-ikanago, 
Hypoptychus  dybowskii  Steindachner",  in  the  waters  around  Hok- 
kaido. Month.  J.  Hokkaido  Fish  Res.  Stn.  14(8):  14-24.  (In  Jap- 
anese.) 

Ishiyama,  R,  1951.  Revision  of  the  Japanese  mugilid  fishes  especially 
based  upon  the  osteological  characters  of  the  cranium.  Jpn.  J.  Ich- 
thyol. 1:230-250.  (In  Japanese.  English  summary.) 

Ivankov,  V.  N.,  AND  V.  P.  Kurdiayeva.  1973.  Systematic  differences 
and  the  ecological  importance  of  the  membranes  in  fish  eggs.  J. 
Ichthyol.  (Engl.  Transl.  Vopr.  Ikhtiol.)  13:864-873. 

IvANTSOFF,  W.  1978.  Taxonomic  and  systematic  review  of  the  Aus- 
tralian fish  species  of  the  family  Atherinidae  with  references  to 
related  species  of  the  Old  World.  Unpubl.  PhD  dissertation.  Mac- 
quarie  University,  Sydney. 

Iwami,  T,,  andT.  Abe.  1980.  Records  of  adults  of  some  scopelarchid 
fishes  from  the  western  North  Pacific  and  the  Southern  Ocean,  with 
osteological  notes  on  five  species  of  the  genus  Benthalbella.  U.  O. 
No.  31:1-20. 

IwAMOTO,  T.  1966.  The  macrourid  fish  Cetonurus  glohiceps  in  the 
Gulf  of  Mexico.  Copeia  1966:439-442. 

.    1974,    Nezumia  {Kuronezuniia)  bubonis.  a  new  subgenus  and 

species  of  grenadier  (Macrouridae:  Pisces)  from  Hawaii  and  the 
western  North  Atlantic.  Proc.  Calif  Acad.  Sci.,  Fourth  Ser..  39(22): 
507-516. 

.    1978.    Eastern  Pacific  macrourids  of  the  genus  Coelorhynchus 

Gioma  (Pisces:  Gadiformes),  with  description  of  a  new  species  from 
Chile.  Proc,  Calif  Acad.  Sci.,  Fourth  Ser.,  41(12):307-337. 

.    1979.   Eastern  Pacific  macrourine  grenadiers  with  seven  bran- 

chiostegal  rays  (Pisces:  Macrouridae).  Proc.  Calif  Acad.  Sci.  42(5): 
135-179. 

,  AND  J.  C.  Staiger.    1976.    Percophidid  fishes  of  the  genus 


Chrionema  Gilbert.  Bull.  Mar.  Sci.  26:488-498. 

,  AND  D.  L.  Stein.    1974.  A  systematic  review  of  the  rattail  fishes 

(Macrouridae:  Gadiformes)  from  Oregon  and  adjacent  waters.  Oc- 
cas.  Pap.  Calif  Acad.  Sci.,  No,  1 1 1:1-79. 

Jacino,  A.  1909.  Uovo  e  larve  di  rraf/jvp^erwisp.  Arch.  Zool.  Napoli 
3(4):479-484. 

Jackson,  P.  D.  1978.  Spawning  and  early  development  of  the  river 
blackfish,  Gadopsis  marmoralus  Richardson  (Gadopsiformes:  Ga- 
dopsidae)  in  the  McKenzie  River,  Victoria.  Aust.  J.  Mar.  Fresh- 
water Res.  29:293-298. 

James,  G.D.  1976a,  Eggs  and  larvae  o( Ihe  IrevaWy  Caran.xgeorgianus 
(Teleostei:  Carangidae).  N.Z.  Mar.  Freshwater  Res.  10:301-310. 

.    1976b.    Cyllus  iraversi  Hullon:  Suveni\e  (onn  of  C.  venlralis 

Barnard  and  Davies  (Pisces,  Zeidae).  J.  R.  See.  N.Z.  6(4):493-498. 


712 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


James,  P.  S.  B.  R.  1970.  Micrognathus  brevirosrns  (Ruppell)  (Family 
Syngnathidae:  Pisces)— a  new  record  from  the  Indian  Seas  with 
observations  on  its  early  development.  J.  Mar.  Biol.  Assoc.  India 
12:158-162. 

Janssen,  J.,  AND  G.  R.  Harbison.  1981.  Fish  in  salps:  the  association 
of  squaretails  (Tetragonurus  spp.)  with  pelagic  tunicates.  J.  Mar. 
Biol.  Ass.  U.  K.  61:917-927. 

Jenkins,  G.  P.,  N.  E.  Milward  and  R.  F.  Hartwick.  1984.  Identi- 
fication and  description  of  larvae  of  Spanish  mackerels,  genus 
Scomberomorus  (Teleostei:  Scombridae),  in  shelf  waters  of  the  Great 
Bamer  Reef.  Aust.  J.  Mar.  Freshwater  Res.  35:341-353. 

Jensen,  A.  S.  1935.  The  Greenland  halibut  (Reinhardtius  hippoglos- 
soides).  its  development  and  migrations.  K.  Dan.  Vidensk.  Selsk. 
Skr.  9(4):  1-32. 

.    1 948.   Contributions  to  the  ichthyofauna  of  Greenland.  Spolia 

Zool.  Mus.  Haun.  9:1-176. 

Jensen,  R.J.  1981.  Wagner  networks  and  Wagner  trees:  a  presentation 
of  methods  for  estimating  parsimonious  solutions.  Taxon  30:576- 
590. 

Jerzmansica,  A.  1968.  Ichtyofaune  des  couches  a  menilite  (flysch  des 
Karpathes).  Acta  Palaeontol.  Pol.  13(3):379-488. 

Jespersen,  p.  1942.  Indo-Pacific  leplocephalids  of  the  genus  Angiulta. 
Systematic  and  biological  studies.  Dana-Rep.  Carlsberg  Found.  22. 

,  AND  A.  V.  TAning.    1919.   Some  Mediterranean  and  Atlantic 

Stemoptychidae.  Preliminary  Note.  Saertryk  af  Vidensk.  Medd. 
Dansk  Naturh.  Foren.  70:215-226. 

,  AND .    1926.  Mediterranean  Stemoptychidae.  Rep.  Dan. 

Oceanogr.  Exped.  Mediterr.  2  (A.  12):  1-59. 

Jhingran,  V.  G.  1975.  Fish  and  fisheries  of  India.  Hindustan  Pub- 
lishing Corporation.  Delhi,  India. 

Jillett,  J.  B.  1 968.  The  biology  of  Acanlhoclinus  quadndactylus  (Bloch 
and  Schneider)  (Teleostei  — Blennioidea).  II.  Breeding  and  devel- 
opment. Aust.  J.  Mar.  Freshwater  Res.  19:9-18. 

Job,  T.  J.  1940.  On  the  breeding  and  development  of  Indian  "mos- 
quito-fish" of  the  genera  Aplocheilus  McCelland  and  Oryzias  Jordan 
and  Snyder.  Rec.  Indian  Mus.  (Calcutta)  42:51-79. 

,  AND  S.  Jones.    1938.   Studies  on  the  development  of  the  Indian 

garfish,  Tylosurus  sirongylunis  (van  Hass.)  with  notes  on  the  young 
stages  of  Hemirhamphus  gaimardi  Cuv.  &  Val.  Rec.  Indian  Mus. 
(Calcutta)  40(3):245-253. 

Johansen,  F.  1912.  The  fishes  of  the  Danmark  Expedition.  Medd. 
Gronl.  45(12):633-675, 

John,  C.  M.  1950.  Early  stages  in  the  development  of  the  sucker  fish 
Echeneis  naucrales  Linn.  Bull.  Cent.  Res.  Inst.  Univ.  Travancore 
India  lC(l):47-55. 

John,  M.  A.  1951a.  Pelagic  fish  eggs  and  larvae  of  the  Madras  coast. 
J.  Zool.  Soc.  India  2:267-268. 

.    1951b.    Pelagic  fish  eggs  and  larvae  of  the  Madras  coast.  J. 

Zool.  Soc.  India  3:41-69. 

Johnsen,  S.  1921.  Ichthyologiske  notiser  I.  Bergens  Mus.  Aarbok  1918- 
1919.  Naturvid.  Raekke  No.  6.  (English  summary.) 

.    1927.    On  some  balhypelagic  stages  of  macrourid  fishes.  Nytt 

Mag.  Natur\'id.  65:221-241. 

Johnson,  C.  R.  1 969.  Contnbutions  to  the  biology  of  the  showy  snail- 
fish,  Liparis  pulchellus  (Lipahdae).  Copeia  1969:830-835. 

,  D.  A.  Ratkowskv  AND  R.  W.  G.  White.    1983.    Multivariate 

analysis  of  the  phenolypic  relationships  of  the  species  of  Paraga- 
laxias  and  Galaxias  (Pisces:  Galaxiidae)  in  Tasmania.  J.  Fish  Biol. 
23:49-63. 

,  R.  W.  G.  White  AND  Y.  A.  E.  Beck.    1981.   Cytotaxonomy  of 

seven  species  of  Galaxias  (Pisces:  Galaxiidae)  in  Tasmania.  Ge- 
netica  56:17-21. 

Johnson,  G.  D.  1975.  The  procurrent  spur,  an  undescribed  perciform 
character  and  its  phylogenetic  implications.  Occas.  Pap.  Calif.  Acad. 
Sci.  121:1-23. 

.    1978.    Development  of  fishes  of  the  Mid-Atlantic  Bight.  An 

Atlas  of  egg.  larval  and  juvenile  stages.  Vol.  IV.  Carangidae  through 


Ephippidae.  U.S.  Fish  Wildl.  Serv.  Biol.  Serv.  Program  FWS/OBS- 
78/12. 

.  1980.  The  limits  and  relationships  of  the  Lutjanidae  and  as- 
sociated families.  Bull.  Scripps  Inst.  Oceanogr.  Univ.  Calif  24. 

.    1983.  Niphon  spinosus: &  primitive  epinephelineserranid,  with 

comments  on  the  monophyly  and  intrarelationships  of  the  Serran- 
idae.  Copeia  1983:777-787. 

,  AND  P.  Keener.  1984.  Aid  to  identification  of  Amencan  grou- 
per larvae.  Bull.  Mar.  Sci.  34:106-134. 

Johnson,  R.  K..  1970.  A  second  record  of  Korsogasler  nanus  Parr 
(Beryciformes:  Korsogasteridae).  Copeia  1970:758-760. 

.     1974a.    A  Macristium  larva  from  the  Gulf  of  Mexico  with 

additional  evidence  for  the  synonymy  of  Macnslntm  with  Balhy- 
iflMrwi  (Myctophiformes:  Bathysauridae).  Copeia  1974:973-977. 

.    1974b.   A  revision  of  the  alepisauroid  family  Scopelarchidae 

(Pisces:  Myctophiformes).  Fieldiana  Zool.  66. 

.    1975.    A  new  myctophid  fish.  Bolinichthys  distofax.  from  the 

western  and  central  north  Pacific  Ocean,  with  notes  on  other  species 
o(  Bolinichthys.  Copeia  1975:53-60. 

.    1982.    Fishes  of  the  families  Evermannellidae  and  Scopelar- 


chidae: systematics,  morphology,  interrelationships  and  zoogeog- 
raphy. Fieldiana  Zool.  N.S.  12. 

— ,  and  D.  M.  Cohen.  1974.  Results  of  the  research  cruises  of 
FRV  "Walther  Herwig"  to  South  America.  30.  Revision  of  the 
chiasmodontid  fish  genera  Dysalotus  and  Kali,  with  descriptions 
of  two  new  species.  Arch.  Fischereiwiss.  25:13-46. 

— ,  and  R.  M.  Feltes.  1984.  A  new  species  of  r/wc/^erna  (Sal- 
moniformes:  Photichthyidae)  from  the  Red  Sea  and  Gulf  of  Aqaba, 
with  comments  on  the  depauperacy  of  the  Red  Sea  mesopelagic 
fish  fauna.  Fieldiana  Zool.  N.  S.  22. 

— ,  AND  D.  W.  Greenfield.    1976.    A  new  chaenopsid  fish.  Em- 


blemana  hylloni.  from  Isla  Roatan,  Honduras.  Fieldiana  Zool.  70(2): 
13-28. 

Johnson,  S.  E.  1918.  Osteology  of  the  gruntfish,  Rhamphocotlus  nch- 
ardsoni.  J.  Morphol.  31:461-477. 

JoLicoEUR.  P.  1963.  The  multivanate  generalization  of  the  allometry 
equation.  Biometrics  19:497-499. 

Jolly,  V.  H.  1967.  Observations  on  the  smelt  Relropinna  lacustris 
Stokell.  N.Z.  J.  Sci.  10:330-355. 

Jones,  A.  1972.  Studies  on  egg  development  and  larval  rearing  of 
turbot,  Scophthalmus  maximus  L..  and  bnll  Scophthalmus  rhom- 
bus L.,  in  the  laboratory.  J.  Mar.  Biol.  Assoc.  U.K.  52:965-986. 

Jones,  A.  C.  1962.  The  biology  of  the  euryhaline  fish  Leptocollus 
armatus  armalus  Girard.  Univ.  Calif  Publ.  Zool.  67(4):32 1-368. 

Jones,  G.  G.,  and  M.  A.  Tabery.  1980.  Larval  development  of  the 
banded  killifish  (Fundulus  diaphanus)  with  notes  on  the  distribu- 
tion in  the  Hudson  River  estuary,  p.  25-35.  In:  Proc.  Fourth  Ann. 
Urval  Fish  Conf  L.  A.  Fuiman  (ed.).  U.S.  Fish  Wildl.  Serv.,  Na- 
tional Power  Plant  Team,  Ann  Arbor,  Mich.,  FWS/OBS-80/43. 

Jones,  M.  P..  F.  G.  T.  Holliday  and  A.  E.  G.  Dlinn.  1966.  The 
ultrastructure  of  the  epidermis  of  larvae  of  the  herring  (Clupea 
harengus)  in  relation  to  the  rearing  salinity.  J.  Mar.  Biol.  Assoc. 
U.K.  46:235-239. 

Jones,  P.  W.,  F.  D.  Martin  and  J.  D.  Hardy,  Jr.  1978.  Development 
of  fishes  of  the  mid-Atlantic  bight,  an  atlas  of  egg.  larval,  and 
juvenile  stages.  Volume  1.  Acipenseridae  through  Ictaluridae.  U.S. 
Fish  Wildl.  Serv.,  Biol.  Serv.  Prog.  FWS/OBS-78/12. 

Jones,  R.,  and  W.  B.  Hall.  1974.  Some  observations  on  the  popu- 
lation dynamics  of  the  larval  stage  in  the  common  gadoids,  p.  87- 
102.  /«.■  The  early  life  history  offish.  J.  H.  S.  Blaxter  (ed.).  Springer 
Verlag,  Berlin. 

Jones,  S.  1937.  Observations  on  the  breeding  habits  and  development 
of  certain  brackish  water  fishes  of  Adyar,  Madras.  Proc.  Indian 
Acad.  Sci.  Sect.  B  5:261-289. 

.    1938.   On  the  breeding  habits  and  development  of  a  cypnnid, 

Danio  (Danio)  malabaricus  (Jerdon)  in  Ceylon.  Ceylon  J.  Sci.  Sec. 
C  6:79-90. 


LITERATURE  CITED 


713 


,  AND  M.  Kljmaran.    1962.   Notes  on  eggs,  larvae  and  juveniles 

of  fishes  from  Indian  waters.  Xll.  Myriprtslis  inurdjan.  Xlll.  Hol- 

ocenlrus  %p.  Indian  J.  Fish.  Sect.  A  9(  1):  155-167. 
.AND .    1964.   On  the  fishes  of  the  genus  .S'c7;»!(y/£'rM  Gil- 

tay  from  the  Indian  Ocean.  J.  Mar.  Biol.  Assoc.  India  6:257-264. 
,  AND .     1967.    Notes  on  eggs,  larvae,  and  juveniles  of 

fishes  from  Indian  waters.  XV.  Pegasus  volilans  Linnaeus.  XVI. 

Daaytoptena  onentalis  (Cuvier  and  Valenciennes).  XVII.  Dacty- 

loptena  macracanlhus  Bleeker).  Indian  J.  Fish.  1 1:232-246. 
.AND .    1950.    Spawning  habits  and  development  of  the 

Gangetic  anchovy,  Selipinna  phasa  (Hamilton).  Curr.  Sci.  19:25. 
,  AND .    1951.   Notes  on  the  bionomics  and  developmental 

stages  of  some  Indian  flatfishes.  J.  Zool.  Soc.  India  3:71-83. 
,  AND  .     1952.    Observations  on  the  development  and 

systematics  of  the  fishes  of  the  genus  Coilia  Gray.  J.  Zool.  Soc. 

India  4: 17-36. 
.  AND  V.  B.  Pantulu.    1958.   On  some  larval  and  juvenile  fishes 

from  the  Bengal  and  Orissa  coasts.  Indian  J.  Fish.  5:1  18-143. 
Jordan.  D.  S.     1892.    Relations  of  temperature  to  vertebrae  among 

fishes.  Proc.  U.S.  Natl.  Mus.  14:107-120. 
.    1923.    A  classification  of  fishes  including  families  and  genera 

as  far  as  known.  Stanford  Univ.  Publ.  Biol.  Sci.  3(2):77-243. 
.  and  B.  W.  Evermann.    1896-1900.    The  fishes  of  North  and 

Middle  Amenca.  U.S.  Natl.  Mus.  Bull.  47. 
,  AND  C.  H.  Gilbert.    1899.    The  fishes  of  the  Bering  Sea.  p. 


433-492.  In:  Fur-seals  and  fur-seal  islands  of  the  North  Pacific 
Ocean,  Pt.  3.  D.  S.  Jordon  (ed.).  Washington,  D.C. 

,  AND  E.  C.  Starks.    1904.   A  review  of  the  Cottidae  or  sculpins 

found  in  the  waters  of  Japan.  Proc.  U.S.  Natl.  Mus.  27:231-335. 

JiDE.  D.  J.  1982a.  Family  Cypnnodontidae,  p.  477-486.  In:  Identi- 
fication of  larval  fishes  of  the  Great  Lakes  basin  with  emphasis  on 
the  Lake  Michigan  drainage.  N.  A.  Auer  (ed.).  Great  Lakes  Fish. 
Comm.  Spec.  Publ.  82-3. 

.  1982b.  Family  Gadidae,  codfishes,  p.  470-476.  In:  Identifi- 
cation of  larval  fishes  of  the  Great  Lakes  basin  with  emphasis  on 
the  Lake  Michigan  drainage.  N.  A.  Auer  (ed.).  Great  Lakes  Fish. 
Comm.  Spec.  Publ.  82-3. 

JiiNGERSEN.H.  F.  E.  1910.  Ichthyotomical  contributions.  II.  The  struc- 
ture of  the  Aulostomidae,  Syngnathidae  and  Solenostomidae.  K. 
Dan.  Vidensk.  Selsk,  Skr.  Ser.  8,  7:268-364. 

K.ADAM,  K.  M.  1961.  The  development  of  the  skull  in /Verop/j/i.  Acta. 
Zool.  (Stockh.)  42:257-298. 

Kahan,  D.,  G.  Uhlig,  D.  Schwenzer  and  L.  Horowitz.  1982.  A 
simple  method  for  cultivating  harpacticoid  copepods  and  offering 
them  to  fish  larvae.  Aquaculture  26:303-310. 

K,MLL,  W.  M.  1967.  Ecology  and  behavior  of  the  cyprinodontid  fishes 
Jordanellaflohdae Goode  and  Bean.  Floridichthys carpio  (Gunther) 
and  Cypnnodon  variegatus  Lacepede.  LIrhpubl.  PhD  dissertation. 
Cornell  Univ.,  Ithaca,  NY. 

Kaiser,  C.  E.  1973.  Age  and  growth  of  horse  mackerel  Trachiinis 
mMrp/iy/ off" Chile.  Trans.  Am.  Fish.  Soc.  102(3):591-595. 

Kamlva,  T.  1925.  The  pelagic  eggs  and  larvae  of  fishes  in  the  Tateya- 
ma  Bay  (Pref  Chiba)  3rd  Report.  J.  Imp.  Fish.  Inst.  (Tokyo)  21(3): 
27-35. 

Kanazawa,  R.  H.  1952.  More  new  species  and  new  records  of  fishes 
from  Bermuda.  Fieldiana  Zool.  34(7):7I-100. 

Kanidev,  A.  N.  1961.  Embryonic  development  of  .4//u'n«a /(epienis 
and  A.  mochon  pontica.  Tr.  Karadag.  Biol.  Stn.  17:24-45. 

Karamchandani,  S.  J.,  AND  M.  P.  MoTWANi.  1955.  Early  life-history, 
bionomics  and  breeding  of /?«a  r;7a  (Hamilton).  J.  Zool.  Soc.  India 
7:115-126. 

.and .    1956.   On  the  larval  development  of  four  species 

of  freshwater  cat  fishes  from  the  River  Ganga.  J.  Zool.  Soc.  India 
8(1):  19-34. 

Karrer,  C.  1976.  Uber  drei  mesopelagische  fisharten  aus  dem  Golf 
von  Guinea.  Mitt.  Zool.  Mus.  Berl.  52:177-182. 

.    1982.    Anguilliformes  du  Canal  de  Mozambique  (Pisces.  Te- 

leostei).  Faune  Trop.  23:1-1 16. 


Katayama,  M.  1960.  Fauna  Japonica  Serranidae.  (Pisces).  Tokyo 
News  Serv.,  Ltd..  Tokyo. 

Kaufman,  L.  S..  and  K.  F.  Liem.  1982.  Fishes  of  the  suborder  La- 
broidei  (Pisces:  Perciformes):  phylogeny,  ecology,  and  evolutionary 
significance.  Breviora  (472):  1-1 9. 

K^WAGLiCHi,  K.,  AND  K.  Aioi.  1972.  Myctophid  fishes  of  the  genus 
Myctophum  (Myctophidae)  in  the  Pacific  and  Indian  Oceans.  J. 
Oceanogr.  Soc.  Jpn.  28:161-175. 

,  AND  J.  L.  Butler.    In  press.    Fishes  of  the  genus  Nansenia 

(Microstomatidae)  with  descriptions  of  seven  new  species.  Contrib. 
Sci.  (Los  Ang.). 

,  H.  Ikeda,  M.  Tamura  and  S.  Ueyanagi.    1972.  Geographical 

distribution  of  surface  migrating  myctophid  fishes  (genus  Mycto- 
phum) in  the  tropical  and  subtropical  Pacific  and  Indian  Oceans. 
Bull.  Far.  Seas  Fish.  Res.  Ub.  (Shimizu)  6:23-37. 

,  and  R.  M ARUMO.    1967.   Biology  of  Gonostoma  gracile  (Gon- 

ostomatidae).  1.  Morphology,  life  history  and  sex  reversal,  p.  253- 
269.  In:  Information  Bulletin  on  Planktology  in  Japan.  Commem- 
oration Number  of  Dr.  Y.  Matsue. 

.  AND  J.  Mauchline.    1982.   Biology  of  myctophid  fishes  (family 

Myctophidae)  in  the  Rockall  Trough.  Northeastern  Atlantic  Ocean. 
Biol.  Oceanogr.  1:337-373. 

Ka-i  A,  C.  M.,  A.  E.  DizoN,  S.  D.  Hendrix.  T.  K.  Kaza.ma  and  M.  K. 
K.  Queenth.  1982.  Rapid  and  spontaneous  maturation,  ovula- 
tion, and  spawning  of  ova  by  newly  captured  skipjack  tuna,  Kal- 
suwonus  pelamis.  Fish.  Bull.  U.S.  80:393-396. 

Keene,  M.  1973.  Mclamphaes  ehelingi.  a  new  species  of  beryciform 
fish  (Melamphaidae)  from  the  North  Atlantic.  Copeia  1973:794- 
800. 

Kelvin,  T.  M.,  G.  Phillips,  J.  E.  Thomerson  and  D.  C.  Taphorn. 
1980.  Egg  chorion  ornamentation  and  taxonomy  of  the  annual 
killifishes  Auslrofundutus.  Rachovia.  and  Terranalos.  Am.  Soc.  Ich- 
thyol.  Herpetol.,  60lh  Annual  Meeting.  Fort  Worth.  Texas.  (Ab- 
stract.) 

Kelley,  S.  MS.  A  description  of  laboratory-reared  larvae  of  the  labrid 
fish,  Lachnolaimus  maximus. 

Kendall,  A.  W.,  Jr.  1972.  Descnption  of  black  sea  bass,  Cen/ropmnj 
striata  (Linnaeus),  larvae  and  their  occurrence  north  of  Cape  Look- 
out. North  Carolina,  in  1966.  Fish.  Bull.  U.S.  70:1243-1260. 

.    1976.    Predorsal  and  associated  bones  in  serranid  and  gram- 

mistid  fishes.  Bull.  Mar.  Sci.  26:585-592. 

.     1977.    Relationships  among  American  serranid  fishes  based 

on  the  morphology  of  their  larvae.  Unpubl.  PhD  dissertation.  Univ. 
Calif  San  Diego,  San  Diego,  CA. 

.    1979.  Morphological  comparisons  of  North  American  sea  bass 

larvae  (Pisces:  Serranidae).  NOAA  Tech.  Rep.  NMFS  Circ.  428. 

.    1982.   Elucidating  relationships  among  fishes  using  larval  char- 


acters. Mimeogr. 

,  AND  M.    P.    FaHAY. 


1979.    Larva  of  the  serranid  fish  Gonio- 


plectrus  hispanus  with  comments  on  its  relationships.  Bull.  Mar. 
Sci.  29:117-121. 

,  AND  D.  Gordon.     1981.    Growth  rate  of  Atlantic  mackerel 

(Scomber  scombrus)  larvae  in  the  middle  Atlantic  Bight.  Rapp.  P. 
V.  Reun.  Cons.  Int.  Explor.  Mer  178:337-341. 

,  C.  D.  Jennings,  T.  M.  Beasley,  R.  Carpenter  and  B.  L.  J. 

SoMAYAjuLU.  1983.  Discovery  of  a  fish  (zoarcid)  egg  cluster  bur- 
ied 10-12  cm  deep  in  Washington  continental  slope  sediments. 
Mar.  Biol.  74:193-200. 

,  andB.  ViNTER.   1984.   Development  of  hexagrammids  (Pisces: 

Scorpaeniformes)  in  the  Northeastern  Pacific  Ocean.  NOAA  Tech. 
Rep.  NMFS  2. 

Kendall,  W.  C,  and  E.  L.  Gole^sborough.  1911.  Reports  on  the 
scientific  results  of  the  expedition  to  the  tropical  Pacific,  in  charge 
of  Alexander  Agassiz,  by  the  U.S.  Fish  Commission  Steamer  "Al- 
batross," from  August  1899  to  March  1900.  13.  The  Shore  fishes. 
Mem.  Mus.  Comp.  Zool.  Harvard  26(7):293-344. 

Kennedy,  M.,  and  T.  Champ.    1971.   Occurrence  of  eggs  of  Echiodon 


714 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


drummondi  Thompson  on  the  coast  of  County  Kerry.  Ir.  Fish. 

Invest.  Ser.  B  (Mar.)7:55-59. 
AND  p.  FiTZMAURicE.    1969.    Pelagic  eggs  and  young  stages  of 

fishes  taken  on  the  south  coast  of  Ireland  in  1967.  Ir.  Fish.  Invest. 

Ser.  B  (Mar.)5:5-36. 
Khaldinova,  N.  a.     1936.    Study  of  spawning,  eggs,  and  larvae  of 

Elegmus  (Gadus)  namga  Pall.  Zool.  Zh.  15(2):32 1-339.  (In  Rus- 
sian, English  summary.) 
Khan,  N.  Y.    1972.    Comparative  morphology  and  ecology  of  the  pe- 
lagic larvae  of  nine  Cottidae  (Pisces)  of  the  northwest  Atlantic  and 

St.  Lawrence  drainage.  Unpubl.  PhD  thesis.  Univ.  Ottawa,  Canada. 
AND  D.  J.   Faber.     1974.     A  comparison  of  larvae  of  the 

deepwater  and  fourhom  sculpin.  Myoxocephalus  quadricorms  L. 

from  North  Amenca.  1 .  Morphological  development,  pp.  703-7 1 2. 

In:  The  early  life  history  offish.  Blaxter,  J.  H.  S.  (ed).  Springer- 

Verlag,  New  York. 
KiDO,  K.    1983.   New  and  rare  liparidid  species  from  the  Okhotsk  and 

Bering  Seas  and  their  adjacent  waters.  Jpn.  J.  Ichthyol.  29:374- 

384. 
Kjliachenkova,  V.  A.    1970.    Development  and  distribution  of  eggs 

and  larvae  of  Trachurus  Irachurus  L.  Rapp.  P.  V.  Reun.  Cons.  Int. 

Explor.  Mer.  159:194-198. 
Kjm,  Y.  U.,  P.  Chin,  T.  Y.  Lee  and  Y.  J.  Kang.    1981.   Studies  on  the 

fish  larvae  of  coastal  waters  in  Korea.  Publ.  Inst.  Mar.  Sci.  Natl. 

Fish.  Univ.  Busan  13:1-35. 
Kindred,  J.  E.    1921.    The  chondrocranium  of  Synanathus  fuscus.  J. 

Morphol.  35:425-456. 
KiNG,  D.  P.  F.,  M,  J.  O'Toole  and  A.  A.  Robertson.    1977.    Early 

development  of  the  south  African  maasbanker,  Trachurus  trachu- 
rus, at  controlled  temperatures.  Fish.  Bull.  S.  Afr.  9:16-22. 
Kjnne,  O.    1977.    Research  Cultivation:  (13)  Pisces,  p.  968-1035.  In: 

Marine  ecology.  Vol.  Ill,  part  2.  O.  Kinne  (ed.).  John  Wiley  &  Sons, 

Chichester. 
Kirschbaum,  F.    1984.    Reproduction  of  weakly  electric  teleosts:  just 

another  example  of  convergent  development?  Environ.  Biol.  Fishes 

10:3-14. 
,  and  J.  p.  Denizot.    1975.    Sur  la  differentiation  des  electro- 

recepteurs  chez  Marcusenius  sp.  (Mormyrides)  et  Eigenmanma  vi- 

rescens  (Gymnotoides),  poissons  electriques  a  faible  decharge.  C. 

R.  Hebd.  Sciences  Acad.  Sci.  Ser.  D.  Sci.  Nat  28:419-422. 
,  and  G.  W.  M.  Westby.    1975.    Development  of  the  electric 

discharge  in  mormyrid  and  gymnotid  fish  {Marcusenius  sp.  and 

Eigenmannia  virescens).  Experientia  (Basel)  31:1290-1293. 
Klawe,  W.  L.    1960.  Animals  from  night-light  collections.  In:  Physical, 

chemical,  and  biological  observations  in  the  eastern  Tropical  Pacific 

Ocean  — Scot  Expedition,  April-June  1958.  R.  W.  Holmes  and  M. 

Blackburn  (eds.).  U.S.  Fish  Wildl.  Serv.  SSRF  345. 
Klee,  a.  J.    1982.    Comments  on  the  genus  Rivulus.  J.  Am.  Killifish 

Assoc.  15(3):76-79. 
Klein-Macphee,  G.,  W.  H.  Howell  and  A.  D.  Beck.    1982.    Com- 
parison of  a  reference  strain  and  four  geographical  strains  of  Artemia 

as  food  for  winter  flounder  (Pseudopleuronectes  americanus)  larvae. 

Aquaculture  29:279-288. 
Kliever,  R.  G.    1976.    Natural  history  of  May nea  californica  (Pisces: 

Zoarcidae)  in  a  drift  seaweed  habitat  in  the  Monterey  Submarine 

Canyon,  Monterey  Bay,  California.  Unpubl.  MA  Thesis.  San  Jose 

State  Univ.,  San  Jose,  CA. 
Kluge,  A.  G.,  AND  J.  S.  Farris.    1969.   Quantitative  phyletics  and  the 

evolution  of  anurans.  Syst.  Zool.  18:1-32. 
Kobayashi,  K.    1957.   Larvae  and  young  of  the  sable  fish,  ,-l«op/opoOTa 

fimbria  (Pallas),  from  the  sea  near  the  Aleutian  Islands.  Bull.  Jpn. 

Soc.  Sci.  Fish  23:376-382.  (In  Japanese.  English  abstract.) 
.    1961a.   Young  of  the  wolflSsh,  Anarchichas  orientalis  (Pallas). 

Bull.  Fac.  Fish.  Hokkaido  Univ.  12(l):l-4.  (In  Japanese,  English 

abstr.) 
.    1961b.    Larvae  and  young  of  the  quillfish,  Plilichlhys  goodei 

(Bean),  from  the  Okhotsk  Sea.  Bull.  Fac.  Fish.  Hokkaido  Univ. 

12(l):5-8.  (In  Japanese,  English  abstr.) 


-.    1961c.    Larvae  and  young  of  the  sand  lance,  .-I  »imo(/i7e.s /le.v- 

aplerus  (Pallas)  from  the  nonh  Pacific.  Bull.  Fac.  Fish.  Hokkaido 
Univ.  12:111-120. 

.    1 962.    Lai-vae  of  the  smooth  lumpsucker,  Aptocychis  venlri- 

cosus  (Pallas),  with  discussion  on  revision  of  the  taxonomy  of  the 
species.  Bull.  Fac.  Fish.  Hokkaido  Univ.  13(3):  153-1 64. 

,  AND  K.  Abe.    1962.    Studies  on  the  larvae  and  young  of  fishes 

from  the  boundary  zones  off' the  south-eastern  coast  of  Hokkaido, 
Japan.  Bull.  Fac.  Fish.  Hokkaido  Univ.  13:165-179.  (In  Japanese, 
English  summary.) 

,  andM.  Igarashi.    1961.   Larvae  and  young  of  GV /-('//a  pKMtVa/a 

Gray  from  Hakodate  Bay  in  southern  Hokkaido.  Bull.  Fac.  Fish. 
Hokkaido  12:121-127.  (In  Japanese,  English  abstract.) 

,  M.  Mikawa,  AND  J.  Ito.    1968.    Descriptions  of  the  young  and 

one,  immature  adult  specimen  of  the  coster  dory,  Altocyllus  ver- 
rucosus (Gilchrist)  from  the  northern  pari  of  the  Pacific.  Bull.  Fac. 
Fish.  Hokkaido  Univ.  19(1):  1-5. 

Kobayashi,  T.,  Y,  Dotsli  and  N.  Miura.  1973.  Egg  development 
and  rearing  experiments  of  the  larvae  of  the  mud  skipper,  Per- 
iophthalmus  cantonensis.  Bull.  Fac.  Fish.  Nagasaki  Univ.  33:49- 
62. 

Kobayasi,  H.,  and  S.  Moriyana.  1957.  On  the  development  of  a 
loach,  Barbalula  toni  areas  (Jordan  et  Fowler).  Jpn.  J,  Ichthyol. 
6:177-183. 

Kobelkowsky,  A.  D.,  and  A.  M.  Resendez.  1972.  Estudio  compar- 
ativo  del  endosqueleto  de  Mugil  cephahis  y  Mugil  curema  (Pisces, 
Perciformes).  An.  Inst.  Biol.  Univ.  Nac.  Auton.  Mex.  Ser.  Cienc. 
Mar.  Limnol.  43(1):33-81. 

KoBLiTSicMA,  A.  F.  1981.  Opredelitel"  molodi  presnovodnykh  ryb. 
Legkaia  i  pishevaia  promyshlennost",  Moskova. 

KocK,  K.  H.  1981.  Fischereibiologische  Untersuchungen  an  drei  ant- 
arktischen  Fischarien:  Champsocephaius  gunnari  Lonnberg,  1905, 
Chaenocephalus  aceratus  (Lonnberg,  1906)  und  Pseudochaen- 
ichlhys  georgianus  Norman,  1937  (Notothenioidei,  Channichthyi- 
dae).  Mitt.  Inst.  Seefisch.  32:1-226. 

KoEFOED,  E.  1909.  Poissons.  Croisiere  oceanographique  accompli  a' 
bord  de  la  Belgica  dans  la  mer  de  Gronland  1905.  Bruxelles. 

.    1927.    Fishes  from  the  sea-bottom.  Rep.  Sci.  Results  Michael 

Sars  N.  Atl.  Deep-Sea  Exped.  4(Pt.  1):  1-147. 

.     1953.    Synentognathi,  Solenichthyes,  Anacanthini,  Beryco- 


morphi,  Xenoberyces.  Rep.  Sci.  Results  Michael  Sars  N.  Atl.  Deep- 

Sea  Exped.  4(Pt.  2,  No.  3):3-36. 
KoENiG,  C.  C,  AND  R.  J.  Livingston.    1976.    The  embryological  de- 
velopment of  the  diamond  killifish  (Adinia  xentca).  Copeia  1976: 

435-445. 
KoHNO,  H.,  T.  HosHiNO,  F.  Yasiida  and  Y.  Taki.    1982.    Larval  me- 

lanophore  patterns  of  Thunnus  alatunga  and  T.  thynmis  from  the 

Mediterranean.  Jpn.  J.  Ichthyol.  28:461-465. 
K0K.HANOVA,  N.  A.     1957.    Razvitie  shchipovki  (Cohitis  taenia  L.). 

Vopr.  Ikhtiol.  8:89-101. 
KoMAROv,  Y.  A.    1964.    Some  data  of  the  reproduction  of  the  horse 

mackerel- rrac/jwr;/.?  trachurus  L.  from  the  Southwest  African  coast. 

Tr.  Atl.  Nauchno-Issled.  Inst.  Rybn.  Khoz.  Okeanogr.  1  1:87-99. 
Kopec,  J.  A.    1 949.   Ecology,  breeding  habits  and  young  stages  of  Cren- 

ichthys  baileyi.  a  cyprinodont  fish  of  Nevada.  Copeia  1949:56-61. 
KORMENDY.  A.  C.    1975.    Microorganisms,  p.  82-108.  In:  Principles 

and  techniques  of  scanning  electron  microscopy.  Vol.  3.  M.  A. 

Hayat  (ed.).  Van  Nostrand  Reinhold  Co..  New  York. 
KosTER,  W.  J.    1948.    Notes  on  the  spawning  activities  and  the  young 

slzges  of  Plancterus  kansae  (GiTman).  Copeia  1948:25-33. 
KoTLYAR,  A.  N.     1981.    A  new  family,  genus  and  species  of  Beryci- 

formes,  Hispidoberycidae  fam.,  Hispidoberyx  ambagiosus  gen.  et 

sp.  n.  (Beryciformes).  J.  Ichthyol.  (Engl.  Transl.  Vopr.  Ikhtiol.) 

21(3):9-13. 
KoTTHAUs,  A.   1969.   2.  FischedesIndischenOzeans.  A.  Systematisch- 

er  Teil,  V.  Solenichthyes  und  Anacanthini.  Meteor  Forschungser- 

geb.  Reihe  D.  Biol.  4:32-46. 
.    1970.   Flagelloserranus  a  new  genus  of  serranid  fishes  with  the 


LITERATURE  CITED 


715 


description  of  two  new  species  (Pisces,  Percomorphi).  Dana-Rep. 
Carlsberg  Found.  78. 

.    1972.    Fische  des  Indischen  Ozeans.  A.  Systematischer  Teil, 

IX.  Iniomi  (Nachtrag:  Fam.  Myctophidae).  Meteor  Forschungser- 
geb.  Reihe  D  Biol.  12:12-35. 

KouMANS,  F.  P.  193 1 .  A  preliminary  revision  of  the  genera  of  gobioid 
fishes  with  united  ventral  fins.  Drukkerij  Imperator.  N.V..  Lisse. 

Kdvalevskava,  N.  V.  1964.  Study  of  development  of  the  "two  winged" 
flying  fishes  of  the  genus  Exocoelus  (Exocoelidae,  Pisces).  Tr.  Inst. 
Okeanol.  Akad.  Nauk  SSSR  73:204-223.  (In  Russian,  English  sum- 
mary.) 

.    1965.   The  eggs  and  larvae  of  synentognathous  fishes  (Belon- 

iformes,  Pisces)  of  the  Gulf  of  Tonkin.  Tr.  Inst.  Okeanol.  Akad. 
Nauk.  SSSR  80:124-146.  (In  Russian,  English  summary.) 

.  1972.  New  materials  on  reproduction,  development  and  dis- 
tribution of  larvae  and  juveniles  of  flying  fishes  of  the  genus  Hi- 
rundichthys  (Exocoetus,  Pisces)  in  the  Pacific  and  Indian  Oceans. 
Tr.  Inst.  Okeanol.  Akad.  Nauk  SSSR  93:42-69.  (In  Russian,  English 
summary.) 

.    1982.    Superfluous  reproduction  and  development  of  flying 

fishes  of  the  family  Exocoetidae.  J.  Ichthyol.  (Engl.  Transl.  Vopr. 
Ikhtiol.)  22(4):48-54. 

KowTAL,  G.  V,  1972.  Observations  on  the  breeding  and  larval  de- 
velopment of  Chilka  'Sabal'  Eleutheronema  telradeclyturn  (Shaw). 
India  J.  Fish.  19(l-2):70-75. 

KoZHOV,  M.  1963.  Lake  Baikal  and  its  life.  Monographiae  Biologicae, 
Vol.  II.  Dr.  W.  Junk.  The  Hague. 

KozLOV,  B.  M.  1952.  Observations  on  the  development  of  navaga 
eggs.  Izv.  Tikhookean  Nauchno-Issled.  Inst.  Rybn.  Khoz.  Okean- 
ogr.  34:261-262.  (In  Russian.) 

Kramer,  D.  1960.  Development  ofeggs  and  larvae  of  Pacific  mackerel 
and  distribution  and  abundance  of  larvae  1 952-56.  U.S.  Fish  Wildl. 
Ser.  Fish.  Bull.  60:393-438. 

Kraul,  S.  1983.  Results  and  hypotheses  for  the  propagation  of  the 
grey  mullet,  MugU  cephatus  L.  Aquaculture  30:273-284. 

Krefft,  G.  1 96 1 .  a  contnbution  to  the  reproductive  biology  of  Hel- 
icolenus  daclylopterus  (De  la  Roche,  1809)  with  remarks  on  the 
evolution  of  the  Sebaslinae.  Rapp.  P.  V.  Reun.  Cons.  Perm.  Int. 
Explor.  Mer  150:243-244. 

.    1970.    Zur  Systematik  und  Verbreitung  der  Gattung  Lampa- 

dena  Goode  and  Bean.  1896  (Osteichthyes,  Myctophoidei,  Myc- 
tophidae) im  Atlantischen  Ozean,  mit  Beschreibung  einer  neen  Art. 
Ber.  Dtsch.  Wiss.  Komm.  Meeresforsch.  21:271-289. 

.    1974.    Investigations  on  midwater  fish  in  the  Atlantic  Ocean. 


Ber.  Dtsch.  Wiss.  Komm.  Meeresforsch.  23:226-254. 
Kryzhanovskii,  S.  G.    1947.    Sistema  semeistva  karpovykh  ryb  (Cy- 

prinidae).  Zool.  Zh.  26:53-64. 
.    1948.    Ekologicheskie  gruppy  ryb  i  zakonomemosti  ikh  raz- 

vitiia.  Izv,  Tikhookean  Nauchno-Issled.  Insl.  Rybn.  Khoz.  Okean- 

ogr.  27:5-114. 
.     1949.    Ekologo-morphologicheskie  zakonomemosti  razvitiia 

karpovykh,  v'iunoviykh  i  somovykh  ryb(Cyprinoidei  i  Siluroidei). 

Tr.  Inst.  Morfol.  Zhivotn.,  Akad.  Nauk  SSSR  1:1-232. 
,  A.  I.  Smirnov  and  S.  G.  Soin.    1951.    Materialy  po  razvitiiu 

ryb  r.  Amura.  Truy  Amurskoi  Ikhtiologicheskoi  Ekspeditsii  1945- 

1949.  Mater.  Poznaniyu  Fauny  Flory  SSSR  otd.  Zool.  24:5-222. 
KuBO,  I.    1939.    Note  on  the  development  of  a  teleost,  Thalassoma 

cupido  (Temminck  and  Schlegel).  Bull.  Jpn.  Soc.  Sci.  Fish.  8(3): 

165-167.  (In  Japanese.  English  summary.) 
KucHNOw,  K.  P.,  andJ.  R.  Scott.    1977.   Ultrastruclure  of  the  chorion 

and  its  micropyle  apparatus  in  the  mature  Fundulus  heleroclitus 

ovum.  J.  Fish  Biol.  10:197-201. 
KuMLMAN,  D.,  G.  Qliantz  AND  U.  WiTT.     1981.    Rearing  of  turbot 

larvae  {Scophlhahnus  ma.ximus  L.)  on  cultured  food  organisms  and 

postmetamorphosis  growth  on  natural  and  artificial  food.  Aqua- 
culture  23:183-196. 
KiiLKARNi,  C,  V.    1940.   On  the  systematic  position,  structural  modi- 


fications, bionomics  and  development  of  a  cyprinodont  fishes  from 

the  province  of  Bombay.  Rec.  Indian  Mus.  (Calcutta)  42:379-423. 
.    1950.   Breeding  habits,  eggs  and  early  life  history  of  the  Indian 

shad.  HUsa  ilisha  (Ham.),  in  the  Narbada  River,  Proc.  Natl.  Inst, 

Sci.  (India)  16:169-176. 
Kuntz,  A.   1914a.  Noteson  the  habits,  morphology  of  the  reproductive 

organs  and  embryology  of  the  viviparous  fish  Gambusia  affinis. 

Bull.  Bur.  Fish.  Wash.  33(1913):181-I90. 
.    1 9 1 4b.   The  embryology  and  larval  development  of  Bairdiella 

chrysura  and  Anchovia  mitchilli.  Bull.  Bur.  Fish.  Wash.  33:1-19. 
.     1916.    Notes  on  the  embryology  and  larval  development  of 

five  species  of  teleosteon  fishes.  Bull.  Bur.  Fish.  Wash.  34(1914): 

407-429. 
,  AND  L.  Radcliffe.    1917.   Notes  on  the  embryology  and  larval 

development  of  twelve  teleostean  fishes.  Bull.  Bur.  Fish.  Wash.  35: 

87-134. 
KuNZ,  Y.     1964.    Morphologische  Sludien  iiber  die  embryonale  und 

postembryonal  Entwicklung  bei  Teleostiem  mit  besonderer  Be- 

rijcksichtigung  des  Dotlersystems  und  der  Leber.  Rev.  Suisse  Zool. 

71:445-525. 
Kro.  C-M..  H.  Shehadeh  and  K.  K.  Milisen.    1973.    A  preliminary 

report  on  the  development,  growth  and  survival  of  laboratory  reared 

larvae  of  the  grey  mullet,  Mugil  cephalus  L.  J.  Fish  Biol.  5:459- 

470. 
Kurata,  H.    1956.    On  the  rearing  of  larvae  of  the  flatfish,  Liopsetta 

obscura  in  small  aquaria.  Bull.  Hokkaido  Reg.  Fish.  Res.  Lab.,  Fish. 

Agency  13:20-28. 
Kuroda,  K.,T.  Yamamoto  and  Y.  Hirano.    1982.   Development  and 

identification  of  the  egg  of  the  Pacific  mackerel  (Scomber  japonicus 

Houttuyn)  and  a  similar  egg.  Bull.  Tokai  Reg.  Fish.  Res.  107:33- 

52. 
KuTHALiNGAM,  M.  D.  K.    1959a.    A  contribution  to  the  life  histories 

and  feeding  habits  of  horse-mackerels,  Megalaspis  cordyla  (Linn.) 

and  Caranx  mate  (Cuv.  &  Val.)  and  notes  on  the  effect  of  absence 

of  light  on  the  development  and  feeding  habits  of  larvae  and  po%X- 

larvae  of  Megalaspis  cordvla.  J.  Madras  Univ.  Sect.  B  29(2):79- 

96. 
.     1959b.    The  life  history  and  feeding  habits  of  a  tripod  fish 

( Tnacanlhus  breviroslns  Temm.  &  Schleg.)  of  the  Indian  Sea.  Treu- 

bia  25:159-164. 
KuwABARA,  S.    1954.    Occurrence  of  luminous  organs  on  the  tongue 

of  two  scopelid  fishes,  Neoscopeliis  macrolepidotus  and  N.  micro- 

chir.  J.  Schimonoseki  Coll.  Fish.  3:283-287. 
Kuz'min-Karovaev.  G.  A.    1930.    Biology  of  the  White  Sea  navaga. 

Tr.  Nauchno.  Inst.  Rybn.  Khoz.  (5)  4:83-86.  (In  Russian.) 
Kvenseth,  P.  G.,  AND  V.  Oiestad.    1984.    Large-scale  rearing  of  cod 

fry  on  the  natural  food  production  of  an  enclosed  pond,  p.  644- 

655.  In:  The  propagation  of  cod.  E.  Dahl,  D.  Danielssen,  E.  Moks- 

ness  and  P.  Solemdal  (eds.).  Flodevigen  Rapp.  I. 
Kyi.e,  H.  M.    1913.    Flat-fishes  (Helerosomata).  Rep.  Dan.  Oceanogr. 

Exped.  Mediterr.  2(A.I):1-150. 
.    1921.   The  asymmetry,  metamorphosis  and  origin  of  flat-fish- 
es. Philos.  Trans.  R.  Soc.  London  Ser.  B  21 1:75-129. 
KvusHiN,  K.    1968.   The  embryonic  and  larval  stages  of //ew/rnpferwj 

v///oj«i  (Pallas).  Bull.  Fac.  Fish.  Hokkaido  Univ.  18:277-289. 
.    1970.    Embryonic  development  and  larvae  of  G.vmnoca«?/!!« 

herzensteini  JovAan  and  Starks.  Jpn.  J.  Ichthyol.  17:74-79. 

1975.    The  embryonic  and  larval  development,  growth,  sur- 


vival and  changes  in  body  form,  and  the  effect  of  temperature  on 
these  characteristics  of  the  smooth  lumpsucker.  Aplocychis  ventn- 
cosus  (Pallas).  Bull.  Fac.  Fish.  Hokkaido  Univ.  26(l):49-72. 

Laale,  H.  W.  1980.  The  perivitellinc  space  and  egg  envelopes  of  bony 
fishes:  a  review,  Copeia  1980:210-226. 

Laciardere,  F.  1980.  Developpement  du  ceteau,  Dicologoglossa  cu- 
neata  (Moreau)  (Poissons-Soleidae).  Cybium,  3rd  Ser.  4(4):61-81. 

,  and  a.  Aboussohan.  1981.  Developpement  du  ceteau,  Di- 
cologoglossa cuneala  (Moreau,  1881)  (Pisces,  Pleuronectiformes, 
Soleidae):  II  — Description  des  larves.  Cybium,  3rd  Ser.  5(2):53-72. 


716 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Lai,  J.  1 979.  The  development  of  stages  of  mullet,  Mugil  so-nig  Bas- 
ilewsky  L.  Acta.  Oceanol.  Sin.  1(1):  157-1 75. 

Lake,  J.  S.  1967.  Rearing  experiments  with  five  species  of  Australian 
freshwater  fishes.  II.  Morphogenesis  and  ontogeny.  Aust.  J.  Mar. 
Freshwater  Res.  18:155-173. 

Lam,  T.  J.  1982.  Applications  of  endocnnology  to  fish  culture.  Can. 
J.  Fish.  Aquat.  Sci.  39:111-137. 

Laroche,  J.  L.,  S.  Richardson  AND  A.  A.  Rosenberg.  1982.  Age  and 
growth  of  a  pleuroneclid,  Parophrys  velulus.  during  the  pelagic  lar- 
val period  in  Oregon  coastal  waters.  Fish.  Bull.  U.S.  80:93-104. 

Laroche,  W.  A.  1977.  Descnplion  of  larval  and  early  juvenile  ver- 
milion snapper,  Rhomboplites  aurorubens.  Fish.  Bull.  U.S.  75:547- 
554. 

.    1981.   Development  of  larval  smooth  flounder,  Z.;opieKa/)Mr- 

nami.  with  a  redescription  of  winter  flounder,  Pseudopleuronectes 
amencanus  (Family  Pleuronectidae).  Fish.  Bull.  U.S.  78:897-909. 

,  AND  S.  L.  Richardson.    1980.    Development  and  occurrence 

of  larvae  and  juveniles  of  the  rockfishes  Sebasles  JIavidits  and  Se- 
bastes  melanops  (Scorpaenidae)  off  Oregon.  Fish.  Bull.  U.S.  77: 
901-924. 

,  and .    1981.    Development  of  larvae  and  juveniles  of 

the  rockfishes  Sebasles  enlomelas  and  5.  zacemrus  (Family  Scor- 
paenidae) and  occurrence  ofli' Oregon,  with  notes  on  head  spines  of 
S.  mrslinus.  S.  flavidus  and  S.  melanops.  Fish.  Bull.  U.S.  79:231- 
257.' 

,  D.  Rupee  and  S.  L.  Richardson.   MS.    Young  carangid  fishes 

from  the  Gulf  of  Mexico:  A  generic  diagnosis  of  larvae,  with  ad- 
ditional descriptive  notes  including  early  development  of  three 
species.  Gulf  Coast  Research  Lab.,  Ocean  Springs,  MS  39564. 

L.\siv£R,  R.  1975.  Field  criteria  for  survival  of  anchovy  larvae:  the 
relation  between  inshore  chlorophyll  maximum  layers  and  suc- 
cessful first  feeding.  Fish.  Bull.  U.S.  73:453-462. 

.    1981.    The  role  of  a  stable  ocean  in  larval  fish  survival  and 

subsequent  recruitment,  p.  80-87.  In:  Marine  fish  larvae.  Mor- 
phology, ecology  and  relation  to  fishenes.  R.  Lasker  (ed.).  Univer- 
sity of  Washington  Press,  Seattle,  WA. 

,  H.  M.  Feder,  G.  H.  TheilackerandR.C.  May.  1970.  Feed- 
ing, growth  and  survival  of  Engraulis  mordax  reared  in  the  labo- 
ratory. Mar.  Biol.  5:345-353. 

,  AND  K.  Sherman  (eds.).    1981.    The  early  life  history  offish: 


recent  studies.  Rapp.  P.V.  Reun.  Cons.  Int.  Explor.  Mer  178:1- 

607. 
,  and  L.  T.  Threadgold.    1968.   "Chloride  cells"  in  the  skin  of 

larval  sardine  (Sardinops  caendea).  Exp.  Cell  Res.  52:582-590. 
Lau,  S.  R.,  and  p.  L.  Shafland.    1982.    Larval  development  of  the 

snook,  Ceniropomus  undecimalis  (Pisces:  Centropomidae).  Copeia 

1982:618-626. 
Lauder,  G.  V.,  Jr.    1980.   On  the  relationship  of  the  myotome  to  the 

axial  skeleton  in  vertebrate  evolution.  Paleobiology  6:51-56. 
,  AND  K.  F.  LiEM.    1983.    The  evolution  and  interrelationships 

of  the  Actinopterygian  fishes.  Bull.  Museum  Comp.  Zool.  Harv. 

Univ.  150(31:95-197. 
Laurence,  G.  C.    1981.    Overview  — modelling— an  esoteric  or  poten- 
tially utilitarian  approach  to  understanding  larval  fish  dynamics. 

Rapp.  P.  V.  Reun.  Cons.  Int.  Explor.  Mer  178:3-6. 
Lea,  R,  N.    1980.    Systematics  and  zoogeography  of  cusk-eels  of  the 

family  Ophidiidae,  subfamily  Ophidiinae.  from  the  eastern  Pacific 

Ocean.  Unpubl.  PhD  dissertation.  Univ.  Miami,  Coral  Gables,  FL. 
Leach,  G.  C.    1923.    Artificial  propagation  of  whitefish.  grayling  and 

lake  trout.  Rep.  U.S.  Comm.  Fish.  Appendix  3,  Doc.  949:1-32. 
Lebour,  M.  V.    1919.   The  young  of  the  Gobiidae  from  the  neighbor- 
hood of  Plymouth.  J.  Mar.  Biol.  Assoc.  U.K.  12:48-80. 
.    1921.   The  larval  and  post-larval  stages  of  the  pilchard,  sprat, 

and  herring  from  the  Plymouth  District.  J.  Mar.  Biol.  Assoc.  U.  K. 

12:427-457. 
.     1927.    The  eggs  and  newly  hatched  young  of  the  common 

blennies  from  the  Plymouth  neighbourhood.  J.  Mar.  Biol.  Assoc. 

U.K.  14:647-650. 
Le  Danois,  Y.    1959.   Etude  osteologique,  myologique  et  systematique 


des  poissons  du  sous— ordre  des  Orbiculales.  Ann.  Inst.  Oceanogr. 

36(1). 
.    1961.    Remarques  sur  les  poissons  orbiculates  du  sous-ordre  ^ 

des  Ostracioniformes.  Mem.  Mus.  Natl.  Hist.  Nat.  Ser.  A  Zool.  29: 

207-338. 
Legand,  M.     1967.    Cycles  biologiques  des  poissons  mesopelagiques 

dans  Test  de  I'ocean  Indian.  Premiere  note.  Scopelopsis  midti- 

punclalus  Brauer,  Gonostoma  sp..  Notolychnus  valdiviae  Brauer. 

Cah  ORSTOM  Ser.  Oceanogr.  5(4):47-7l. 
Leibi,  M.  M.    1979a.   Leptocephalus  larvae  of  the  eel  family  Ophichth- 

idae.  I.  Ophichlhus  gomesi  Ca^le\nau.  Bull.  Mar.  Sci.  29:329-343. 
.    1979b.   Morphological  development  of  the  eel  Myrophis  punc- 

tatus  (Ophichthidae)  from  hatching  to  metamorphosis,  with  em- 
phasis on  the  developing  head  skeleton.  Bull.  Mar.  Sci.  29:509- 

521. 
.    1981.    Larval  morphology  of  the  eels  5a«'a«;t7!?/;i'i /la^ca/J- 

ium.   B    scuticaris.   Ophichthus   melanoporus  and    O.   ophis 

(Ophichthidae),  with  a  discussion  of  larval  identification  methods. 

Bull.  Mar.  Sci.  31:46-71. 
.    1982.  Leptocephalus  larvae  ofthetnbeSphagebranchini  (Pisces. 

Ophichthidae)  in  the  western  North  Atlantic.  Bull.  Mar.  Sci.  32: 

220-236. 
.     1984.    Leptocephalus  larvae  of  the  tribe  Callechelyini  (An- 

guilliformes,  Ophichthidae,  Ophichthinae)  in  the  western  North 

Atlantic.  Bull.  Mar.  Sci.  34:398^23. 
.    In  press.    Ophichthidae.  /«.  Leptocephalus  larvae  of  the  Gulf 


of  Mexico  and  Yucatan  Channel.  D.  G.  Smith  and  W.  H.  Hulet 

(eds.).  Stud.  Trop.  Oceanogr.  (Miami). 
Leim,  a.  H.,  and  W.  B.  Scott.    1966.    Fishes  of  the  Atlantic  coast  of 

Canada.  Bull.  Fish.  Res.  Board  Can.  155. 
Leis,  J.  M.    1977.    Development  of  the  eggs  and  larvae  of  the  slender 

mola,  Ranzania  laevis  (Pisces,  Molidae).  Bull.  Mar.  Sci.  27:448- 

466. 
.    1978.    Systematics  and  zoogeography  of  the  porcupine  fishes 

{Diodon.  Diodontidae,  Tetraodonliformes),  with  comments  on  egg 

and  larval  development.  Fish.  Bull.  LI.S.  76:535-567. 
.     1982.     Hawaiian  creediid  fishes  (Cryslallodytes  cookei  and 

Limnichthys  donaldsoni):  development  of  eggs  and  larvae  and  use 

of  pelagic  eggs  to  trace  coastal  water  movement.  Bull.  Mar.  Sci.  32: 

166-180. 
.     1983.    Coral  reef  fish  larvae  (Labndae)  in  the  East  Pacific 

Barrier.  Copeia  1983:826-828. 
,  andB.  GC5LDMAN.    1983.   Studies  On  the  biology  of  larval  fishcs 

in  the  Lizard  Island  area.  Northern  Great  Barrier  Reef,  p.  22 1  -22  5 . 

In:  Proc.  Inaugural  Great  Barrier  Reef  Conf  i.  T.  Baker.  R.  M. 

Carter,  P.  W.  Sammarco  and  K.  P.  Stark  (eds.).  J.  Cook  University 

Press,  Townsville,  Australia. 
,  AND  J.  T.  Mover.    MS.    Development  of  eggs  and  larvae  of 

three  Indo-Pacific  trunkfishes  (Tetraodonti formes,  Ostraciidae). 
,  AND  D.  S.  Rennis.    1983.   The  larvae  of  Indo-Pacific  coral  reef 


fishes.  New  South  Wales  University  Press,  Sydney. 
Lentz,  T.  L..  AND  J.  P.  Trinkhaus.    1967.    A  fine  structural  study  of 

cytodifferentiation  during  cleavage,  blastula.  and  gastrula  stages  of 

Fundidus  hcteroditus.  J.  Cell.  Biol.  32:121-138. 
Leonard,  S.  B.    1971.    Larvae  of  the  fourspot  flounder,  Hippoglossina 

oblonga  (Pisces:  Bothidae),  from  the  Chesapeake  Bight,  western 

North  Atlantic.  Copeia  1971:676-681. 
LEONci,  R.    1971.   Induced  spawning  of  the  northern  anchovy  Engraulis 

mordax  Girard.  Fish.  Bull.  U.S.  69:357-360. 
.     1977.    Maturation  and  induced  spawning  of  captive  Pacific 

mackerel.  Scomber japomcus.  Fish.  Bull.  U.S.  75:205-21  I. 
Letaconnoux,  R.    1951.    Contribution  a  I'etude  des  especes  du  genre 

Trachuriis  et  specialement  du  Trachurus  irachiirus  (L.).  Mem.  Off. 

Sci.  Pech.  Mant.  15:1-67. 
Levings,  C.  D.    1 969.   The  zoarcid  Lycodopsis pacifica  in  outer  Burtard 

Inlet,  British  Columbia.  J.  Fish.  Res.  Board  Can.  26:2403-2412. 
LlAO,  I.  C,  J.  C.  JUARIO,  S.  KUMACiAI,  H.  Nakajima.  M.  Natividad 

and  p.  Buri.    1979.    On  the  induced  spawning  and  larval  rearing 

of  the  milkfish,  Chanos  chanos  (Forskal).  Aquaculture  18:75-93. 


LITERATURE  CITED 


717 


LiMBAUGH,  C.  1962.  Life  history  and  ecological  notes  on  the  tubenose, 
Aulorhynchus  IJavidus.  a  hemibranch  fish  of  western  North  Amer- 
ica. Copeia  1962:549-555. 

LiNDBERG,  G.  U.,  AND  Z.  V.  Krasyukova.  1975.  Fishcs  of  the  Sea  of 
Japan  and  the  adjacent  areas  of  the  Sea  of  Okhotsk  and  the  Yellow 
Sea.  Part  4.  Nauka.  Leningrad.  (In  Russian.) 

.  and  M.  I.  Legeza.    1955.    A  review  of  the  genera  and  species 

of  the  subfamily  Cyclopterinae  (Pisces).  Bull.  Zool.  Inst.  Acad.  Sci. 
USSR  18:389-458. 

,  AND .    1969.   Fishes  of  the  Sea  of  Japan  and  the  adjacent 

areas  of  the  Sea  of  Okhotsk  and  the  Yellow  Sea.  Part  2.  Teleostomi. 
Xll.  Acipenseriformes  — XXVlll.  Polynemiformes.  Akad.  Nauk. 
SSSR.  No.  84: 1-389.  (First  published  in  1 965.  Transl.  from  Russian 
by  Isr.  Prog.  Sci.  Transl.  cat.  no.  1852.  Available  from:  NTIS, 
Springfield,  VA;  TT68-50361.) 

LiNDSEV,  C.  C.  1975.  Pleomerism,  the  widespread  tendency  among 
related  fish  species  for  vertebral  number  to  be  correlated  with  max- 
imum body  length.  J.  Fish.  Res.  Board  Can.  32:2453-2469. 

LiPPSON,  A.  J.,  AND  R.  L.  MoRAN.  1974.  Manual  for  identification  of 
early  developmental  stages  of  fishes  of  the  Potomac  River  Estuary. 
Md.  Dep.  Nat.  Resour.  Power  Plant  Siting  Program  PPSP-MP-13: 
236-241. 

LisovENKO,  L.  A.,  andM.  F.  SvETLOV.  1981.  Some  data  on  the  biology 
of  Paraliparis  (Family  Liparidae)  off  the  South  Shetlands.  J.  Ich- 
thyol.  (Engl.  Transl.  Vopr.  Ikhtiol.)  20(1):131-135. 

,  A.  V.  Barabanov  AND  V.  N.  Efremenko.    1982.   New  data  on 

the  reproduction  of  the  "southern  putassu".  Micromesistius  aus- 
tralis  (Gadidae),  from  the  Falkland-Patagonian  zoogeographical  re- 
gion. J.  Ichthyol.  (Engl.  Transl.  Vopr.  Ikhtiol.)  22(4):55-67. 

Llewellyn,  L.  C  1973.  Spawning,  development,  and  temperature 
tolerence  of  the  spangled  perch,  Madigania  unicolor  (Gunther), 
from  inland  waters  in  Australia.  Aust.  J.  Mar.  Freshwater  Res.  24: 
73-94. 

.    1974.  Spawning,  development  and  distribution  of  the  southern 

pigmy  perch  Nannoperca  australis  australis  Gunther  from  inland 
waters  of  eastern  Australia.  Aust.  J.  Mar.  Freshwater  Res.  25:121- 
149. 

LOBEL,  P.,  AND  R.  Johannes.  1980.  Nesting,  eggs  and  larvae  of  trig- 
gerfishes  (Balistidae).  Environ.  Biol.  Fishes  5:251-252. 

Lo  Bianco,  S.  1903.  Le  pesche  abissali  esequite  da  F.  A.  Krupp  col 
yacht  (Puritan)  nelle  adiacenze  di  Capri  ed  in  altre  localita  del 
Mediterraneo.  Mitt.  Zool.  Sta.  Neapel  16:109-279. 

.    1908a.    Uova  e  larve  di  Trachipterus  taenia  Bl.  Mitt.  Zool. 

Sta.  Neapel.  19(1):  1-1 7. 

.    1908b.   Sviluppo  larvale,  metamorfos!  biologia  della  "Tnglia 

di  fango"  {.\fullus  harhatus  Lin.)  Mitt.  Zool.  Stn.  Neapel  19:18. 

.    1909.    Notize  biologiche  riguardanti  specialmente  il  periodo 

di  maturita  sessuale  delgi  animali  del  Golfo  di  Napoli.  Mitt.  Zool. 
Sta.  Neapel  19(4):5 13-763. 

.    1911.    Su  alcuni  stadi  postlarvali  appartenenti  a  Gadidi  ran 

del  golfo  di  Napoli.  Mitt.  Zool.  Sta.  Neapel  20:170-187. 

LofKET,  N.  A.    1970.    Eyes  for  the  deep.  Spectrum  78:10-15. 

Loc  KwooD.  S.  J..  J.  H.  Nkhols  and  W.  A.  Dawson.  1981.  The 
estimation  of  a  mackerel  {Scomber  scomhrus  L.)  spawning  stock 
size  by  plankton  survey.  J.  Plankton  Res.  3:217-233. 

LoEB,  V.  J.  1979.  Larval  fishes  in  the  zooplankton  community  of  the 
North  Pacific  central  gyre.  Mar.  Biol.  53:173-191. 

LoNOHiRST,  A.  R.  1971.  The  clupeoid  resources  of  tropical  seas. 
Oceanogr.  Mar.  Biol.  Annu.  Rev.  9:349-386. 

LoNNiNii,  S.  1972.  Comparative  electron  microscopic  studies  of  te- 
leostean  eggs  with  special  reference  to  the  chorion.  Sarsia  49:41- 
48. 

,  AND  B.  E.  Hagstrom.     1975.    Scanning  electron  microscope 

studies  of  the  surface  of  the  fish  egg.  Astarte  8:17-22. 

Loos,  J.  J.,  AND  L.  A.  Fliiman.  1978.  Subordinate  taxa  of  the  genus 
Nolropis:  a  preliminary  survey  of  their  developmental  traits,  p.  1- 
50.  In:  Proc.  First  Symp.  on  Freshwater  Larval  Fish.  L.  L.  Olmsted 
(ed.).  Southeastern  Electric  Exchange. 

,  ,    E.    K.   Jankowski    AND   N.    R.    Foster.      1979. 


Notes  on  early  life  histories  of  cyprinoid  fishes  of  the  upper  Potomac 
River,  p.  93-139.  In:  Proceedings  of  a  workshop  on  freshwater 
larval  fishes.  R.  Wallus  and  C.  W.  Voigtlander  (eds).  Tennessee 
Valley  Authority.  Noms,  TN. 

Lopez  RojAS,  H.,  AND  A.  Machado  Allison.  1975.  Algunos  aspectos 
del  desarrollo  y  crecimiento  de  Loricaria  laticeps  (Osteichthyes, 
Siluriformes,  Loricaridae).  Acta  Biol.  Venez.  9(I):51-76. 

Lough,  R.  G.,  M.  Pennington,  G.  R.  Bolz  and  A.  A.  Rosenberg. 
1982.  Age  and  growth  of  larval  Atlantic  herring,  Clupea  harengus 
L,,  in  the  Gulf  of  Maine-Georges  Bank  region  based  on  otolith 
growth  increments.  Fish.  Bull.  U.S.  80:187-199. 

Louw,  E.,  and  M.  J.  OTooLE.  1977.  Larval  development  ofSardinops 
OfWtoa  (Pisces:  Clupeidae).  Ann.  S.  Afr.  Mus.  72(7):125-145. 

Lowell,  N.  E.  1971.  Some  aspects  of  the  life  history  and  spawning  of 
the  moi  (Polydaclylus  saxtitis).  Unpubl.  MS  thesis.  Univ.  Hawaii, 
Honolulu. 

LowENSTAM,  H.  A.  1980.  Bioinorganic  constituents  of  hard  parts,  p. 
3-16.  In:  Biogeochemistry  of  Amino  Acids.  P.  E.  Hare  (ed).  John 
Wiley  &  Sons,  New  York. 

.    1981.  Minerals  formed  by  organisms.  Science21 1:1 126-1  131. 

,  AND  J.  E.  Fitch.    1978.    Morphology,  mineral  diversity  and 

ontogenetic  mineralogic  changes  in  extant  elasmobranch  otoconia. 
Geol.  Soc.  Am.  Abstr.  10:1  14. 

,  and .     1981.    Vatente  formation  in  otoliths  of  higher 

teleost  fishes.  Abstract.  Pacific  Division.  AAAS,  Annual  Meeting, 
June  14-19,  1981,  Univ.  of  Oregon.  Eugene. 

Llibbock,  R.  1975.  Fishes  of  the  family  Pseudochromidae  (Perci- 
formes)  in  the  northwest  Indian  Ocean  and  Red  Sea.  J.  Zool.  (Lond.) 
176(2):115-157. 

Lund,  W.  A.,  Jr.,  AND  B.  C  Mar<  I,  Jr.  1975.  Early  development  of 
the  grubby,  Myoxocephalus  aeneas  (Mitchell).  Biol.  Bull.  (Woods 
Hole)  149:373-383. 

Lundberg,  J.  G.  1972.  Wagner  networks  and  ancestors.  Syst.  Zool. 
21:398-413. 

Llither,  G.  1975a.  New  characters  for  consideration  in  the  taxonomic 
appraisal  of  grey  mullets.  Aquaculture  5:107.  (Abstract.) 

.     1975b.    On  the  little-known  mullet,    I'alamugil  buchanani 

(Bleeker)  (Mugilidae:  Pisces)  in  the  Indian  waters.  J.  Mar.  Biol. 
Assoc.  India  16(1):290-291. 

LuTKEN,  C.  1880.  Spolia  Atlantica.  Bidrag  til  kundaskab  on  forafor- 
andringer  nos  fiske  under  deres  vaext  og  udvikling.  saerlight  nus 
nogle  of  Atlantic  Lavets  honosfiske.  Kgt.  Dan.  Vidensk.  Selsk.  Biol. 
Skr.  5{12):409-613.  (In  Danish.  French  summar>.) 

Lythgoe,  J.,  AND  G.  Lythgoe.  1975.  Fishes  of  the  sea.  Anchor  Press/ 
Doubleday,  Garden  City,  New  York. 

Maccall,  A.  D.  1980.  Theconsequencesof  cannibalism  in  the  stock- 
recruitment  relationship  of  planktivorous  pelagic  fishes  such  as 
Engraulis.  IOC  Workshop  Rep.  No.  28:201-220. 

MacDonald,  C.  M.  1978.  Morphological  and  biochemical  system- 
atics  of  Australian  freshwater  and  estuarine  percichthyid  fishes. 
Aust.  J.  Mar.  Freshwater  Res.  1978:667-698. 

Macer,  C.  T.  1967.  Ammodytidae.  Cons.  Int.  Explor.  Mer,  Fiches 
Identification  Oeufs  Larves  Poissons.  2. 

MacGregor,  J.  S.  1976.  Ovarian  development  and  fecundity  of  five 
species  of  California  current  fishes.  Calif  Coop.  Oceanic  Fish.  In- 
vest. Rep.  18:181-188. 

M^^HADO  Allison,  A.,  and  H.  Lopez  Rojas.  1975.  Etapas  del  de- 
sarrollo de  Loncanichthys  lypiis  (Bleeker)  1864  (Osteichthyes,  Si- 
luriformes, Loncanidae).  Acta  Biol.  Venez.  9(1):93-1  19. 

Major.  P.  F.  1973.  Scale  feeding  behaviour  of  the  leatherjacket.  Scom- 
beroides  lysan  (Forskal)  and  two  species  of  the  genus  OligopHles 
(Pisces:  Carangidae).  Copeia  1973:151-154. 

Makeeva,  a.  p.  1 976.  Characteristics  of  the  early  de\  elopmental  stages 
of  a  bitterling  Rhodeus  ocellalus  occllalus  new  to  the  ichthyofauna 
of  the  USSR.  J.  Ichthyol.  (English  Trans.  Vopr.  Ikhtiol.)  16:756- 
767. 

,  and  I.  N.  Ryabov.    1973.    Reproduction  and  development  of 

Opsariichthys  uncirostris (Temminck  et  Schlegel)  and  its  systematic 


718 


ONTOGEl^fY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


position  in  the  family  Cyprinidae.  J.  Ichthyol.  (English  Transl.  Vopr. 
Ikhtiol.)  13:392-399. 

Maksimov,  V.  P.  1969.  Onlhe  spav/ning  of  Naiicralesducloi:  Tr.  Atl. 
Nauchno-Issled.  Inst.  Rybn.  Khoz.  Okeanogr.  25:206-209.  (In  Rus- 
sian, taken  from  Biol.  Abstr.,  1971,  abstr.  no.  42506.) 

Makushok,  V.  M.  1958.  The  morphology  and  classification  of  the 
northern  blennioid  fishes  (Stichaeoidae,  Blennioidei.  Pisces).  Tr. 
Zool.  Inst.  Akad.  Nauk  SSSR  25:3-129.  (Transl.  by  A.  R.  Gosline, 
1959,  Ichthyol.  Lab..  U.S.  Fish  Wildl.  Serv.,  U.S.  Natl.  Mus.,  Wash.. 
D.C.) 

.  1961.  The  group  Neozoarcmae  and  its  place  in  the  classifi- 
cation of  fishes  (Zoarcidae.  Blennioidei,  Pisces).  Tr.  Zool.  Inst.  Akad. 
Nauk  SSSR  43:198-224. 

.   1 966.  The  specific  identity  of  Nematonums  longifilis  (Gunther. 

1877)  and  N.  clarki  (Jordan  and  Gilbert,  1898)  and  some  remarks 
on  age-dependant  variations  in  Macruridae  (Pisces),  p.  147-172. 
In:  Fishes  of  the  Pacific  and  Indian  Oceans.  Biology  and  Distri- 
bution. T.  S.  Rass  (ed.).  Tr.  Inst.  Okeanol.  Akad.  Nauk.  SSSR  73: 
147-172. 

Malloy,  R.,  and  F.  D.  Martin.  1982.  Comparative  development  of 
redfin  pickerel  {Esox  americanus  americanus)  and  the  eastern  mud- 
minnow  (Umbra  pygmaea),  p.  70-72.  In:  Proc.  5th  Annual  Larval 
Fish  Conference.  C.  F.  Bryan,  J.  V.  Conner  and  F.  M.  Truesdale 
(eds.).  Louisiana  Cooperative  Fishery  Unit,  L.S.U.,  Baton  Rouge, 
LA. 

Manacop.  p.  R.  1936.  A  new  phallostethid  fish  with  notes  on  its 
development.  Philipp.  J.  Sci.  59:375-381. 

Mancuso,  B.  1926.  Contributo  alia  conoscenza  dello  svilupo  larvale 
di  Physiculus  dalwigki  Kp.  Mem.  R.  Comm.  Tallasogr.  Ital.  124. 

Mansueti,  a.  J.  1964.  Early  development  of  the  yellow  perch,  Perca 
flavescens.  Chesapeake  Sci.  5(I-2):46-66. 

,  AND  J.  D.  Hard>.  1967.  Development  of  fishes  of  the  Ches- 
apeake Bay  region:  An  atlas  of  egg,  larval,  and  juvenile  stages.  Part 
I.  Nat.  Resour.  Inst.  Univ.  MD  Port  City  Press.  Baltimore.  MD. 

Mansueti,  R,  J.  1958.  Eggs,  larvae,  and  young  of  the  striped  bass, 
Roccus  saxalilis.  Md.  Dep.  Res.  Educ.  CBL  Contrib.  1 12. 

.    1963.   Symbiotic  behavior  between  small  fishes  and  jellyfishes 

with  new  data  on  that  between  the  stromateid,  Pepnlus  atepidotus. 
and  the  Scyphomedusa,  Chrvsaora  qumquecirrha.  Copeia  1 963:40- 
80. 

.     1964.    Eggs,  larvae,  and  young  of  the  white  perch,  Roccus 

americanus.  with  comments  on  its  ecology  in  the  estuary.  Chesa- 
peake Sci.  5(l-2):3-45. 

Marak,  R.  R.  1967.  Eggs  and  early  larval  stages  of  the  offshore  hake, 
Merluccius  albidus.  Trans.  Am.  Fish.  Soc.  96:227-228. 

Marathe,  V.  B.,  AND  D.  V.  Bal.  1958.  A  brief  comparative  account 
of  the  axial  skeleton  of  six  polynemids  from  Bombay  waters.  J. 
Univ.  Bombay  268:139-151. 

Marchal,  E.  G.  1966.  Oeufs,  larves  et  post-larves  de  Tanchois  du 
Golfe de  Guinee,  Anchoviella guineensis(Q\ac]:\c et  Rossignol).  Doc. 
Sci.  Prov.  (Abidjan)  No.  005SR. 

.    1967.   Cle  provisoire  de  determination  des  oeufs  et  larves  des 

Clupeidaeet  Engraulidae  Ouest-africains.  Doc.  Sci.  Prov.  (Abidjan) 
No.  0I4SR. 

Marinaro.  J.-Y.  1971.  Contribution  a  I'etude  des  oeufs  et  larves 
pelagiques  de  poissons  mediterraneens.  V.  Oeufs  pelagiques  de  la 
bale  d' Alger.  Pelagos  3(l):l-l  15. 

Marino,  R.  P.,  AND  J.  K.  Dooley.  1982.  Phylogenetic  relationships 
of  the  tilefish  family  Branchiostegidae  (Perciformes)  based  on  com- 
parative myology.  J.  Zool.  (Lond.)  1 982(1 96):  15 1-1 63. 

Marion,  A.  F.  1894a.  Sur  la  peche  et  la  reproduction  du  "Sioudet" 
(Atherina  hepsetus  L.).  Ann.  Mus.  Hist.  Nat.  Marseille  (Zool.) 
4:93-99. 

.    1894b.    Oeufs  flotlants  et  alevins  observes  dans  le  Golfe  de 

Marseille  durant  I'annee  1 890.  Ann.  Mus.  Hist.  Nat.  Marseille  4(  1 1 ): 
112-121. 

Markle,  D.  F.  1982.  Identification  of  larval  and  juvenile  Canadian 
Atlantic  gadoids  with  comments  on  the  systematics  of  gadid 
subfamilies.  Can.  J.  Zool.  60(12):3420-3438. 


— .  1984.  Phosphate  buffered  formalin  for  long  term  preservation 
of  formalin  fixed  ichthyoplankton.  Copeia  1984:525-528. 

— ,  AND  L.-A.  Frost.  MS.  Comparative  morphology  and  iden- 
tification of  planktonic  fish  eggs  from  the  Nova  Scotian  Shelf 

— ,  AND  J.  E.  Olney.  1980.  A  description  of  the  vexillifer  larvae 
of  Pyramodon  venlralis  and  Snydendia  canma  (Pisces,  Carpaidae) 
with  comments  on  classification.  Pac.  Sci.  34:173-180. 

— ,  W.  B.  Scott  AND  A.  C.  KoHLER.  1980.  New  and  rare  records 
of  Canadian  fishes  and  the  influence  of  hydrography  on  resident 
and  nonresident  Scotian  Shelf  ichthyofauna.  Can.  J.  Fish.  Aquat. 
Sci.  37:49-65. 

— ,  AND  C.  A.  Wenner.  1979.  Evidence  of  demersal  spawning  in 
the  mesopelagic  zoarcid  fish  Melanostigma  allanlicum.  with  com- 
ments on  demersal  spawning  in  the  alepocephalid  fish  Xenoder- 
michlhys  copei.  Copeia  1979:363-366. 

,  J.  T.  Williams  AND  J.  E.  Olney.    1983.    Description  of  a  new 


species  of  Echiodon  (Teleostei,  Carapidae)  from  Antarctic  and  ad- 
jacent seas.  Proc.  Biol.  Soc.  Wash.  97(1). 

Marliave,  J.  B.  1975.  The  behavioral  transformation  from  the  plank- 
tonic larval  stage  of  some  marine  fishes  reared  in  the  laboratory. 
Unpubl.  PhD  dissertation.  Univ.  British  Columbia,  Vancouver. 

.     1976.    A  theory  of  storm-induced  drift  dispersal  of  the  gas- 

terosteid  fish  Aulorhynchus Jlavidus.  Copeia  1976:794-796. 

.    1978.   Spawning  and  yolk-sac  larvae  of  the  agonid  fish  species. 

Agonomalus  mozinoi  Wilimovsky  and  Wilson.  Syesis  1 1:285-286. 

.     1981.    Spawn  and  larvae  of  the  Pacific  sandfish.  Trichodon 

tnchodon.  Fish.  Bull.  U.S.  78:959-964. 

Marshall,  N.  B.  1953.  Egg  sizes  of  Arctic,  Antarctic  and  deep-sea 
fishes.  Evolution  7:328-341. 

.    1955.    Alepisauroid  fishes.  Discovery  Rep.  27:303-336. 

.    I960.   Swimbladder  structure  of  deep-sea  fishes  in  relation  to 

their  systematics  and  biology.  Discovery  Rep.  31:1-122. 

.    1961.    A  young  Macnslium  and  the  ctenothrissid  fishes.  Bull. 

Br.  Mus.  (Nat.  Hist.)  Zool.  7:355-370. 

.    1962.  Observations  on  the  Heteromi,  an  order  of  teleost  fishes. 

Bull.  Br.  Mus.  (Nat.  Hist.)  Zool.  9:249-270. 

.    1964.    Bathypelagic  macround  fishes.  Copeia  1964:86-93. 

.    1965.  Systematic  and  biological  studies  oflhemacrourid  fishes 


(Anacanthini-Teleostii).  Deep-Sea  Res.  12:299-322. 
.    1966a.    Family  Scopelosauridae.  p.  194-204. />;.■  Fishes  of  the 

western  North  Atlantic.  D.  M.  Cohen  (ed.).  Mem.  Sears  Found. 

Mar.  Res.  No.  1  (Pt.  5). 
.    1966b.    The  relationships  of  the  anacanthine  fishes.  Macru- 

ronus,  Lyconus.  and  Steiridachnena.  Copeia  1966:275-280. 
— .    1971.    Explorations  in  the  life  of  fishes.  Harvard  Univ.  Press, 

Cambridge.  MA. 

— .    1973.   Family  Macouridae.  p.  496-665. /«.  Fishes  of  the  west- 
em  North  Atlantic.  D.  M.  Cohen  (ed.).  Mem.  Sears  Found.  Mar. 

Res.  No.  1  (Pt.  6). 
— ,  AND  D.  M.  Cohen. 


1973.    Order  Anacanthini  (Gadiformes). 

Characters  and  synopsis  of  families,  p.  479-495.  In:  Fishes  of  the 

western  North  Atlantic.  D.  M.  Cohen  (ed.).  Mem.  Sears  Found. 

Mar.  Res.  No.  I  (Pt.  6). 
,  AND  T.  IwAMOTO.    1973.    Family  Macrouridae  (Genera  Coe- 

lorhynchus.  Coryphaenoides,  Hymenocephahis  and  Nczumia).  p. 

496-665.  In:  Fishes  of  the  western  North  Altantic.  D.  M.  Cohen 

(ed.).  Mem.  Sears  Found.  Mar.  Res.  No.  I  (Pt.  6). 
,  and  J.  C.  Staiger.    1975.    Aspects  of  the  structure,  relation- 
ships, and  biology  of  the  deep-sea  fish  Ipnops  murrayi  (Family 

Bathypteroidae).  Bull.  Mar.  Sci.  25:101-111. 
Marshall,  T.  C.    1965.    Fishes  of  the  Great  Barrier  Reef  and  coastal 

waters  of  Queensland.  Livingston  Publ.  Co..  Narberth,  PA. 
Martin,  F.  D.,  AND  G.  E.  Drewry.    1978.    Development  of  the  fishes 

of  the  Mid-Atlantic  Bight.  An  atlas  of  egg,  larval  and  juvenile  stages. 

Vol.  VI.  Stromateidae  through  Ogcocephalidae.  U.S.  Fish  Wildl. 

Serv.  Biol.  Serv.  Prog.  FWS/OBS-78/12. 
Martin,  R.  A.,  .AND  J.  H.  FiNHCANE.    1969.   Reproduction  and  ecology 

ofthelongnose  killifish.  Q.  J.  Fla.  Acad.  Sci.  31(2):101-l  1 1. 
Martinez,  A.  M.    1983.    Identification  of  brook,  brown,  rainbow,  and 


LITERATURE  CITED 


719 


cutthroat  trout  larvae.  Unpubl.  MS  Thesis.  Colorado  State  Univ., 

Ft.  Collins,  CO. 
Maruyama,  K.    1970.    Some  deep-water  fishes  from  off  the  Tohoku 

and  adjacent  regions.  Bull.  Tohoku  Reg.  Fish.  Res.  Lab.  30:43-66. 

(In  Japanese.) 
Marx,  H.,  AND  D.  B.  Rabb.    1972.   Phyletic  analysis  of  fifty  characters 

of  advanced  snakes.  Fieldiana  Zool.  63:1-321. 
Masterman,  a.  T.    1901.    A  contribution  to  the  life  histories  of  the 

cod  and  whiting.  Trans.  R.  Soc.  Edinb.  40:1-14. 
Masuda,  H.,  C.  Araga  and  T.  Yoshino.     1975.    Coastal  fishes  of 

Southern  Japan.  Tokai  Univ.  Press,  Tokyo. 
Masuda,  S.,  and  T.  Ozawa.    1979.    Reexamination  of  the  holotypes 

of  Bregmaceros  japonicus  Tanaka  and  B.  neclabamis  Whitley.  Jpn. 

J.  Ichthyol.  25:266-268. 
Masuda,  T.,  and  K.  Tanaka.    1962.    Young  of  labroid  and  scaroid 

fishes  from  the  central  Pacific  of  Japan.  J.  Tokyo  Univ.  Fish.  48: 

1-98. 
Masurekar,  v.  B.    1968.    Eggs  and  development  stages  of  Tylosurus 

crocodilus  (LesueuT).  J.  Mar.  Biol.  Assoc.  India.  9:70-76. 
Matarese,  A.  C.  AND  S.  F.  Borton.    MS.    Reproductive  biology  and 

development  of  the  Pacific  spiny  lumpsucker,  Eumicrolremus  orbis 

(Cyclopteridae). 
,  AND  J.  B.  Marliave.    1982.   Larval  development  of  laboratory- 
reared  rosylip  sculpin,  Ascelichthvs  rhodonis  (Cottidae).  Fish.  Bull. 

U.S.  80:345-355. 
,  S.  L.  Richardson  AND  J.  R.  Dunn.    1981.   Larval  development 

of  Pacific  tomcod,  Microgadus  pro.ximiis.  in  the  northeast  Pacific 

Ocean  with  comparative  notes  on  larvae  of  walleye  pollock,  Ther- 

agra  chakogramma.  and  Pacific  cod,  Gadus  macrocephalus  (Gad- 

idae).  Fish.  Bull.  U.S.  78:923-940. 
,  AND  B.  Vinter.    MS.    Development  and  occurrence  of  larvae 

of  the  longfin  Irish  lord,  Hemilepidotus  zapus  (Cottidae). 
Mather,  F.  J.    1958.    A  preliminary  review  of  the  amberjacks,  genus 

Senota.  of  the  western  Atlantic.  Proc.  3rd  Int.  Game  Fish  Conf 
Matias,  J.  R.    1982.   Embryonic  diapause  in  annual  fishes:  Evaporative 

water  loss  and  survival.  Experientia  (Basel)  38:1315-1316. 
Matsubara,  K.    1936.  A  review  oftwo  genera  of  the  Japanese  sculpins, 

Ereiinias  and  Marukawichthys  with  special  reference  to  their  sys- 
tematic positions.  J.  Imp.  Fish.  Inst.  (Tokyo)  31:97-1  14. 
.    1942.    On  the  metamorphosis  of  a  clupeoid  fish  Plerothnssus 

gissu  Hilgendorf  J.  Imp.  Fish.  Inst.  (Tokyo)  35(I):1-16. 
.    1943.    Studies  on  the  scorpaenoid  fishes  of  Japan.  Trans.  Sig- 

enkagaku  Kenyusho.  Nos.  1  and  2. 
.    1955.    Fish  morphology  and  hierarchy.  Part  II.  Ishizaki-Sho- 

ten,  Tokyo. 
,  AND  T.  IwAi.     1952.    Studies  on  some  Japanese  fishes  of  the 

family  Gempylidae.  Pac.  Sci.  6:193-212. 
,  AND .    1958.  Anatomy  and  relationships  of  the  Japanese 


fishes  of  the  family  Gempylidae.  Mem.  Coll.  Agnc.  Kyoto  Uni\ 
Fish  Ser.  Spec.  No.,  p.  23-54. 

,  AND .    1959.   Description  of  a  new  sandfish,  Kraemana 

sexradiata.  from  Japan,  with  special  reference  to  its  osteology.  J. 
Wash.  Acad.  Sci.  49:27-32. 

,  AND  A.  OcHiAi.    1955.   A  revision  of  the  Japanese  fishes  of  the 

family  Platycephalidae  (the  flatheads).  Mem.  Coll.  Agric.  Kyoto 
Univ.  68:1-109. 

,  ,  K.  Am,aok.a  and  I.  Nak,amura.     1964.    Revisional 

study  of  the  trachinoid  fishes  of  the  family  Champsodontidae  from 
the  waters  around  Japan,  and  Tonking  Bay.  Bull.  Misaki  Mar.  Biol. 
Inst.  Kyoto  Univ.  6:1-20. 

Matsui,  K.  1949.  Illustration  of  the  normal  courses  of  development 
in  the  fish,  Oryzias  latipes.  Jikken-keitai  Gaku  5:33-42.  (In  Jap- 
anese.) 

Matsumoto,  W.  M.  1968.  Morphology  and  distnbution  of  larval 
wahoo  Acanthocybium  solandn  (Cuvier)  in  the  central  Pacific  Ocean 
fishery.  U.S.  Fish  Wildl.  Serv.  Fish.  Bull.  66:299-322. 

,  E.  H.  Ahlstrom,  S.  Jones,  W.  L.  Klawe,  W.  J.  Rr  hards  and 

S.  Ueyanagi.  1972.  On  the  clarification  of  larval  tuna  identifi- 
cation particularly  in  the  genus  Thunnus.  Fish.  Bull.  U.S.  70:1-12. 


Matsuoka,  M.,  and  T.  Iwai.    1983.    Adipose  fin  cartilage  found  in 

some  teleostean  fishes.  Jpn.  J.  Ichthyol.  30:37-46. 
Matsuura,  K.   1979.  Phytogeny  ofthesuperfamilyBalistoidea (Pisces: 

Tetraodontiformes).  Mem.  Fac.  Fish.  Hokkaido  Univ.  26:49-169. 
.    1982.  A  list  ofthe  fishes  collected  in  the  Palau  and  Yap  Islands. 

Proc.  Jpn.  Soc.  Syst.  Zool.  23:80-89. 
Matsuura,  Y.    1975.   A  study  of  the  life  history  of  Brazilian  sardine, 

Sardinetla  brasiliensis.  III.  Development  of  sardine  larvae.  Bol. 

Inst.  Oceanogr.  23:17-29. 
,  AND  M.  Katsur.aoawa.    1981.    Larvae  and  juveniles  of  grey 

triggerfish,  Balisies  capriscus.  from  Southern  Brazil.  Jpn.  J.  Ich- 
thyol. 28:267-275. 
Mattson.C.  R.,  AND  B.  L.  Wing.    1978.   Ichthyoplankton  composition 

and  planktonic  volumes  from  inland  coastal  waters  of  southeastern 

Alaska,  April-November  1972.  NOAA  Tech.  Rep.  NMFS  SSRF 

723. 
Maul,  G.  E.    1951.    Monografia  dos  Peixes  do  Museu  Municipal  do 

Funchal,  Familia  Macrouridae  e  Merlucciidae.  Bol.  Mus.  Munic. 

Funchal  No.  5  (Art.  12):5-55. 
.   1976.  The  fishes  taken  in  bottom  trawls  by  R.V.  Meteor  during 

the  1967  Seamount  Cruises  in  the  Northeast  Atlantic.  Meteor  For- 

schungsergeb.  22:1-69. 
,  AND  E.  KoEFOED.    1950.   On  a  new  genus  and  species  of  mac- 

rourid  fish,  Phalacromacrurus  panthennus.  Ann.  Mag.  Nat.  Hist., 

Ser.  12,  3:970-976. 
May,  R.  C.     1970.    Feeding  larval  marine  fishes  in  the  laboratory:  a 

review.  Calif  Coop.  Oceanic  Fish.  Invest.  Rep.  14:76-83. 
.    1974.   Larval  mortality  in  marine  fishes  and  the  critical  period 

concept,  p.  3-19.  In:  The  eariy  life  history  offish.  J.  H.  S.  Blaxter 

(ed.).  Springer  Verlag,  Berlin. 
,  D.  Popper  and  J.  P.  McVey.    1974.    Rearing  and  larval  de- 
velopment oi  Siganus  canaliculatus  (Park)  (Pisces:  Siganidae).  Mi- 

cronesica  10:285-298. 
Mayer,  F.    1933.    Cuhanichthys  cubensis  Eigenmann.  Blatt.  Aquar.- 

Terranenk.  44(22):375-376. 
Mayer,  G.  F.    1972.   Systematics,  functional  anatomy,  and  ecology  of 

the  cardinalfish  genus  Epigonus  (Apogonidae).  Unpubl.  PhD  dis- 
sertation. Harvard  Univ.,  Cambridge,  MA. 
Mayo,  C.  A.    1973.   Rearing,  growth,  and  development  of  the  eggs  and 

larvae  of  seven  scombrid  fishes  from  the  straits  of  Florida.  Unpubl. 

PhD  thesis.  University  of  Miami,  Coral  Gables,  FL. 
Mair,  E.    1969.   Principles  of  systematic  zoology.  McGraw-Hill  Publ., 

New  York. 
.    1981.   Biological  classification:  toward  a  synthesis  of  opposing 

methodologies.  Science  214:510-516. 
Mazza,  F.    1902.    La  Lebias  calarilana  Bon.  Segmenlazione  dell'uovo 

e  stadi  successivi  di  sviluppo.  Atti  Soc.  Ligust.  Sci.  Nat.  Genova 

13:181-223. 
M.AZZARELLi,  G.    1917.    Contributo  alia  conosccnza  dcllc  mctamorfosi 

deW Eretmophonis  kleinenbergi  Gigl.  Pubbl.  Staz.  Zool.  Napoli  2. 
McAllister,  D.  E.    1963.   A  revision  of  the  smelt  family,  Osmeridae. 

Natl.  Mus.  Can.  Bull.  191. 
.    1968.    Evolution  of  branchiostegals  and  classification  of  tel- 

eostome  fishes.  Natl.  Mus.  Canada  Bull.  221. 
,  AND  J.  E.  Randall.    1975.   A  new  species  of  centrolophid  fish 

from  Easter  Island  and  Rapa  Iti  Island  in  the  South  Pacific.  Natl. 

Mus.  Nat.  Sci.  (Ottawa)  Publ.  Biol.  Oceanogr.  8. 
McCann,  C,  AND  D.  G.  M(  Knight.    1980.   The  marine  fauna  of  New 

Zealand:  macrourid  fishes  (Pisces:  Gadida).  N.Z.  Oceanogr.  Inst. 

Mem.  61:1-91. 
McCosKER,  J.  E.    1977.   The  osteology,  classification  and  relationships 

ofthe  eel  family  Ophichthidae.  Proc.  Calif.  Acad.  Sci.  4 1(  1 ):  1-1 23. 
.    1979.   The  snake  eels  (Pisces.  Ophichthidae)  of  the  Hawaiian 

Islands,  with  the  descnption  of  two  new  species.  Proc.  Calif.  Acad. 

Sci.  42(2):57-67. 

1982.   A  new  genus  and  two  new  species  of  remarkable  Pacific 


worm  eels  (Ophichthidae,  subfamily  Myrophinae).  Proc.  Calif  Acad. 
Sci.  43:59-66. 
— ,  AND  R.  H.  Rosenblatt.    1972.    Eastern  Pacific  snake-eels  of 


720 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


the  genus  Callechelys  (Apodes:  Ophichthidae).  Trans.  San  Diego 
Soc.  Nat.  Hist.  17:15-24. 

McCuLLOCH,  A.  R.  1926.  Report  on  some  fishes  obtained  by  the  F.I. S. 
"Endeavor"  on  the  coasts  of  Queensland.  New  South  Wales.  Vic- 
toria. Tasmania,  South  and  south-western  Australia.  Part  V,  p.  157- 
216.  Commonw.  Aust.  Dep.  of  Trade  and  Customs. 

McCuLLV,  H.  1961.  The  comparative  anatomy  of  the  scales  of  the 
serranid  fishes.  Unpubl.  PhD  dissertation.  Stanford  Univ.,  Stan- 
ford. CA. 

.     1970.    Amputation  and  replacement  of  marginal  spines  in 

ctenoid  percoid  scales.  Proc.  Calif  Acad.  Sci.  38:41 1-414. 

McDowALL,  R.  M.  1968a.  The  ?>\a\\x^  of  Nesogalaxias  neocaledonicus 
(Weber  and  de  Beaufort)  (Pisces:  Gala.xiidae).  Breviora  286:1-8. 

.  1 968b.  Gataxias  maculatus  (Jenyns).  the  New  Zealand  white- 
bait. Fish.  Bull.  N.Z.  2:1-84. 

.    1 969.   Relationships  of  galaxioid  fishes  with  further  discussion 

of  salmoniform  classification.  Copeia  1969:796-824. 

.    1970.    The  galaxiid  fishes  of  New  Zealand.  Bull.  Mus.  Comp. 

Zool.  Harv.  Univ.  139:341-431. 

.    1971a.    Fishes  of  the  family  Aplochitonidae.  J.  R.  Soc.  N.Z. 

1:37-52. 

.    1971b.    The  galaxiid  fishes  of  South  America.  Zool.  J.  Linn. 

Soc.  50:33-73. 

.    1973a.   The  status  of  the  South  African  galaxiid  (Pisces:  Gal- 

axiidae).  Ann.  Cape  Prov.  Mus.  Nat.  Hist.  9(5):91-101. 

.    1973b.    Limitation  of  the  genus  Sraf/!i'ga/aA7a.s  Eigenmann, 

1928  (Pisces:  Galaxiidae).  J,  R.  Soc.  N.Z.  3:193-197. 

.    1973c.    Relationships  and  taxonomy  of  the  New  Zealand  tor- 


rent fish,  Chehnarrichlhys  foslen  Haast  (Pisces:  Mugiloididae).  J. 

R.  Soc.  N.Z.  3:199-217. 
— .    1976.    Fishes  of  the  family  Prototroctidae  (Salmoniformes). 

Aust.  J.  Mar.  Freshwater  Res.  27:641-659. 
— .     1978.    New  Zealand  freshwater  fishes— a  guide  and  natural 

history.  Heinemann  Educational  Books,  Auckland. 
— .    1979.    Fishes  of  the  family  Retropinnidae  (Pisces:  Salmoni- 
formes)—a  taxonomic  revision  and  synopsis.  J.  Soc.  N.Z.  9:85- 

121. 
— .    1980.    Freshwater  fishes  and  plate  tectonics  in  the  southwest 

Pacific.  Palaeogeogr.  Palaeoclimatol.  Palaeoecol.  31:337-351. 
— .    1982.    The  centrolophid  fishes  of  New  Zealand  (Pisces:  Stro- 

mateoidei).  J.  R.  Soc.  N.Z.  12:103-142. 
— ,  AND  G.  A.  Eldon.    1 980.   The  ecology  of  whitebait  migrations. 

Fish.  Bull.  N.Z.  20:1-171. 
— ,  andR.  S.  Frankenberg.    1981.  The  galaxiid  fishes  of  Australia 

(Pisces:  Galaxiidae).  Rec.  Aust.  Mus.  33:433-605. 
-,  D.  A.  Robertson  AND  R.  Saito.    1975.   Occurrence  of  galaxiid 


larvae  and  juveniles  in  the  sea.  N.Z.J.  Mar.  Freshwater  Res.  9:1-9. 
McDowell,  S.  B.    1973.    Order  Heteromi  (Notacanthiformes),  p.  1- 

228.  In:  Fishes  of  the  western  North  Atlantic.  D.  M.  Cohen  (ed.). 

Mem.  Sears  Found.  Mar.  Res.  No.  1  (Pt.  6). 
McGiNNis,  R.  F.    1982.    Biogeography  of  lantemfishes  (Myctophidae) 

south  of  SO-S.  Biol.  Antarct.  Seas  12,  Antarct.  Res.  Ser.  35:1-1 10. 
McIntosh,  W.  C.    1892.    Contributions  to  the  life-histones  and  de- 
velopment of  the  food-  and  other  fishes.   10th  Annu.  Rep.  Fish. 

Board  Scotl.,  Pt.  3:273-322. 
.    1893.    Contributions  to  the  life-histones  and  development  of 

the  food-  and  other  fishes.  Scotl.  Fish.  Board  Annu.  Rep.  1 1(1892): 

239-249. 
,  AND  A.  T.  Mastermann.    1897.  The  life-histories  of  the  British 

marine  food  fishes.  Cambridge  Univ.  Press,  London. 
,  AND  E.  E.  Prince.    1890.  On  the  development  and  life-histories 

of  the  teleostean  food-  and  other  fishes.  Trans.  R.  Soc.  Edinb.  35: 

665-946. 
McKenney,  T.  W.     1959.    A  contribution  to  the  life  history  of  the 

squirrel  fish  Holocenlrus  vexillarius  Poey.  Bull.  Mar.  Sci.  Gulf  Car- 

ibb.  9:174-221. 
.    1961.   Larval  and  adult  stages  of  the  stromateoid  fish  Pie/?« 

regulus  with  comments  on  its  classification.  Bull.  Mar.  Sci.  Gulf 

Caribb.  11:210-236. 


,  E.  C.  Alexander  AND  G.  S.  Voss.    1958.    Early  development 

and  larval  distribution  of  the  carangid  fish,  Caranx  crysos  (Mitchill). 

Bull.  Mar.  Sci.  Gulf  Caribb.  8:167-200. 
Ml  Kenzie,  R.  A.,  AND  S.  N.  TiBBO.    1963.   An  occurrence  of  the  opah, 

Lampris  reg;i«(Bonnaterre)  in  the  northwest  Atlantic.  J.  Fish.  Res. 

Board  Can.  20:1097-1099. 
M<  Kern,  J.  L.,  H.  F.  Morton  AND  K.  V.  KosKi.    1974.    Development 

of  steelhead  trout  (Salmo  gairdneri)  otoliths  and  their  use  for  age 

analysis  and  for  separating  summer  from  winter  races  from  hatchery 

stocks.  J.  Fish.  Res.  Board  Can.  31:1420-1426. 
M(  Lellan,  T.    1977.    Feeding  strategies  of  the  macrourids.  Deep-Sea 

Res.  24:1019-1036. 
Mc  Mahon,  T.  E.,  AND  J.  C.  Tash.    1979.    Effects  of  formalin  (buffered 

and  unbuffered)  and  hydrocholoric  acid  on  fish  otoliths.  Copeia 

1979:155-156. 
McMillan,  C.  W.    1979.    The  embryological  development  of  .Rn'M/wi 

marmoratus  (Pisces:  Cyprinodontidae).  Unpubl.  MS  thesis.  College 

of  Charleston. 
McMillan,  H.  M.    1961.    An  addition  to  the  knowledge  of  the  fish 

Retropinna  anisodon  Stokell  (Retropinnidae).  Trans.  R.  Soc.  N.Z. 

Zool.  1(10):139-144. 
Mead,  G.  W.     1957.    An  Atlantic  record  of  the  zeoid  fish  Parazen 

pacificus.  Copeia  1957:235-237. 
.    1963.    Observations  on  fishes  caught  over  the  anoxic  waters 

of  the  Cariaco  Trench,  Venezuela.  Deep-Sea  Res.  10:251-257. 

.    1972.    Bramidae.  Dana-Rep.  Carlsberg  Found.  81. 

,  E.  Bertelsen  AND  D.  M.  Cohen.    1964.    Reproduction  among 


deep  sea  fishes.  Deep-Sea  Res.  1 1:569-596. 

Meek,  S.  E.,  AND  S.  F.  Hildebrand.  1923.  The  marine  fishes  of  Pan- 
ama. Field  Mus.  Nat.  Hist.  Publ.  Zool.  Ser.  15(215). 

Meldrim,  J.  W.  1980.  Novumbra  hubbsi.  p.  128.  In:  Atlas  of  North 
American  freshwater  fishes.  D.  S.  Lee,  C.  R.  Gilbert,  C.  H.  Hocutt, 
R.  E.  Jenkins,  D.  E.  McAllister  and  J.  R.  Staufler,  Jr.  (eds).  North 
Carolina  State  Mus.  Nat.  Hist.,  Raleigh,  NC. 

Menasse,  V,  1970.  The  spined  loach,  Cobilis  taenia.  Trop.  Fish  Hob- 
by. 1 8(6):36-37,  40-41,44. 

Menon,  A.  G.  K.  1977.  A  systematic  monograph  of  the  tongue  soles 
of  the  genus  Cynoglossus  Hamilton-Buchanan  (Pisces:  Cynoglos- 
sidae).  Smithson.  Contrib.  Zool.  238. 

Mercado  S.,  J.  E.,  and  A.  Ciardelli.  1972.  Contribucion  a  la  mor- 
fologia  y  organogenesis  de  los  leptocefalos  del  sabalo  Megalops 
allanlicus  (Pisces:  Megalopidae).  Bull.  Mar.  Sci.  22:153-184. 

Merrett,  N.  R.  1963.  Pelagic  gadoid  fish  in  the  Antarctic.  Nor.  Hval- 
fangst-tidsskr.  52:245-247. 

.     1978.    On  the  identity  and  pelagic  occurrence  of  larval  and 

juvenile  stages  of  rattail  fishes  (family  Macroundae)  from  60°N, 
20°W  and  53°N,  20°W.  Deep-Sea  Res.  25:147-160. 

.  1980.  Balhytyphlops  sewe/i  {Pisces:  Chlorophthalmidae)  a  se- 
nior synonym  of  B.  azorensis,  from  the  eastern  North  Atlantic  with 
notes  on  its  biology.  Zool.  J.  Linn.  Soc.  68:99-109. 

,  J.  Badcock  and  p.  J.  Herring.    1973.    The  status  of  Ben- 

ihatbetla  infans  (Pisces.  Myctophoidei),  its  development,  biolu- 
minescence,  general  biology  and  distribution  in  the  eastern  North 
Atlantic.  J.  Zool.  Lond.  170:1-48. 

,  AND  N.  B.  Marshall.    1981.    Observations  on  the  ecology  of 

deep-sea  bottom-living  fishes  collected  off  northwest  Africa  (08°- 
27°N).  Prog.  Oceanogr.  9:185-244. 
-,  Y.  I.  Sazonov  and  Y.  N.  Shcherbachev.    1983.   A  new  genus 


and  species  of  rattail  fish  (Macrouridae)  from  the  eastern  North 

Atlantic  and  eastern  Indian  Ocean,  with  notes  on  its  ecology.  J. 

Fish  Biol.  22:549-561. 
Meshkov,  M.  M.    1967.    The  phases  in  the  development  of  Lola  lata 

(Linne).  Izv.  Vses.  Nauchno-Issled.  Inst.  Ozem.  Rechn.  Rybn.  Khoz. 

62:181-194. 
Methot,  R.  D.    1981.   Spatialcovariationofdailygrowlh  rates  of  larval 

northern   anchovy,   Engraulis   mordax.   and   northern   lampfish, 

Stenobrachnis  leucopsarus.  Rapp.  P.  V.  Reun.  Cons.  Int.  Explor. 

Mer  178:424-431. 


LITERATURE  CITED 


721 


,  AND  D.  Kramer.     1979.    Growth  of  nonhem  anchovy,  En- 

graulis  mordax.  larvae  in  the  sea.  Fish.  Bull.  U.  S.  77:1  124-1 127. 
Mettee,  M.  F.,  Jr..  and  E.  C.  Beckman.  HI.     1978.    Notes  on  the 

breeding  behavior,  embryology  and  larval  development  of  Cyprin- 

odon  vanegatus  Lacepede  in  aquana.  Tulane  Stud.  Zool.  Bot.  20(3- 

4):137-I48. 
Mever-Rochow.  V.  B.    1974.   Leptocephali  and  other  transparent  fish 

larvae  from  the  south-eastern  Atlantic  Ocean.  Zool.  Anz.  192(3/ 

4):240-25l. 
MicHENER,  C.     1977.    Discordant  evolution  and  the  classification  of 

allodapine  bees.  Syst.  Zool.  26:32-56. 
Mi<  KEVK  H,  M.  F.    1982.    Transformation  series  analysis.  Syst.  Zool. 

31:461-478. 
Middauoh,  D.  P.,  G.  I.  Scott  AND  J.  M.  Dean.    1981.    Reproductive 

behavior  of  the  Atlantic  silverside,  Menidia  mcnidia  (Pisces,  Ath- 

erinidac).  Environ.  Biol.  Fishes  6:269-276. 
Miller,  D.    1958.    A  key  to  some  of  the  more  common  larval  fishes 

of  the  Gulf  of  Maine.  NMFS  Woods  Hole  Lab.  Misc.  Rep.  58- 1 . 
,  AND  R.  R.  Marak.    1959.    The  early  larval  stages  of  the  red 

hake,  Urophycis  chuss.  Copeia  1959:248-250. 
.  AND  R.  R.  Marak.    1962.    Early  larval  stages  of  the  fourspot 

flounder.  Paralichthys  oblongus.  Copeia  1962:454-455. 
Miller,  D.  J.    1952.    Development  through  the  prelarval  stage  of  ar- 
tificially fertilized  eggs  of  the  Pacific  sardine  (Sardinops  caeiidca). 

Calif.  Fish  Game  38:587-595, 
,  AND  R.  N.  Lea.    1972.    Guide  to  the  coastal  marine  fishes  of 

California.  Calif  Dep.  Fish  Game  Fish  Bull.  157. 
Miller,  G.  L.,  and  S.  C.  Jorgensen.     1973.    Meristic  characters  of 

some  marine  fishes  in  the  western  Atlantic  Ocean.  Fish.  Bull.  U.S. 

71:301-312. 
Miller.  J.  M.,  and  B.  Y.  SuMiDA.    1974.   Development  and  life  history 

of  Caranx  mate  (Carangidae).  Fish.  Bull.  U.S.  72:497-514. 
,  AND  J.  W.  Tucker.  Jr.     1979.    X-radiography  of  larval  and 

juvenile  fishes.  Copeia  1797:535-538. 
.  W.  Watson  and  J.  M.  Leis.    1979.    An  atlas  of  nearshore 


marine  fish  larvae  of  the  Hawaiian  Islands.  Sea  Grant  Misc.  Rep. 

UNlHl-SEAGRANT-MR-80-02.  Univ.  Hawaii  Sea  Grant  College 

Program.  Honolulu. 
Miller.  P.  J.    1961.    Evidence  for  the  breeding  in  Manx  waters  of  the 

sand-smelt.  Athenna  preshvler  C  &  V.  Rep.  Mar.  Biol.  Sta.  Point 

Erin  74:27-28. 
.    1973.    The  osteology  and  adaptive  features  of  Rhyacichlhys 

aspro  (Teleostei:  Gobioidei)  and  the  classification  of  gobioid  fishes. 

J.  Zool.  (Lond.)  171:397-434. 
Miller,  R.  R.     I960.    Systematics  and  biology  of  the  gizzard  shad 

(Dorosoma  cepedianum)  and  related  fishes.  U.S.  Fish  Wildl.  Serv. 

Fish.  Bull.  60:371-392. 
,  AND  J.  C.  Briggs.    1962.    Daclylos^opus  ainnis,  a  new  sand 

stargazer  from  rivers  of  the  Pacific  slope  of  southern  Mexico.  Occas. 

Pap.  Univ.  Mich.  Mus.  Zool.  627. 
Miller,  R.  V.     1969.    Constancy  of  epibranchial  organs  and  fourth 

epibranchial  bones  within  species  groups  of  clupeid  fishes.  Copeia 

1969:308-312. 
,  AND  J.  W.  VAN  Landingham.     1969.    Additional  procedures 

for  effective  enzyme  clearing  and  staining  of  fishes.  Copeia  1969: 

829-830. 
MiLLiKEN,  D.  M.    1975.   Species  of  the  genus  Bregmaceros  in  the  Gulf 

of  Mexico  and  Caribbean  Sea,  including  the  description  of  a  new 

species  from  the  Cariaco  Trench.  Unpubl.  MS  thesis.  Univ.  South 

Florida,  Tampa,  FL. 
,  and  E.  D.  Holde.    1984.    A  new  species  of  Bregmacerotidae 

(Pisces),  Bregmaceros  canton,  from  the  western  Atlantic  Ocean. 

Bull.  Mar.  Sci.  In  press. 
MiiwARD,  N.  H.    1966.    Development  of  eggs  and  early  larvae  of  the 

Australian  smelt  Rctropinna  semoni  (Weber).  Proc.   Linn.  Soc. 

N.S.W.  90:218-227. 
MiNAMi,  T.    1981a.   The  early  life  history  ofa  flounder  I'seudorbombus 

pentophlhalmus.  Bull.  Jap.  Soc.  Sci.  Fish.  47:849-856. 
.    1981b.    The  early  life  history  ofa  sole  Heteromycteris  japon- 


icus.  Bull.  Jap.  Soc.  Sci.  Fish.  47:857-862.  (In  Japanese.  English 
abstract.) 

.     1981c.    The  early  life  history  ofa  flounder  Limanda  yoko- 

hamae.  Bull.  Jpn.  Soc.  Sci.  Fish.  47:141  1-1419. 

.     1982a.    The  early  life  history  ofa  flounder  Pleuronkhthys 

cornutus.  Bull.  Jap.  Soc.  Sci.  Fish.  48:369-374.  (In  Japanese.  En- 
glish abstract.) 

.    1 982b.    The  early  life  history  of  a  tongue  fish  Paraplagusia 

japonica.  Bull.  Jap.  Soc.  Sci.  Fish.  48:1041-1046. 

MiNOPRio.  J.  L.  1944.  Considerations  sobre  el  desarrollo  embrionano 
del  Perjerrey.  Acta.  Zool.  Lilloana  2:9-14. 

MisiTANO,  D.  1978.  Description  of  laboratory-reared  larvae  of  the 
Buffalo  Sculpin.  Enophrvs  bison  (Pisces,  Cottidae).  Copeia  1978: 
635-642. 

.  1980.  A  record  of  internal  fertilization  in  the  roughback  scul- 
pin Chitonotus  pugetensis  with  descriptive  notes  of  their  early  lar- 
vae. Copeia  1980:162-165. 

MiTANi,  F.  1960.  Fishery  biology  of  the  yellowtail,  Seriola  qumque- 
radiata  T.  &  S.,  inhabiting  the  water  around  Japan.  Mem.  Fac. 
Agric.  Kinki  Univ.  1:81-300. 

Mitchell,  C  P..  AND  B.  P.  Penlington.  1982.  Spawning  of  Ga/a.v(aj 
fascialus  Gray  (Salmoniformes:  Galaxiidae).  N.Z.  J.  Mar.  Fresh- 
water Res.  16:131-133. 

,  AND  D.  ScoTT.  1979.  Muscle  myogens  in  New  Zealand  Gal- 
axiidae. N.Z.  J.  Mar.  Freshwater  Res.  13:285-294. 

MiTO,  S.  1954.  Breeding  habits  ofa  blennioid  fish,  Salarias  enosimae. 
Jpn.  J.  Ichthyol.  3:144-152,  (In  Japanese.  English  summary.) 

.    1955.   Breeding  habits  of  a  percoid  fish.  P/M/op5Sf»;e(o/;.  Sci. 

Bull.  Fac.  Agric.  Kyushu  Univ.  15(l):95-99. 

.    1956a.    On  the  egg  development  and  hatched  larvae  of  Lelh- 

rinus  nematacanthus  Bleeker.  Sci.  Bull.  Fac.  Agric.  Kyushu  Univ. 
15:497-500. 

.    1956b.  On  the  egg  development  and  hatched  larvae  of  Op/e^- 

nathus  fascialus  (Temmni:}/.  et  Schlegel).  Sci.  Bull.  Fac.  Agric.  Kyu- 
shu LJniv.  15:501-506. 

.    1957a.    Egg  development  and  hatched  larvae  of  G/>W/a  pi/nc- 


tata  Gray  (Girellidae,  Teleostei).  Jpn.  J.  Ichthyol.  6:105-108.  (In 

Japanese,  English  summary.) 
.    1957b.    On  the  egg  development  and  larvae  of  a  Japanese  sea 

bass,  Lateolabrax japonicus  (Cuvier).  Sci.  Bull.  Fac.  Agric.  Kyushu 

Univ.  16:1 15-124.  (In  Japanese.  English  summary.) 
.     1958.    Studies  on  the  eggs,  larvae  and  juvenile  of  Japanese 

fishes.  Series  I.  In:  K.  Uchida  et  al.  1958.  Fish  Dep..  Fac.  Agric. 

Kyushu  Univ..  Fukuoka.  Japan.  (In  Japanese.) 
.    1960.    Egg  development  and  hatched  lar\ae  of  the  common 

dolphin-fish.  Coryphaena  hippunis  Linne.  Bull.  Jap.  Soc.  Sci.  Fish. 

26(3):223-226.  (In  Japanese.  English  abstract.) 
.    1961a.    Pelagic  fish  eggs  from  Japanese  waters— I.  Clupeina. 

Chanina,  Stomiatina,  Myctophida,  Anguillida,  Belonida  and  Syng- 

nathida.  Sci.  Bull.  Fac.  Agric.  Kyushu  Univ.  18:285-310. 
.    1961b.   Pelagic  eggs  from  Japanese  waters— II.  Lamprida,  Zei- 

da,  Mugilina,  Scombrina,  Carangina  and  Stromateina.  Sci.  Bull. 

Fac.  Agric.  Kyushu  Univ.  18:451-465. 
— .    1962a.    Pelagic  fish  eggs  from  Japanese  waters— V.  Calliony- 

mina  and  Ophidiina.  Sci.  Bull.  Fac.  Agric.  Kyushu  Univ.  19:377- 

380. 
.    1962b.    Pelagic  fish  eggs  from  Japanese  waters— VI.  Labrina. 

Sci.  Bull.  Fac.  Agric.  Kyushu  Univ.  19:493-502.  (In  Japanese.  En- 
glish summary.) 
— .    1962c.    Pelagic  fish  eggs  from  Japanese  waters— VII.  Chaeto- 


donlina.  Balistina  and  Ostraciontina.  Sci.  Bull.  Fac.  Agric.  Kyushu 

Univ.  19:503-506.  (In  Japanese,  English  summary.) 
— .    I962d.    On  the  egg  development  and  early  lanae  ofa  trachi- 

noid  fish,  Champsodon  snyden  (Franz).  Bull.  Jpn.  Soc.  Sci.  Fish. 

28(5):499-503. 
— .     1963.    Pelagic  fish  eggs  from  Japanese  waters— III.  Percina. 

VIII.  Cottina.  IX.  Echeneida  and  PIcuroneclida.  Jpn.  J.  Ichthyol. 

11:39-102. 
— .    1966.    Fish  eggs  and  larvae.  In:  Illustrated  encyclopedia  of  the 


722 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


marine  plankton  of  Japan,  Vol.  7,  S.  Motoda  (ed.).  Soyo-Sha,  To- 
kyo. (In  Japanese.) 

.    1979a.    A  perspective  on  morphology  and  phylogeny  offish 

eggs  and  larvae.  Kaiyo  Kaguku  (Mar.  Sci.)  1 1:87-92.  (In  Japanese.) 

.    1979b,    Fish  eggs,  Kaiyo  Kaguku  (Mar.  Sci.)  1 1:126-130.  (In 

Japanese.) 

,  AND  K.  UcHiDA.    1958.   On  the  egg  development  and  hatched 

larvae  of  a  scorpaenoid  fish,  Pterois  lunulata  Temminck  et  Schlegel. 
Sci.  Bull.  Fac.  Agric.  Kyushu  Univ.  16:381-385.  (In  Japanese,  En- 
glish summary,) 

,  M.  Ukawa  AND  M.  HioucHi.    1967.    On  the  larval  and  young 

stages  of  a  serranid  fish,  Epmephelus  akaara  (Temmmck  et  Schle- 
gel). Bull.  Naikai  Reg.  Fish.  Res,  Lab.  25:337-347. 

MocHioKA,N.,S.  K^KLiDA  andO.  Tabeta.  1982.  Congrid  leptocepha- 
li  in  the  western  North  and  Middle  Pacific.  I.  Exterilium  Ariosoma- 
type  larvae,  J,  Fac.  Appl.  Biol.  Sci,  Hiroshima  Univ,  21:35-66, 

Moffatt,  N,  M.  1981.  Survival  and  growth  of  northern  anchovy 
larvae  on  low  zooplankton  densities  as  affected  by  the  presence  of 
a  Chlorella  bloom.  Rapp.  P.  V.  Reun.  Cons.  Int.  Explor.  Mer  178: 
475-480. 

MoHR,  E.  1926.  Die  Gattung  Ze«aR'/iop/en«  Gill.  Zool.  Jahrb.  Jena 
Abt.  Syst,  52:231-266, 

,     1936a.    Hemirhamphiden-Studien  IV,-VI,  IV.  Die  Gattung 

Dermogenys  van  Hasselt.  Mitt.  Zool.  Mus.  Bed.  21(l):34-55. 

.     1936b.    Hemirhamphiden-Studien  IV.-VI.  V.  Die  Gattung 

Nomorhainphus  Weber  &  de  Beaufort.  Mitt.  Zool.  Mus.  Berl.  21(1): 
55-58. 

.     1936c.    Hemirhamphiden-Studien  IV.-VI.  VI.  Die  Gattung 

Hemirhamphodon  Bleeker  1866.  Mitt.  Zool.  Mus.  Berl.  21(1):59- 
64. 

1937.    Revision  der  Centnscidae  (Acanthopterygii  Centrisci- 


formes)  (in  German).  Dana-Rep.  Carlsberg  Found.  18. 
MoHSiN,  A,  K.  M.    1978.    The  osteology  of  the  grey  mullet, /.;ia  ii/^- 

vindis  (Valenciennes).  The  branchiocranium,  vertebral  column,  and 

girdles.  Malayan  Nat.  J.  31(4):203-218. 
MoK,  H.  K.,  AND  S.  C.  Shen.     1983.    Osteology  and  phylogeny  of 

Squamipinnes.  Spec.  Publ.  Taiwan  Mus.  Zool.  Ser.  I, 
MoNOD,  T.     1968.    Le  complexe  urophore  des  poissons  teleosteens. 

Mem.  Inst.  Fondam.  Afr.  Noire.  81. 
MONTALENTI,  G.    1933.    Famiglia  5:  Maenidae,  p.  383-384.  In:  Uova, 

larve  e  stadi  giovanili  di  Teleostei.  Fauna  Flora  Golfo  Napoli 

Monogr.  38. 
.    1937a.    Famiglia  6:  Mullidae,  p,  391-398.  In:  Uova,  larve  e 

stadi  giovanili  di  Teleostei.  Fauna  Flora  Golfo  Napoli  Monogr.  38. 
1937b.  Cepolidae,  p.  407-4 1 0.  In:  Uova,  larve  e  stadi  giovanili 


di  Teleostei.  Fauna  Flora  Golfo  Napoli  Monogr.  38. 

MoNTOLio,  M.  A,  1976,  Estudio  taxonomico  y  morfometrico  de  los 
estadios  larvales  de  dos  especies  de  Carangidae  Decapterus  punc- 
tatus  (Agassiz,  1829)  y  Caranx  crysos  (Mitchill,  1815)  y  su  distri- 
bucion  en  el  Golfo  de  Mexico.  Rev.  Invest.  Inst.  Nac.  Pesca  2:85- 
125. 

MooK,  D.  1977.  Larval  and  osteological  development  of  the  sheeps- 
head,  Archosargitsprobalocephalus {Pisces:  Sparidae).  Copeia  1977: 
126-133. 

MooKERjEE,  H.  K.  1946.  Identification  of  eggs  of  common  freshwater 
fishes  of  Bengal.  Sci.  Cuh.  1 1(1):48. 

,  AND  S,  R,  Mazumdar,    1950.   Some  aspects  of  the  life  history 

oi  Clarias  batrachus  (Linn.).  Proc.  Zool.  Soc.  Bengal  3(l):71-83. 

Moore,  C.  J.  1980.  Spawning  of  Menidia  menidia  (Pisces:  Atherini- 
dae).  Copeia  1980:886-887. 

Moore,  D,  1962,  Development,  distribution,  and  comparison  of  rud- 
der fishes  Kyphosus  sectatnx  (Linnaeus)  and  A',  incisor  (Cuvier)  in 
the  western  North  Atlantic,  U,  S,  Fish  Wildl.  Serv.  Fish.  Bull.  196: 
451-480, 

MoREiRA,  C.  1920.  Recherches  sur  la  reproduction  de  l'//op//ai  ma/- 
abaricus (Bloch)  et  sur  I'lncubation  d'oeufs  de  Salmo  fario  au  Bresil. 
Bull.  Soc.  Zool.  Fr.  44:329-336, 

MoRELAND,  J.     1960.    A  new  species  and  genus  of  congiopodid  fish 


from  southern  New  Zealand.  Rec.  Dom.  Mus.  (Wellington)  3:241- 

246, 
Moreno,  C,  A.    1980.   Observations  on  food  and  reproduction  in  Tre- 

matomus  bernacchii  (Pisces:  Nototheniidae)  from  the  Palmer  Ar- 
chipelago, Antarctica.  Copeia  1980:171-173. 
Mori,  K,    1984,    Early  life  history  of  Luljanus  villa  (Lutjanidae)  in 

Yuya  Bay,  the  Sea  of  Japan.  Jpn.  J.  Ichthyol.  30:374-392. 
,  S.  Uevanagi  AND  Y.  NisHiKAWA.    1971.    The  development  of 

artificially  fertilized  and  reared  larvae  of  the  yellowfin  tuna,  Thun- 

nus  albacares.  Bull.  Far  Seas  Fish.  Res.  Lab.  (Shimizu)  5:219-232. 
MoRiN,  R.  P.,  andK.  W.  Able.    1983.   Patterns  of  geographic  variation 

in  the  egg  morphology  of  the  fundulid  fish,  Fundulus  hcteroditus. 

Copeia  1983:726-740, 
MoRiTA,  S.     1984.    History  of  the  hemng  fishery  and  review  of  the 

artificial  propagation  technique  for  herring  in  Japan.  Can.  J.  Spec. 

Publ.  Fish.  Aquat.  Sci.  In  press. 
MoRK,  J.,  P.  SoLEMDAL  AND  G.  Sl'ndness.     1983.    Identification  of 

marine  fish  eggs:  a  biochemical  genetics  approach.  Can.  J.  Fish. 

Aquat.  Sci.  40:361-369. 
MoRovic,  D.    1953.  Sur  la  determination  des  mugesadnatiquesd'apres 

la  forme  de  I'otolith  sagitta.  Notes  Inst.  Oceanogr.  Split  9:1-7. 
Morris,  D,  E,,  and  R,  K^navama.    1964-1969.    Life  history  study  of 

the  Moi,  Polydactylus  sextiUs.  Job  Completion  Rep.,  Proj.  No. 

F-S-R-1 1.  Div.  Fish  Game,  Honolulu,  Hawaii.  Mimeogr. 
Morris,  R,    1951.    Early  development  of  the  cottid  fish  Clinocottus 

recalvus  (Greeley).  Calif  Fish  Game  37:281-300. 
.    MS.    Some  notes  on  the  early  life  history  of  the  night  smelt, 

Spirinchus  slarksi  (Fisk). 
Morrow,  J.  E.,  Jr.    1964a.    Family  Chauliodontidae,  p.  274-289. /«. 

Fishes  of  the  western  North  Atlantic.  H.  B.  Bigelow  (ed.).  Mem. 

Sears  Found.  Mar.  Res.  No.  1  (Pt.  4). 
.     1964b.    Stomiatidae,  p.  290-310.  In:  Fishes  of  the  western 

North  Atlantic.  H.  B.  Bigelow  (ed).  Mem.  Sears  Found.  Mar.  Res. 

No.  1  (Pt.  4). 
.    1964c.    Family  Malacosteidae,  p,  523-549.  In:  Fishes  of  the 

western  North  Atlantic.  H.  B.  Bigelow  (ed).  Mem.  Sears  Found. 

Mar.  Res.  No.  1  (Pt.  4). 
.    1980.  The  freshwater  fishes  of  Alaska.  Alaska  Northwest  Publ. 

Co.,  Anchorage, 
,  and  R.  H.  Gibbs,  Jr. 


1964,  Family  Melanostomiatidae,  p. 
351-510.  In:  Fishes  of  the  western  North  Atlantic.  H.  B.  Bigelow 
(ed.).  Mem.  Sears  Found.  Mar.  Res.  No.  1  (Pt.  4).  (Genera  Metan- 
oslomias  and  Eustomias  by  R.  H.  Gibbs,  Jr,,  the  remainder  by  J. 
E.  Morrow,  Jr.) 

,  AND  S.  J.  Harbo.    1 969.  A  revision  of  the  sailfish  genus  Islioph- 

orus.  Copeia  1969:34-44. 

MosER,  H.  G.  1972.  Development  and  geographical  distnbution  of 
the  rockfish,  Sebasles  macdonaldi  (Eigenmann  and  Beeson,  1893), 
family  Scorpaenidae,  off  southern  California  and  Baja  California. 
Fish.  Bull.  U.S.  70:941-958. 

.    1974.    Development  and  distribution  of  larvae  and  juveniles 

oi Sebaslolobus  {Pisces;  Family  Scorpaenidae).  Fish.  Bull.  U.S.  72: 
865-884. 

.     1981.    Morphological  and  functional  aspects  of  marine  fish 

larvae,  p.  89-131.  In:  Marine  fish  larvae.  Morphology,  ecology  and 
relation  to  fisheries.  R.  Lasker  (ed).  Univ.  Washington  Press,  Se- 
attle. 

,  and  E.  H,  Ahlstrom.     1970.    Development  of  lantemfishes 

(family  Myctophidae)  in  the  California  Current.  Part  I.  Species  with 
narrow-eyed  larvae.  Nat.  Hist.  Mus.  Los  Ang.  Cty.  Sci.  Bull.  7. 

,  and .    1972.   Development  of  the  lantemfish,  Scopelop- 

sis  multipunclalus  Brauer  1 906,  with  a  discussion  of  its  phylogenetic 
position  in  the  family  Myctophidae  and  its  role  in  a  proposed 
mechanism  for  the  evolution  of  photophore  patterns  in  lantem- 
fishes. Fish.  Bull.  U.S.  70:541-564, 

,  AND .  1974.  Role  of  larval  stages  in  systematic  inves- 
tigations of  marine  teleosts:  the  Myctophidae,  a  case  study.  Fish, 
Bull,  U,S.  72:391-413. 


LITERATURE  CITED 


723 


.AND .    1978.    Larvae  and  pelagic  juveniles  of  blackgill 

rockfish.  Sehastes  melanoslomus.  taken  in  midwaler  trawls  off 
southern  California.  J.  Fish.  Res.  Board  Can.  35:981-996. 

, .  D.  Kramer  and  E.  G.  Stevens.    1973.    Distribution 

and  abundance  of  fish  eggs  and  larvae  in  the  Gulf  of  California. 
Cal.  Coop.  Oceanic  Fish.  Invest.  Rep.  17:1  12-128. 

, andE.  M.Sandknop.    1977.  Guide  to  the  identification 

of  scorpionfish  larvae  in  the  eastern  Pacific  with  comparative  notes 
on  species  of  Sebastes  and  Helicolemis  from  other  oceans.  NOAA 
Tech.  Rep.  NMFS  Circ.  402. 

— .  AND  J.  L.  Butler.    1981.  Description  ofreared  larvae  and  early- 


juveniles  of  the  calico  rockfish,  Sebastes  dallii.  Calif.  Coop.  Oceanic 
Fish.  Invest.  Rep.  22:88-95. 

,  AND .   MS.    Descriptions  of  reared  larvae  of  six  species 

of  Sebastes.  In:  Proc.  of  a  workshop  on  widow  rockfish  {Sebastes 
entomelas).  W.  H.  Lenarz  and  D.  Gunderson  (eds.). 

MosHER,  C.  1954.  Observations  on  the  spawning  behavior  and  the 
early  larval  development  of  the  Sargassum  fish,  Histno  histno  (Lin- 
naeus). Zoologica  (N.Y.)  39:141-152. 

MoiiLTciN,  L.  1970.  The  1970  longfin  smelt  population  spawning  run 
in  Lake  Washington  with  notes  on  egg  development  and  changes 
in  the  population  since  1964.  LInpubl.  MS  thesis.  Univ.  Washing- 
ton, Seattle. 

Mover,  J.  T.  1979.  Mating  strategies  and  reproductive  behaviour  of 
ostraciid  fishes  at  Miyake-jima.  Japan.  Jpn.  J.  Ichthyol.  26:148- 
160. 

MujiB.  K.  A.  1967.  The  cranial  osteology  of  the  Gadidae.  J.  Fish.  Res. 
Board  Can.  24:1315-1375. 

.    1969.    Cranial  osteology  of  Brosme  bromse  with  notes  on  its 

position  within  the  subfamily  Lotinae  (Gadidae).  J.  Fish.  Res.  Board 
Can.  ItAlX-AJ,!. 

Mi'KHACHEVA,  V.  A.  1957.  Notes  on  the  development  of  the  Far 
Eastern  na\aga(Eleginus gracilis1'\\e^\\K).  Tr.  Inst.  Okeanol.  Akad. 
Nauk  SSSR  20:356-370.  (In  Russian.  Transl.  by  Am.  Inst.  Biol. 
Sci.,  1959,  p.  291-302.) 

.     1964.    The  composition  of  species  of  the  genus  Cyclothone 

(Pisces,  Gonoslomatidae)  in  the  Pacific  Ocean,  p.  98-146.  In:  Fish- 
es of  the  Pacific  and  Indian  Oceans.  Biology  and  Distribution.  T. 
S.  Rass(ed.).  Tr.  Inst.  Okean.  73.  (English  translation  1966  by  Isr. 
Program  Sci.  Transl.) 

.     1978.    A  review  of  the  species  of  Diplophos  Giinther  (Gon- 

ostomatidae,  Osteichthyes)  and  their  vertical  and  geographical  dis- 
tribution. Trans.  P.  P.  Shirshov  Inst.  Ocean.  1  1  1:10-27.  (In  Rus- 
sian, English  abstract.) 

,  AND  O.  A.  ZviAGiNA.    1960.    Development  of  the  Pacific  cod, 

Gadus  morhua  macrocephalus  Tilesius.  Tr.  Inst.  Okeanol.  Akad. 
Nauk  SSSR  31:145-165.  (Transl.  from  Russian  by  W.  E.  Ricker, 
1969,  Fish.  Res.  Board  Can..  Transl.  Ser.  393.) 

MOller,  H.,  AND  G.  Sterba.  1963.  Elektronenmikroskopische  Un- 
tersuchungen  iiber  Bildung  und  Struktur  der  EihiiUen  bei  Knock- 
enfischen.  II.  Die  Eihiillen  jungerer  und  alterer  Oozytcn  von  Cy- 
no/eft(a5 /)W/<:)K;VSteindachner(Cyprinodontidae).  Zool.  Jahrb.  Abt. 
Anal.  80(4):469-488. 

Munk,  O.  1966.  Ocular  anatomy  of  some  deep-sea  teleosts.  Dana- 
Rep.  Carlsberg  Found.  70. 

.  AND  E.  Bertelsen.     1980.    On  the  esca  light  organ  and  its 

associated  light-guiding  structures  in  the  deep-sea  anglerfish  Chaen- 
ophryne  draco.  Vindensk.  Medd.  Dan.  Naturhist.  Foren.  142:103- 
129. 

MuNRO.  I.  S.  R.  1945.  Postlarval  stages  of  Australian  fishes— No.  1. 
Mem.  Queensl.  Mas.  12:148-152. 

.     1950.    Revision  of  Bregmaceros  with  descriptions  of  larval 

stages  from  Australasia.  Proc.  R.  Soc.  Queensl.  61:37-53. 

1954.    Eggs  and  larvae  of  the  four-winged  flying  fish,  l/irun- 


Dasson  i^earf;  (Whitley)  (Blenniidae).  Aust.  J.  Mar.  Freshwater  Res. 

6:30-34. 
.    1967.    The  Fishes  of  New  Guinea.  New  Guinea  Dep.  Agric. 

Stock  Fish..  Port  Moresby. 
Mlntz,  W.  R.  a.    1976.    On  yellow  lenses  in  mesopelagic  animals.  J. 

Mar.  Biol.  Assoc.  U.K.  56:963-976. 
Murray,  J.,  and  J.  Hjort.    1912.   The  depths  of  the  ocean.  Macmillan 

and  Co.,  London. 
MusicK,  J.  A.    1973.   A  meristic  and  morphometric  comparison  of  the 

hakes,  Vrophycis  chuss,  and  V.  tenuis  (Pisces,  Gadidae).  Fish.  Bull. 

U.S.  71:479-488. 
Musienko,  L.  N.    1963.    Ichthyoplankton  of  the  Bering  Sea  (data  of 

the  Bering  Sea  Expedition  of  1958-1959),  p.  251-286.  In:  P.  A. 

Moiseev  (ed.).  Soviet  Fisheries  Investigations  in  the  Northeast  Pa- 
cific, I .  Tr.  Vses.  Nauchno-Issled.  Inst.  Morsk.  Rybn.  Khoz.  Okean- 

ogs.  48. 
Musiv,  Y.  I.,  and  V.  A.  Sergi-ienko.    1977.    Fingerlings  of  the  genus 

Lutjanus  (Lutjanidae,  Perciformes)  from  the  Gulf  of  Aden.  J.  Ich- 
thyol. (Engl.  Transl.  Vopr.  Ikhtiol.)  17:151-154. 
MiERS,  G.  S.   1937.  The  deep-sea  Zeomorph  fishes  of  the  family  Gram- 

micolepidae.  Proc.  U.S.  Natl.  Mus.  84:145-156. 
.    1958.    Trends  in  the  evolution  of  teleostean  fishes.  Stanford 

Ichthyol.  Bull.  7(3):27-30. 
.     1960.    A  new  zeomorph  fish  of  the  Family  Oreosomatidae 

from  the  coast  of  California,  with  notes  on  the  family.  Stanford 

Ichthyol.  Bull.  7(4):89-98. 
,  AND  W.  C.  Freihofer.    1966.    Megalomyctendae.  a  previously 

unrecognized  family  of  deep-sea  cetomimiform  fishes  based  on  two 

new  genera  from   the   North   Atlantic.   Stanford   Ichthyol.   Bull. 

8:193-206. 
,  AND  C.  B.  Wade.   1946.  New  fishes  of  the  families  Dactyloscop- 

idae,  Microdesmidae,  and  Antennariidae  from  the  west  coast  of 

Mexico  and  the  Galapagos  Islands  with  a  brief  account  of  the  use 

of  rotenone  fish  poisons  in  ichthyological  collecting.  Allan  Hancock 

Found.  Pac.  Exped.  9(6):I51-179. 
Nafpaktitis,  B.  G.    1975.  Review ofthelantemfish genus A'oroscope/iw 

(family  Myctophidae)  in  the  north  Atlantic  and  Mediterranean. 

Bull.  Mar.  Sci.  25:75-87. 
.    1977.   Family  Neoscopelidae,  p.  1-12. /w.Fishesofthe  western 

North  Atlantic.  R.  H.  Gibbs,  Jr.  (ed).  Mem.  Sears  Found.  Mar. 

Res.  No.  l(Pt.  7). 
.    1978.   Systematics  and  distribution  of  lantemfishes  of  the  gen- 


dichthys  speculiger  (Valenciennes).  .'\ust.  J.  Mar.  Freshwater  Res. 
5(l):64-69. 
.     1955.    Eggs  and  larvae  of  the  sabre-toothed  oyster  blenny 


era  Lobianchia  and  Diaphiis  (Myctophidae)  in  the  Indian  Ocean. 
Nat.  Hist.  Mus.  Los  Ang.  Cty.  Sci.  Bull.  30:1-92. 

,  and  M.  Nafpaktitis.  1969.  Lantemfishes  (family  Myctophi- 
dae) collected  during  cruises  3  and  6  of  the  R'V  "Anton  Bruun" 
in  the  Indian  Ocean.  Nat.  Hist.  Mus.  Los  Ang.  Cty.  Sci.  Bull.  5:1- 
79. 

,  and  J.  R.  Paxton.     1968.    Review  of  the  lantemfish  genus 

Lampadena  with  a  description  of  a  new  species.  Contrib.  Sci.  (Los 
Ang.)  138. 

,  and .    1978.    W(o/vc/i/iJ«.  a  new  genus  of  Myctophidae 

based  on  Diaphus  urolampus.  Copeia  1978:492-497. 

,  R.  H.  Bakus,  J.  E.  Craddcxk,  R.  L.  Haedrich.  B.  H.  Robison 

AND  C.  Karnella.  1977.  Family  Myctophidae,  p.  13-265.  In: 
Fishes  of  the  western  North  Atlantic.  R.  H.  Gibbs,  Jr.  (ed.).  Mem. 
Sears  Found.  Mar.  Res.  No.  l(Pt.  7). 

Nagiec,  C.  1 979.  On  the  developmental  stages  of  the  fins  and  axial 
skeleton  of  whitefish  Coregunus  lavaretus  L.  Pol.  Arch.  Hydrobiol. 
26:529-543. 

Nair,  G.  S.  1958.  On  the  breeding  habits  and  development  of  .4m- 
bassis  gymnocephalus  {Lac).  Bull.  Cent.  Res.  Inst.  Univ.  Travan- 
core,  India  5C(l):69-76. 

N^iR,  R.  V.  1952a.  Studies  on  some  fish  eggs  and  larvae  of  the  Madras 
plankton.  Proc.  Indian  Acad.  Sci.  Sec.  B  35:181-208. 

.  1952b.  Studies  on  some  post-larval  fishes  of  the  Madras  plank- 
ton. Proc.  Indian  Acad.  Sci.  Sect.  B  35:225-244. 

.    1960.    Notes  on  the  spawning  habits  and  early  life-history  of 


724 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


the  oil  sardine,  Sardmella  longiceps  Cuv.  &  Val.  Indian  J.  Fish. 

6:342-359. 
Nakabo,  T.     1982a.    Revision  of  the  family  Draconeltidae.  Jpn.  J. 

Ichthyol.  28:335-367. 
.    1982b.   Revision  of  genera  of  the  dragonets  (Pisces:  Callionym- 

idae).  Publ.  Seto  Mar.  Biol.  Lab.  27:77-131. 
Nakah,\ra,  K.    1962.   A  note  on  the  larvae  of  .-l/i/;.?o/»u  fu/iro.v  Lowe 

and  Erythrocles  schlegeli  (Richardson)  collected  off  the  southern 

coast  of  Kyushu.  Bull.  Jpn.  Soc.  Sci.  Fish.  28:484-488. 
Nakamlira,  L    1974.   Some  aspects  of  the  systematics  and  distribution 

of  billfishes.  NOAA  Tech.  Rep.  NMFS  SSRF-675(2):45-53. 
.    1977.    Gempylidae.  //;.  FAO  species  identification  sheets  for 

fishery  purposes.  Western  Central  Atlantic,  Fishing  Area  31.  Vol. 

2.  FAO.  Rome. 
.    1980.    New  record  of  a  rare  gempylid,  Thyrsiloides  inarleyi. 

from  the  Sea  of  Japan.  Jpn.  J.  Ichthyol.  26:357-360. 
,  AND  E.  Fuji.    1983.    A  new  genus  and  species  of  Gempylidae 

(Pisces:  Perciformes)  from  Tonga  Ridge.  Publ.  Seto  Mar.  Biol.  Lab. 

28(4/6):173-191. 
,  T.  IwAi  AND  K.  Matshbara.    1968.    A  review  of  the  sailfish. 


spearfish.  marlin  and  swordfish  of  the  world.  Kyoto  Univ.  Spec. 

Rep.  4. 
Nakamlira,  M.    1951.  Thelifehistory  ofacyprinid  fish.  Opsariichthys 

uncirostris  (Temminck  et  Schlegel)  in  Lake  Biwa.  Misc.  Rep.  Res. 

Inst.  Nat.  Resour.  (Tokyo)  19-21:70-78. 
Nakamura,  M.  N.    1969.   Cyprinid  fishes  of  Japan:  Studies  on  the  life 

history  of  cyprinid  fishes  of  Japan.  Res.  Inst.  Nat.  Resour.  (Tokyo) 

Spec.  Publ.  4. 
Nakamura,  S.    1936.   Larvae  and  young  of  fishes  found  in  the  vicinity 

of  Kominato,  II-IV.  II.  Athenon  elynnis  loTdiin  and  Starks.  J.  Imp. 

Fish.  Inst.  (Tokyo)  30:219-225. 
.     1937.    Larvae  and  young  of  fishes  found  in  the  vicinity  of 

Kominato,  VII-VIII.  J.  Imp.  Fish.  Inst.  (Tokyo)  32:17-25. 
Naplin,  N.  A.,  andC.  L.  Obenchain.    1980.   A  description  of  eggs  and 

larvae  of  the  snake  eel,  Pisodonophis  crueiilifer  (Ophichlhidae). 

Bull.  Mar.  Sci.  30:413-423. 
Nash,  C.  E.     1973.    Automated  mass-production  of  Anemia  salina 

nauplii  for  hatcheries.  Aquaculture  2:289-298. 
,  C.-M.  Kiio,  W.  D.  MA15I5EN  AND  C.  L.  Paulsen.    1977.   Swim 

bladder  inflation  and  survival  of  Mugil  cephatus  to  50  days.  Aqua- 
culture  12:89-94. 
Nassogne,  A.    1970.   Influence  of  food  organisms  on  the  development 

and  culture  of  pelagic  copepods.  Helgol.  Wiss.  Meeresunters.  20: 

333-345. 
Nellen,  W.    1973a.    Untersuchungen  zur  Verteilung  von  Fischlarven 

und  Plankton  im  Gebiet  der  Grossen  Meteorbank.  Meteor  For- 

schungsergeb.  13:47-69. 
.     1973b.     Fischlarven   des   Indischen   Ozeans.    Meteor   For- 

schungsergeb.  Biol.  14:1-66.  (English  summary.) 
,  AND  G.  Hempel.     1970.    Beobachtungen  am  Ichthyoneuston 

der  Nordsee.  Ber.  Dtsch.  Wiss.  Komm.  Meeresforsch.  21:31  1-348. 
Nellis,  D.  W.     1980.    Reproduction  in  the  ocean  triggerfish  Cantht- 

dennis  sufflamen.  Caribb.  J.  Sci.  16(1-4):  1 67. 
Nelsen.O.  E.    1953.  Comparative  embryology  of  the  vertebrates.  Blak- 

iston,  New  York. 
Nelson,  G.  J.    1 966a.  Osteology  and  relationships  of  the  eel,  Neenche- 

lys  huilendijki.  Copeia  1966:321-324. 
.    1966b.    Gill  arches  of  leleostean  fishes  of  the  order  Anguilli- 

formes.  Pac.  Sci.  20:391-408. 
.    1967.   Gill  arches  of  teleostean  fishes  of  the  family  Clupcidac. 

Copeia  1967:389-399. 
.    1969.    Gill  arches  and  the  phylogeny  of  fishes,  with  notes  on 

the  classification  of  vertebrates.  Bull.  Amer.  Mus.  Nat.  Hist.  141: 

479-552. 
.    1970.   The  hyobranchial  apparatus  of  teleostean  fishes  of  the 

families  Engraulidae  and  Chirocentridae.  Am.  Mus.  Novit.  2410. 
.    1972.   Cephalic  sensory  canals,  pitlines,  and  the  classification 

of  esocoid  fishes,  with  notes  on  galaxiids  and  other  teleosts.  Am. 

Mus.  Novit.  2492:1-49. 


— .  1973.  Relationships  of  clupeomorphs,  with  remarks  on  the 
structure  of  the  lower  jaw  in  fishes,  p.  333-349.  In:  Interrelation- 
ships of  fishes.  P.  H.  Greenwood.  R.  S.  Miles  and  C.  Patterson 
(eds.).  Zool.  J.  Linn.  Soc.  53  (Suppl.  1). 

.    1978.   Ontogeny,  phylogeny,  palaeontology,  and  the  biogenetic 

law.  Syst.  Zool.  27:324-345. 

.    1983.  Anchoa  argouixitlala  with  notes  on  other  eastern  Pacific 

anchovies  and  the  Indo-Pacific  genus  Encrasicholina.  Copeia  1983: 
48-54. 

.    1984.    Identity  of  the  anchovy  Hildebrandichlhys  setiger  with 

notes  on  relationships  and  biogeography  of  the  genera  Engraulis 
and  Cetengraulis.  Copeia  1984:422^27. 

,  AND  N.  Platnick.    1981.   Systematics  and  biogeography.  Cla- 


distics  and  vicariance.  Columbia  Univ.  Press,  New  York. 

Nelson.  J.  S.  1975.  Records  of  a  new  form  of  the  marine  cottid  fish 
Anltpodocoltus  gatatheac  from  the  east  coast  of  New  Zealand.  Natl. 
Mus.  N.Z.  Rec.  l(4):80-86. 

.    1976.    Fishes  of  the  world.  John  Wiley  &  Sons.  New  York. 

.    1 977.   Fishes  of  the  Southern  Hemisphere  genus  Neophrynich- 

ihys  (Scorpaeniformes:  Cottoidei),  with  descriptions  of  two  new 
species  from  New  Zealand  and  Macquarie  Island.  J.  R.  Soc.  N.Z. 
7:485-511. 

.    1978.    Limnichlhys  polyaclis.  a  new  species  of  blennioid  fish 

from  New  Zealand,  with  notes  on  the  taxonomy  and  distribution 
of  other  Creediidae  (including  Limnichthyidae).  N.Z.  J.  Zool. 
5:351-364. 

.     1982.    Two  new  South  Pacific  fishes  of  the  genus  Ehinania 

and  contributions  to  the  systematics  of  Psychrolutidae  (Scorpaen- 
iformes). Can.  J.  Zool.  60:1470-1504. 

Nichols,  J.  H.  1971.  Pleuronectidae.  Cons.  Int.  Explor.  Mer.  Fiches 
d'identification  des  oeufs  et  larves  de  poissons.  No.  4-6:2-18. 

Nichols,  J.  T.,  and  C.  M.  Breder.  Jr.  1927.  The  marine  fishes  of 
New  York  and  southern  New  England.  Zoologica  (N.Y.)  9(1):1- 
192. 

,  AND  C.  M.  Breder.  Jr.    1928.   An  annotated  list  of  the  Synen- 

tognathi  with  remarks  on  their  development  and  relationships.  Col- 
lected by  the  Arcturus.  Zoologica  (N.Y.)  8:423-448. 

Nielsen,  J.  G.  1963a.  On  the  development  of  Cheilodaclylus  varie- 
.?<3««  Valenciennes  1 883  (Cheilodactylidae).  Copeia  1963:528-533. 

.    1963b.   Notes  on  some  Heterosomata  (Pisces)  from  NW  South 

America  with  the  description  of  a  new  genus  and  species  and  a  new 
subspecies  of  Paralichthinae.  Vidensk.  Medd.  Dan.  Naturhist.  For- 
en.  125:375-400. 

.     1963c.    Description  of  two  large  unmetamorphosed  flatfish 

larvae  (Heterosomata).  Vidensk.  Medd.  Dan.  Naturhist.  Foren  1 25: 
401-406. 

.    1964.    Fishes  from  depths  exceeding  6000  meters.  Galathea 

Rep.  7:113-124. 

.     1966.    Synopsis  of  the  Ipnopidae  (Pisces,  Iniomi).  Galathea 


Rep.  8:49-75. 
.     1969.    Systematics  and  biology  of  the  Aphyonidae  (Pisces, 

Ophidioidea).  Galathea  Rep.  10:7-88. 
.    1973.   Stylephondae.  p.  335. //I.  Check-list  of  the  fishes  of  the 

north-eastern  Atlantic  and  of  the  Mediterranean  (Clofnam).  Vol.  I. 

J.  C.  Hureau  and  T.  Monod  (eds.).  UNESCO,  Pans. 
,  .'VND  J.-C.  HuREAi'.     1980.    Revision  of  the  ophidiid  genus 

Spectninculus.  a  senior  synonym  of  Parabassogigas  (Pisces,  Ophi- 

diiformes).  Steenstrupia  6(1  1):  149-1 69. 
,  AND  D.  G.Smith.    1978.  The  eel  family  Nemichthyidae(Pisces, 


Anguilliformes).  Dana-Rep.  Carlsberg  Found.  88:1-71. 

NiKOLSKY,  G.  V.  1961.  Special  ichthyology.  Second  revised  and  en- 
larged edition  (translated  from  Russian,  original  date  of  publication 
1954).  Jerusalem.  (Not  seen;  cited  by  Tyler,  1968.) 

.    1963.    The  ecology  of  fishes.  Academic  Press.  London. 

NisHiKAWA,  Y.  1979.  Early  development  of  the  double-lined  mackerel, 
Grammatorcymis  bicarinatiis  (Quoy  and  Gaimard),  from  the  west- 
em  tropical  Pacific.  Bull.  Far  Seas  Fish.  Res.  Lab.  (Shimizu)  17: 
125-140. 

.    1982.    Early  development  of  the  fishes  of  the  family  Gempy- 


LITERATURE  CITED 


725 


lidae.  I.  Larvae  and  juveniles  of  escolar,  Lepidocybium  flavobnin- 
neum  (Smith).  Bull.  Far  Seas  Fish.  Res.  Lab.  (Shimizu)  ( 19):  1-13. 

,  AND  L  N.AKAMLiRA.     1978.    Post  larvae  and  juveniles  of  the 

gempylid  fish  Neoepinnuta  onentalis  (Gilchrist  and  Von  Bonde), 
from  the  North  Arabian  Sea.  Bull.  Far  Seas  Fish.  Res.  Lab.  (Shim- 
izu) (16):75-91. 

NiSHiMURA.  S.  1964.  Additional  mformation  on  the  biology  of  the 
dealfish,  Trachiplerus  ishikawai  Jordan  and  Snyder.  Bull.  Jpn.  Sea 
Reg.  Fish.  Res.  Lab.  13:127-129. 

.    1966.    Early  life  history  of  the  deep-sea  smelt,  Glossarwdon 

semifasciatus  (Kishinouye)  (Teleostei:  Clupeida).  Pt.  1.  Publ.  Seto 
Mar  Biol.  Ub.  13:349-360. 

NoAKES,  D.  L.  G.,  AND  E.  K.  Balon.  1982.  Life  histories  of  tilapia: 
an  evolutionary  perspective,  p.  61-68.  In:  biology  and  culture  of 
tilapia.  R.  S.  V.  Pullin  and  R.  H.  Lowe-McConnell  (eds.).  ICLARM 
Conference  Proceedings  7.  International  Center  for  Living  Aquatic 
Research  Management,  Manila,  Philippines. 

NoRCROSS,  J.  J.,  W.  H.  Massmann  and  E.  B.  Joseph.  1961.  Investi- 
gations of  inner  continental  shelf  waters  off  lower  Chesapeake  Bay. 
Part  2.  Sand  lance  larvae,  Ammodvles  amencanus.  Chesapeake  Sci. 
2:46-59. 

.  S.  L.  Richardson,  W.  H.  Massman  and  E.  B.  Joseph.    1974. 

Development  of  young  bluefin  (Pomalonms  saltalnx)  and  distri- 
bution of  eggs  and  young  in  Virginia  coastal  waters.  Trans.  Am. 
Fish.  Soc.  103:477-497. 

Norden,  C.  R.  1961.  Comparative  osteology  of  representative  sal- 
monid  fishes  with  particular  reference  to  the  grayling  {Thymallus 
arcticus)  and  its  phylogeny.  J.  Fish.  Res.  Board  Can.  18:679-791. 

Norman,  J.  R.  1926a.  A  report  on  the  flatfishes  (Heterosomata)  col- 
lected by  the  F.  I.  S.  "Endeavour,"  with  a  synopsis  of  the  flatfishes 
of  Australia  and  a  revision  of  the  Rhombosoleinae.  Biol.  Res.  Fish- 
ing Exped.  F.  I.  S.  Endeavour  5:217-308. 

.     1926b.    The  development  of  the  chondrocranium  of  the  eel 

(Anguilta  vulgaris)  with  observations  on  the  comparative  mor- 
phology and  development  of  the  chrondrocranium  in  bony  fishes. 
Philos.  Trans.  R.  Soc.  Lond.  B.  Biol.  Sci.  214:369-464. 

.    1929.    The  teleostean  fishes  of  the  family  Chiasmodontidae. 

Ann.  Mag.  Nat.  Hist.  3(10): 529-544. 

.     1930.    Oceanic  fishes  and  flatfishes  collected  in  1925-1927. 

Part  I.  Oceanic  fishes.  Part  II.  Flatfishes.  Discovery  Rep.  2:261- 
370. 

.  1934.  A  systematic  monograph  of  the  flatfishes  (Heteroso- 
mata). Vol.  I .  Pseltodidae,  Bothidae.  Pleuronectidae.  Br.  Mus.  (Nat. 
Hist.)  Lond. 

.    1937.    Coast  fishes,  part  2,  the  Patagonian  Region.  Discovery 

Rep.  16:1-50. 

.     1938a.    Coast  fishes,  part  3,  the  Antarctic  Zone.  Discovery 

Rep.  18:1-104. 

.    1938b.    On  the  aflinities  of  the  Chilean  fish,  Normanichthvs 


crocken  C\avk.  Copeia  1938:29-32. 
.    1966.    A  draft  synopsis  of  the  orders,  families  and  genera  of 

recent  fishes  and  fish-like  vertebrates.  Br.  Mus.  (Nat.  Hist.)  Lond. 
North,  A.  W.,  AND  M.  G.  White.    1982.    Key  to  fish  postlarvae  from 

the  Scotia  Sea,  Antarctica.  Cybium,  3rd  Ser.  6(1):  13-32. 
Nibelin,0.    1947.  Antarctic  fishes.  Sci.  Results  Norw.  Antarct.  Exped. 

26. 
.    1948.    Fishes  collected  by  the  "Skagerak"  Expedition  in  the 

eastern  Atlantic  1946.  Goteb.  Vetensk.  Samh.  Handl.  6:1-95. 
.    1951.    Subantarctic  and  Antarctic  fishes.  Sci.  Results  Brategg 

Exped.  2. 
Obrhelova.N.    1961.   Vergleichende  Osteologie  der  tertiaren  Susswas- 

serfische  Bohmens  (Gobioidei).  Sb.  Ustred,  LIstavu  Geol.  26:103- 

192. 
OcHiAi.  A.,  AND  K.  Amaora.    1963.    Description  of  larvae  and  young 

of  four  species  of  flatfishes  referable  to  subfamily  Bothidae.  Bull. 

Jpn.  Soc.  Sci.  Fish.  29(2):  127-1 34.  (In  Japanese.) 
,  AND  Y.  Nozawa.    1980.   On  the  metamorphosis  and  burrowing 

habits  of  the  snake  eel,  Muraenichthvs  gymnolus.  Jpn.  J.  Ichthyol. 

27:237-242. 


O'CoNNELL.  C.  P.  1953.  The  life  history  of  the  cabezon,  Scorpaenich- 
thys  mannoralus  (Ayres).  Calif  Dep.  Fish  Game  Fish  Bull.  93:1- 
10. 

.  1976.  Histological  criteria  for  diagnosing  the  starving  con- 
dition in  early  post  yolk-sac  larvae  of  the  northern  anchovy  En- 
graulis  mordax  Girard.  J.  Exp.  Mar.  Biol.  Ecol.  25:285-312. 

.     1980.    Percentage  of  starving  northern  anchovy,  Engraiilis 

mordax.  larvae  in  the  sea  as  estimated  by  histological  methods. 
Fish.  Bull.  U.S.  78:475-489. 

.    1981a.   Development  oforgan  systems  in  the  northern  anchovy 

Engraulis  mordax  and  other  teleosts.  Am.  Zool.  21:429-446. 

.    1981b.    Estimation  by  histological  methods  of  the  percent  of 

starving  northern  anchovy  larvae  in  the  sea.  Rapp.  P.  V.  Reun. 
Cons.  Int.  Explor.  Mer  178:357-360. 

O'Day,  W.  T.  1972.  The  histology  and  fine  structure  of  some  biolu- 
minescent  organs  in  deep  sea  fishes.  Unpubl.  PhD  dissertation. 
Univ.  South.  Calif,  Los  Angeles. 

.    1974.    Bacterial  luminescence  in  the  deep-sea  anglerfish  OnW- 

rodes  acanthias  {GWhen.  1915).  Contnb.  Sci.  (Los  Ang.)  255:1-12. 

Oelschlager,  H.  a.  1 974.  Das  Caudalskelett  von  Lampns  regius  und 
seine  Ableitung  von  I'elUer  hvpselopterus.  Senckenb.  Biol.  55(  1/3): 
77-85. 

.     1976a.    On  the  evolution  and  ecological  adaptations  of  the 

AUotriognathi.  Rev.  Trav.  Int.  Peches  Mant.  40(3, 4):69 1-694. 

.    1976b.   Gotteslachse  und  Bandfische  — Erforschungsgeschichle 

und  Evolution  der  Glanzfischartigcn.  Nat.  Mus.  l06(2):48-54. 

.  1978a.  Comparative  morphological  and  functional  investi- 
gations on  the  pelvis  of  the  oarfishes  (Teleostei:  AUotriognathi). 
Zool.  Beitr.  24(2):  165-192. 

.    1978b.  On  the  morphology  and  biomechanics  of  .4g/'o«rt'/!?/!V'.s 


Phillipps  in  comparison  with  other  AUotriognathi  (Pisces,  Teleos- 
tei). Zool.  Jahrb.  Abt.  Anat.  100:118-153. 
.    1979.   On  the  functional  osteology  of  the  Lophotidae  (Pisces: 

Teleostei).  Zool.  Jahrb.  Abt.  Anat.  102:336-360. 
Ogasawara.  Y..  S.  Akatsu  AND  Y,  Taki.    1978.   Juvenile  stages  and 

effect  of  salinity  on  the  survival  of  larvae  and  juveniles  in  the  striped 

fingerfish,  Monodaclylus  sebae.  Jpn.  J.  Ichthyol.  24:246-250. 
Ohsima,  Y.,  AND  N.  Nakamiira.     1941.    Note  on  the  life  history  of 

Triacanthus  breviroslns  Temminck  et  Schlegel.  Bull.  Jpn.  Soc.  Sci. 

Fish.  10(4): 1 7 1-1 76.  (In  Japanese.  English  summary.) 
OiESTAD,  v.,  AND  E.  MoKSNESs.    1981.    Study  of  growth  and  survival 

of  herring  larvae  (Chipea  harengiis)  using  plastic  bag  and  concrete 

basin  enclosures.  Rapp.  P.  V.  Reun.  Cons.  Int.  Explor.  Mer  178: 

144-149. 
OiokDA,  Y.    1960.   Studies  on  the  freshwater  fishes  of  Japan.  II.  Special 

part.  J.  Fac.  Fish.  Prefect.  Univ.  Mie  4(2):267-588. 
,  AND  R.  Semshi.     1938.    Studies  on  the  early  life  history  of  9 

species  of  freshwater  fishes  of  Japan.  Bull.  Biogeogr.  Soc.  Jpn.  8(15): 

223-263. 
Okamura,  O.     1970a.    Studies  on  the  macrouroid  fishes  of  Japan  — 

morphology,  ecology  and  phylogeny.  Rep.  Usa  Mar.  Biol.  Stn.  17(1- 

2):1-I79. 
.    1970b.   Fauna  Japonica:  Macrourina  (Pisces).  Academic  Press 

Japan. 
Okivama,  M.    1963.    Larvae  and  young  of  the  witch  flounder,  Glyp- 

locephalus  stelleri  (Schmidt)  at  metamorphosis  stages.  Bull.  Jpn. 

Sea  Reg.  Fish.  Res.  Lab.  1  1:101-108. 
.     1964.    Early  life  history  of  the  Japanese  blanquillos,  Brart- 

chioslegus japoniciis japonicus (Hounuyn).  Bull.  Jap.  Sea  Reg.  Fish. 

Res.  Lab.  13:1-14.  (In  Japanese,  English  abstract.) 
.    1967.   Study  on  the  early  life  history  of  a  flounder.  Paralichthys 

olivaceus  (Temminck  et  Schlegel).  I.  Descnptions  of  postlarvae. 

Bull.  Jpn.  Sea  Reg.  Fish.  Res.  Lab.  17:1-12. 
.    1970.    Studies  on  the  early  life  history  of  the  rainbow  runner, 

Elagalis  bipinnulatus  (Quoy  and  Gaimard)  in  the  Indo-Pacific 

Oceans.  Bull.  Far  Seas  Fish.  Res.  Lab.  (Shimizu)  3:167-186. 
.    1971.    Early  life  history  of  the  gonostomatid  fish.  Maurolicus 

muellen  (Gmelin),  in  the  Japan  Sea.  Bull.  Jpn.  Sea  Reg.  Fish.  Res. 

Lab  23:21-53. 


726 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


.    1972.    Morphology  and  identification  of  the  young  ipnopid. 

"Macristiella"  from  the  tropical  western  Pacific.  Jpn.  J.  Ichthyol. 
19:145-153. 

— .  1974a.  Studieson  the  early  life  history  of  a  flounder,  Para/;W;- 
ihys  olivaceus  (Temminck  et  Schlegel).  II.  Descriptions  of  juveniles 
and  the  comparison  with  those  of  the  related  species.  Bull.  Jpn.  Sea 
Reg.  Fish.  Res.  Lab  25:39-61. 

.    1974b.    The  larval  taxonomy  of  the  primitive  myctophiform 

fishes,  p.  609-621.  In:  The  early  life  history  offish.  J.  H.  S.  Blaxter 
(ed.).  Springer- Verlag,  New  York. 

.    1979a.    Manuals  for  the  larval  fish  taxonomy.  ( 1 )  Definition 

and  classification  of  larval  stages.  Kaiyo  to  Seibutsu  (Ocean  and  Its 
Life)  1:54-59.  (In  Japanese.) 

.    1979b.   Larval  morphology  and  interrelationships  of  the  myc- 

tophoid  fishes.  Kaiyo  Kaguku  (Mar.  Sci.)  11:131-145.  (In  Jap- 
anese.) 

.    1980.    Manuals  for  the  larval  fish  taxonomy  (5).  Diversity  of 

Myctophiformes  larvae.  Aquabiology  2:124-129. 

.  1981.  A  larval  Ipnops  and  its  possible  metamorphosing  pro- 
cess. Jpn.  J.  Ichthyol.  28:247-253. 

— .     1982a.    Manuals  for  the  larval  fish  taxonomy  (9).  Percoid 


larvae  and  spination.  Oceans  Mammals  4(2):92-99.  (In  Japanese, 

English  abstract.) 
.    1982b.   Embryoof  ,-l//o/ep«/;o//a«(y/ Jordan  and  Hubbs,  1925. 

Aquabiology  4(4):cover. 
,  AND  H.  Sando.    1976.   Early  life  history  of  sea  raven.  Bull.  Jap. 

Sea  Reg.  Fish.  Res.  Lab.  27:1-10. 
,  AND  K.  Takahashi.    1976.   Larval  stages  of  the  right  eye  floun- 
ders (subfamily  Pleuronectinae)  occurring  in  the  Japan  Sea.  Bull. 

Jpn.  Sea  Reg.  Fish.  Res.  Lab.  27:1 1-34. 
,  AND  S.  Ueyan.agi.     1977.    Larvae  and  juvenile  of  the  Indo- 

Pacific  dogtooth  tuna,  Gymnosarda  umcolor  (Ruppell).  Bull.  Far 

Seas  Fish.  Res.  Lab.  (Shimizu)  15:35-49. 
,  AND .    1978.    Interrelationships  of  scombroid  fishes:  an 

aspect  from  larval  morphology.  Bull.  Far  Seas  Fish.  Res.  Lab. 

(Shimizu)  16:103-113. 
Olney,  J.  E.,  AND  G.  C.  Grant.    1976.    Early  planktonic  larvae  of  the 

blackcheek  tonguefish,  Symphurus  plagiusa  (Pisces:  Cynoglossi- 

dae),  in  the  lower  Chesapeake  Bay.  Chesapeake  Sci.  1 7(4):229-237. 
,  AND  D.  F.  Markle.     1979.    Description  and  occurrence  of 

vexillifer  larvae  of  Echiodon  (Pisces:  Carapidae)  in  the  western 

Norih  Atlantic  and  notes  on  other  carapid  vexilHfers.  Bull.  Mar. 

Sci.  29:365-379. 
,  AND  A.  Naplin.    1 980.    Eggs  of  the  scalloped  ribbon-fish,  Zu 

cristatus,  (Pisces,  Trachipteridae)  in  the  western  North  Atlantic. 

Copeia  1:165-166. 
-,  .^ND  G.  R.  Sedberrv.    In  press.    Dragonet  larvae  (Teleostei, 


Callionymidae)  in  continental  shelf  waters  off  the  eastern  United 

States.  Biol.  Oceanogr. 
Olsen,  v.,  AND  D.  Merriman.    1946.   Studies  on  the  marine  resources 

of  southern  New  England.  IV.  The  biology  and  economic  impor- 
tance of  the  ocean  pout,  Macrozoarces  amencanus  (Bloch  and 

Schneider).  Bull.  Bingham  Oceanogr.  Collect.  Yale  LJniv.  9(4):  1- 

184. 
Orcutt,  H.G.    1950.  The  life  history  of  the  starry  flounder,  P/ar/f/!//;,r5 

stellatus  (Pallas).  Calif  Dep.  Fish  Game  Fish  Bull.  78. 
Orellana,  M.  C,  AND  F.  Balbontin.    1983.    Estudio  comparativo  de 

las  larvas  de  Clupeiformes  de  la  costa  de  Chile.  Rev.  Biol.  Mar. 

(Chile)  19:1-46. 
Orton,  G.  L.    1953a.    Development  and  migrations  of  pigment  cells 

in  some  teleost  fishes.  J.  Morphol.  93:69-99. 
.    1953b.   The  svstematics  of  vertebrate  larvae.  Syst.  Zool.  2:63- 

75. 
.    1955a.   Separation  of  eggs  of  synenthognath  and  allotriognath 

fishes  in  early  embryonic  stages.  Calif  Fish  Game  41:103-105. 
.    1955b.  Early  developmental  series  of  the  California  barracuda, 

Sphyraena  argentea  Girard.  Calif  Fish  Game  41:167-176. 
.     1955c.    The  role  of  ontogeny  in  svstematics  and  evolution. 

Evolution  9:75-83. 


— .  1955d.  Early  developmental  stages  of  the  California  scor- 
pionfish,  Scorpaena  guttata.  Copeia  1955:210-214. 

— .  1957.  The  bearing  of  larval  evolution  on  some  problems  in 
frog  classification.  Syst.  Zool.  6:285-292. 

— .     1964.    The  eggs  of  scomberesocid  fishes.  Copeia  1964:144- 


150. 
O'TociLE.  M.  J.    1977.    Development  and  distnbutional  ecology  of  the 

larvae  of  the  west  coast  sole  Auslroglossus  microlepis.  Fish.  Bull. 

S.  Afr.  9:32-45. 
,  AND  D.  P.  F.  King.     1974.    Early  development  of  the  round 

herring  Etrumeus  teres  (De  Kay)  from  the  southeast  Atlantic.  Vie 

Milieu  24:443-452. 
Owen,  R.     1866.    On  the  anatomy  of  the  vertebrates.  Vol.  1.  Fishes 

and  Reptiles.  Longman  Green.  London. 
Owen,  R.  W.     1980.    Patterning  of  flow  and  organisms  in  the  larval 

anchovy  environment.  lOC  Workshop  Rep.  No.  28:167-200. 
OzAWA,  T.    1976.   Early  life  history  of  the  gonostomatid  fish,  Pollichthys 

mauli.  in  the  oceanic  region  off  southern  Japan.  Jpn.  J.  Ichthyol. 

23:43-54. 
.    1978.    Notosudid  fish  larvae  in  the  ocean  off  southern  Japan. 

Jpn.  J.  Ichthyol.  25:79-88. 
.    1983.    Study  on  the  early  life  history  and  the  distribution  of 

oceanic  fish  larvae  in  the  ocean  ofl"  southern  Japan.  Unpubl.  PhD 

dissertation.  Kyushu  Univ. 
,  AND  S.  Matslh,     1979.    First  record  of  the  schindlerid  fish, 

Schmdleria  praematura.  from  southern  Japan  and  the  south  China 

Sea.  Jpn.  J.  Ichthyol.  25:283-285. 
Padmanabhan,  K.  G.    1961.   The  early  development  of  Sotenostomus 

cyanopterus.  Bull.  Cent.  Res.  Inst.  Trivandrum  80:1-13. 
Padoa,  E.     1947.    Forme  post-larvali  e  giovanili  di  Carapus.  Pubbl. 

Staz,  Zool.  Napoli  20:1 1 1-120. 
.    1956a.    Famiglia  4:  Trichiuridae.  p.  508-513.  In:  Uova,  larve 

e  stadi  giovanili  di  Teleostei.  Fauna  Flora  Golfo  Napoli  Monogr. 

38. 
.    1956b.    Famiglia  10-13:  Stromateidae,  Centrolophidae.  No- 

meidae,  Tetragonuridae,  p.  534-547.  In:  Uova,  larvc,  e  stadi  giov- 
anili di  Teleostei.  Fauna  Flora  Golfo  Napoli  Monogr.  38. 
.    1956c.   Divisione:  Carangiformes,  p.  548-572. /«.■  Uova  larve 

e  stadi  giovanili  di  Telostei.  Fauna  Flora  Golfo  Napoli  Monogr. 

38. 
.    1956d.    Sottordine:  Chromides,  p.  573-576.  In:  Uova,  larve 

e  stadi  giovanili  di  Teleostei.  Fauna  Flora  Golfo  Napoli  Monogr. 

38. 
.    1956e.    Triglidae,  Peristediidae,  Dactylopteridae,  Gobiidae, 

Echneidae,  Jugulares,  Gobiesocidae,  Heterosomata,  Pediculati.  p. 

627-888.  In:  Uova.  larve,  e  stadi  giovanili  di  Teleostei.  Fauna  Flora 

Golfo  Napoli  Monogr.  38. 
.     1956f    Famiglia:  Gobiidae,  p.  648-678.  In:  Uova,  larva  e 


stadi  giovanili  de  Teleostei.  Fauna  Flora  Golfo  Napoli  Monogr.  38. 
.    I956g.    Famiglia  1:  Trachinidae,  p.  687-697.  //;.  Uova.  larvae 

e  stadi  giovanili  di  Teleostei.  Fauna  Flora  Golfo  Napoli  Monogr. 

38. 
.    1956h.    Famiglia  6:  Clinidae,  p.  745-752.  In:  Uova,  larve  e 

stadi  giovanili  di  Teleostei.  Fauna  Flora  Golfo  Napoli  Monogr.  38. 
.    1956i.    Famiglia  8:  Ophidiidae,  p.  755-761.  In:  Uova,  larve 

e  stadi  giovanili  di  Teleostei.  Fauna  Flora  Golfo  Napoli  Monogr. 

38. 
.    1956j.    Famiglia  9:  Carapidae.  p.  761-774.  In:  Uova,  larve  e 

stadi  giovanili  di  Teleostei.  Fauna  Flora  Golfo  Napoli  Monogr.  38. 
.     1956k.    Heterosomata,  p.  783-877.  In:  Uova,  larve  e  stadi 

giovanili  di  Teleostei.  Fauna  Flora  Golfo  Napoli  Monogr.  38. 
Pal,  R.  N.,  and  H.  A.  Khan.    1969.    Breeding  and  development  of 

Heteropneustes  fossdis  (Bloch).  Proc.  Indian  Sci.  Congr.  56:543- 

544. 
Palko,  B.  J.,  G.  L.  Beardsley  and  W.  J.  Richards.    1981.   Synopsis 

of  the  biology  of  the  swordfish,  Xiphias  gladius  Linnaeus.  NOAA 

Tech.  Rep.  NMFS  Circ.  441. 
,  AND  W.  J.  Richards.    1969.   The  rearing  of  cowfish.  Salt  water 

Aquar.  Mag.  5(3):67-70. 


LITERATURE  CITED 


727 


Palmer.  G.  1961.  The  dealfishes  (Trachipteridae)  of  the  Mediterra- 
nean and  northeast  Atlantic.  Bull.  Br.  Mus.  (Nat.  Hist.)  Zool. 
7:335-351. 

.      1973.     Lamprididae.    Regalecidae,    Trachiptendae,    Radii- 

cephalidae,  Lophotidae,  p.  328-334.  In.  S.  C.  HureauandT.  Monod 
(eds.).  Check-list  of  the  fishes  of  the  north-eastern  Atlantic  and  of 
the  Mediterranean  (Clofnam).  Vol.  I.  UNESCO.  Paris. 

Palomera.  I.,  AND  J.  M.  FoRTLiNO.  1981.  Larvas  Icptocefalas  dc  peces 
anguilliformes  en  la  costa  noroccidental  de  Africa.  Res.  Exped. 
Cient.  9:3-15. 

.  AND  P.  Rubies.     1977.    Descnpcion  de  huevos  y  larvas  de 

Microchirus  ocellatus  y  M.  aze\ia  (Pleuronectiformes.  Soleidae)  de 
las  costas  del  NW  de  Africa.  Res.  Exped.  Cient.  B/O  Comide 
6:211-220. 

Panchen,  a.  L.  1982.  The  use  of  parsimony  in  testing  phylogenetic 
hypotheses.  Zool.  J.  Linn.  Soc.  74:305-328. 

Pannella,  G.  1971.  Fish  otoliths:  daily  growth  layers  and  penodical 
patterns.  Science  (Wash.  D.C.)  173:1124-1127. 

.    1980.  Growth  patterns  in  fish  sagittae,  p.  519-560. /«.  Skeletal 

growth  of  aquatic  organisms;  biological  records  of  environmental 
change.  D.  C.  Rhoads  and  R.  A.  Lutz  (eds.).  Plenum  Press,  New 
York. 

Papadopoulu,  C,  G.  D.  Kanias  and  E.  Moraitopoulou-Kassimati. 
1978.  Zinc  content  in  otoliths  of  mackerel  from  the  Aegean.  Mar. 
Pollut.  Bull.  9:106-108. 

, and .  1980.  Trace  element  content  in  fish  oto- 
liths in  relation  to  age  and  size.  Mar.  Pollut.  Bull.  1 1:68-72. 

Parenti.  L.  R.  1981.  A  phylogenetic  and  biogeographic  analysis  of 
cyprinodontiform  fishes  (Teleostei,  Atherinomorpha).  Bull.  Am. 
Mus.  Nat.  Hist.  168:335-557. 

.  1984.  On  the  relationships  of  phallostethid  fishes  (Atherino- 
morpha), with  notes  on  the  anatomy  of  Phalloslethus  dunckeri  Re- 
gan, 1913.  Am.  Mus.  Novit.  2779. 

Parin,  N.  V.  1961.  The  basis  of  classification  of  the  flying  fishes 
(families  Oxyporhamphidae,  Exocoetidae).  Tr.  Inst.  Okeanol.  Akad. 
NaukSSSR  43:92-183. 

.    1967.    Review  of  the  marine  Belonidae  of  the  western  Pacific 

and  Indian  oceans.  Tr.  Inst.  Okean.  Akad.  Nauk  SSSR  84:3-83. 
(In  Russian:  Natl.  Mar.  Fish.  Serv.  Syst.  Lab.  Transl.  No.  68.) 

.    1968.   Scomberesocidae  (Pisces,  Synentognathi)  of  the  eastern 

Atlantic  Ocean.  Atl.  Rep.  10:275-290. 

.    1976.    Two  new  species  of  Benthodesmus  (Trichiuridae,  Os- 

teichthyes)  from  the  western  Tropical  Pacific  Ocean.  Tr.  Inst. 
Okeanol.  Akad.  Nauk  SSSR  104:191-194.  (In  Russian,  English 
abstract.) 

.    1977.    New  species  of  A/e/aA!Oi?;^wa  (Zoarcidae,  Osteichthy- 

es),  M.  inexpeclatum,  from  the  western  equatorial  Pacific.  Tr.  Inst. 
Okeanol.  Acad.  Nauk  SSSR  107:63-67.  ^ 

.    1978.    New  records  of  midwater  fishes  from  off  New  Guinea 

and  Tonga  Islands  with  descriptions  of  two  new  species  of  Eusto- 
mias  (Family  Melanostomiatidae)  and  Benthodesmus  (Family 
Trichiundae).  Tr.  Inst.  Okeanol.  Akad.  Nauk  SSSR  111:156-168. 
(In  Russian,  English  Abstract.) 

,  and  D.  A.  AsTAKHov.    1982.  Studies  on  the  acoustico-lateralis 

system  of  beloniform  fishes  in  connection  with  their  systematics. 
Copeia  1982:276-291. 

,  and  V.  E.  Bekker.    1972.    Materials  on  taxonomy  and  distn- 

bution  of  some  trichiuroid  fishes  (Pisces,  Trichiuroidae:  Scombro- 
labracidae,  Gempylidae.  Trichiuridae).  Proc.  P.  P.  Shirshov  Inst. 
Oceanol.  93:1  10-204  (In  Russian,  English  abstract.) 

,  and  T.  N.  Belvanina.    1972.    New  data  on  the  distribution 


and  morphology  offish  of  the  genus  Macnstwlla.  J.  Ichthyol.  (Engl. 
Transl.  Vopr.  Ikhtiol.)  12:1018-1022. 

,  B.  B.  CoLLETTE  and  Yu.  N.  Shcherbachev.    1 980.   Prehminary 

review  of  the  marine  halfbeaks  (Hemiramphidac,  Beloniformes)  of 
the  tropical  Indo-West  Pacific.  Tr.  Inst.  Okean.  Akad.  Nauk  SSSR 
97:7-173.  (In  Russian,  English  abstract.) 

,  and  N.  N.  Gorbunova.  1964.  On  the  reproduction  and  de- 
velopment of  some  synentognalhs  (Pisces,  Beloniformes)  from  the 


Indian  Ocean  (based  on  material  from  the  R/S  "Vityaz").  Tr.  Inst. 
Okean.  Akad.  Nauk  SSSR  73:224-234. 

,  AND  Y.  I.  KuKUiEv.    1983.    Re-establishment  of  the  validity 

oi Lampns  tmmaculata  Gilchrist  and  the  geographical  distribution 
of  Lampridae.  J.  Ichthyol.  (Engl.  Transl.  Vopr.  Ikhtiol.)  23(1):1- 
12. 

,  AND  S.  Mekhailin.    1981.    A  new  species  of  cutlassfish.  Le/j;- 

dopus  dubnis  Parin  and  Mikhailin  (Trichiuridae),  from  the  eastern 
tropical  Atlantic  Ocean.  J.  Ichthyol.  (Engl.  Transl.  Vopr.  Ikhtiol.) 
21(3):l-8. 

,  and  G.  N.  Pokhil'skaya.  1968.  The  age  variability  and  dis- 
tribution of  a  rare  oceanic  fish  Eumectchthys  fiski  (Pisces,  Lophoti- 
dae). Probl.  Ichthyol.  (Engl.  Transl.  Vopr.  Ikhtiol.)  8:808-812. 

Parker,  G.  H.  1903.  The  optic  chiasma  in  teleosts  and  its  bearing  on 
the  asymmetry  of  the  Heterosomata  (flatfishes).  Bull.  Mus.  Comp. 
Zool.  Harv.  40:221-242. 

Parker.  K.  R.  1980.  A  direct  method  for  estimating  norihem  ancho- 
vy, Engraulis  inordax,  spawning  biomass.  Fish.  Bull.  U.S.  78:541- 
544. 

Parr,  A.  E.  1926.  Investigations  on  the  Cyclopterini.  Bergens  Mus. 
Aarbok  1924-25,  Naturvid.  Raekke  (7). 

.     1927.    The  stomiatoid  fishes  of  the  suborder  Gymnophoto- 

dermi  (Astronesthidae,  Melanostomiatidae,  Idiacanthidae)  with  a 
complete  review  of  the  species.  Bull.  Bingham  Oceanogr.  Collect. 
Yale  Univ.  3. 

.     1933.     Deep-sea  Berycomorphi  and  Percomorphi  from  the 

waters  around  the  Bahama  and  Bermuda  Islands.  Bull.  Bingham 
Oceanogr.  Collect.  Yale  Univ.  3(6). 

.    1946.    The  Macroundae  of  the  western  North  Atlantic  and 


Central  American  seas.  Bull.  Bingham  Oceanogr.  Collect.  Yale  Univ. 
10  (Art.  l):l-99. 

Pati,  S.  1979.  On  the  maturation  and  spawning  of  Chinese  pomfret 
Pampus  chinensis  (Euphrasen)  from  Orissa  coast.  Indian  J.  Fish. 
26:150-162. 

Patterson,  C.  1968.  The  caudal  skeleton  in  mesozoic  acanthopterygi- 
an  fishes.  Bull.  British  Mus.  (Nat.  Hist.)  Geol.  17:5-102. 

.    1981a.   Agassiz,  Darwin,  Huxley,  and  the  fossil  record  of  te- 

leost  fishes.  Bull.  British  Mus.  (Nat.  Hist.)  Geol.  35:213-224. 

.  1981b.  Significance  of  fossils  in  determining  evolutionary  re- 
lationships. Annu.  Rev.  Ecol.  Syst.  12:195-223. 

.    1982.    Morphological  characters  and  homology. /n.  Problems 

of  Phylogenetic  Reconstruction.  K.  A.  Joysey  and  A.  E.  Friday 
(eds.).  Proceedings  of  a  symposium,  Cambridge,  England,  April 
1980.  Published  for  the  Systematics  Association  by  Academic  Press, 
New  York.  Syst.  Assoc.  Spec.  Vol.  21. 

,  AND  D.  E.  Rosen.    1977.   Review  of  ichthyodectiform  and  other 


Mesozoic  teleost  fishes  and  the  theory  and  practice  of  classifying 

fossils.  Bull.  Am.  Mus.  Nat.  Hist.  158:83-172. 
Paulin,  C.  D.    1983.    A  revision  of  the  family  Moridae  (Pisces:  Ana- 

canthini)  within  the  New  Zealand  region.  Natl.  Mus.  N.Z.  Rec.  2(9): 

81-126. 
Paxton,  J.  R.     1967.    A  distributional  analysis  for  the  lantemfishes 

(family  Myctophidae)  of  the  San  Pedro  Basin.  California.  Copeia 

1967:422-440. 
.    1972.   Osteology  and  relationships  of  the  lantemfishes  (family 

Myctophidae).  Nat.  Hist.  Mus.  Los  Ang.  Cty.  Sci.  Bull.  13:1-81. 
.    1979.    Nominal  genera  and  species  of  lantemfishes  (family 

Myctophidae).  Contrib.  Sci.  (Los  Ang.)  322. 
Pearcy,  W.  G.    1962.    Egg  masses  and  early  developmental  stages  of 

the  scorpaenid  fish,  Sebastolobus.  J.  Fish.  Res.  Board  Can.  19:11 69- 

1173. 
Pearse,  a.  G.  E.    1968.    Histochemistry  — theory  and  applied.  Vol.  2. 

Little  Brown,  Boston. 
Pearson,  J.  C.     1938.    Life  history  of  the  striped  bass,  or  rockfish, 

Roccus  saxalilis  (Walbaum).  Bull.  Bur.  Fish.  Wash.  49(28):825- 

851. 
.    1941.    The  young  of  some  marine  fishes  taken  in  lower  Ches- 
apeake Bay.  Virginia,  with  special  reference  to  the  gray  sea  trout 


728 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Cynoscion  regalis  (B\och).  U.S.  Fish  Wildl.  Serv.  Fish.  Bull.  50:79- 

102. 
Peden,  a.  E.     1970.    A  new  cottid  fish,  Nautichthys  robustus.  from 

Alaska  and  British  Columbia.  Natl.  Mus.  Nat.  Sci.  (Ottawa)  Publ. 

Biol.  Oceanogr.  2. 
.    1981.    On  the  identity  of  two  cottid  subspecies, /cW;/u« /)!» - 

chami  burchami  and  I.  h.  fuscescens.  Copeia  1981:482-484. 
,  AND  M.  E.  Anderson.    1978.    A  systematic  review  of  the  fish 

genus  Lycodapus  (Zoarcidae)  with  descriptions  of  two  new  species. 

Can.  J.  Zool.  56:1925-1961. 
,  AND .    1981.    Lycodapus  (Pisces:  Zoarcidae)  of  eastern 

Bering  Sea  and  nearby  Pacific  Ocean,  with  three  new  species  and 

a  revised  key  to  the  species.  Can.  J.  Zool.  59:667-678. 
,  AND  C.  A.  CoRBETT.    1973.   Commensalism  between  a  liparid 

fish,  Careproclus  sp.  and  the  lithodid  box  crab,  Liphotilhodes  for- 

aminatus.  Can.  J.  Zool.  51:555-556. 
Pellegrin.J.    1907.   Liste  des  poissons  recueillis  a  Madagascar  par  M. 

F.  Geay.  Description  d'une  espece  nouvelle.  Bull.  Mus.  Paris  1907: 

201-206. 
.    1914.   Sur  un  atherinide  nouveau  de  Madagascar  appartenant 

au  genre  Bedotia.  Bull.  Soc.  Zool.  Fr.  39:178-180. 
.    1935.    Les  poissons  africains  de  la  famille  des  cromeriides  et 

leurs  affinites.  Arch.  Mus.,  Ser.  6,  12:461-463. 
Penaz,  M.    1968.    Development  of  the  chub,  LpWfKcwi  fcp/!a/(«  (Lin- 
naeus 1758)  m  the  posthatching  penod.  Zool.  Listy  1 7(3):269-278. 
.    1975.   Early  development  of  the  graylmg  Thymallus  ihymallus 

(Linnaeus,  1758).  Acta.  Sci.  Nat.  Acad.  Sci.  Bohemoslov.  Brno. 

9:1-35. 
.  andJ.  GajdCisek.    1979.  Early  development  of  bream,. -l/iranNi 


brama.  from  the  water  reservoir  Mostiste,  Czechoslovakia.  Folia 
Zool.  28(4):347-360. 

Penrith,  M.  J.  1965.  A  new  species  of  flat  fish,  Mancopsella  mitjordi. 
from  South  Africa,  with  notes  on  the  genus  Mancopsella.  Ann.  S. 
Afr.  Mus.  48:181-188. 

Penrith,  M.-L.  1969.  The  systematics  of  the  fishes  of  the  family 
Clinidae  in  South  Africa.  Ann.  S.  Afr.  Mus.  55:1-121. 

Peppar,  J.  L.  1965.  Some  features  of  the  life  history  of  cockscomb 
prickleback,  Anoplarchus  purpurescens.  Unpubl.  MS  thesis.  Univ. 
British  Columbia,  Vancouver. 

Perlmutter,  a.  1939.  An  ecological  survey  of  young  fish  and  eggs 
identified  from  two-net  collections,  p.  11-71.  In:  A  biological  sur- 
vey of  the  salt  waters  of  Long  Island.  Suppl.  28th  Ann.  Rep.,  1938, 
New  York  State  Cons.  Dep.  Albany,  No.  15,  Pt.  2,  Sect.  1. 

,  L.  BoGRAD  and  J.  J.  Pruginin.    1957.    Use  of  the  estuarine 

and  sea  fish  of  the  family  Mugilidae  (grey  mullets)  for  pond  culture 
in  Israel.  Proc.  Fish.  Counc.  Mediterr.,  FAG  4:289-304. 

Pertseva,  T.  A.  1936.  Key  to  pelagic  eggs  of  fishes  of  the  Barents  Sea. 
Pischepromizdat,  Moscow. 

Pertseva-Ostrolimova,  T.  A.  1954.  Materials  on  the  development 
of  far  eastern  flounders  (Pleuroneclidae).  I.  Development  of  the 
yellowfin  so\e-Linianda  aspera  Pallas.  Tr.  Inst.  Okeanol.  Acad.  Nauk. 
SSSR  11:221-232. 

.  1961.  The  reproduction  and  development  of  far  eastern  floun- 
ders. Tr.  Inst.  Okeanol.  Akad.  Nauk.  SSSR.  (In  Russian;  transl. 
avail.  Fish.  Res.  Bd.  Canada,  Transl.  Serv.  No.  856,  1967.) 

.    1964.  Some  morphological  characteristics  of  myctophid  larvae 

(Myctophidae,  Pisces).  In:  Fishes  of  the  Pacific  and  Indian  Oceans, 
biology  and  distribution.  T.  S.  Rass  (ed.).  Tr.  Inst.  Okeanol.  Akad. 
Nauk  SSSR  73:79-97.  (In  Russian,  English  transl.,  U.S.  Dep.  Com- 
merce 65-5120.) 

.  1965.  Ratfish  larvae  from  the  Gulf  of  Tonkin.  Tr.  Inst.  Okean- 
ol. Acad.  Nauk.  SSSR  80:177-220.  (In  Russian.) 

.    1967.    har\a\  ^\a%es  oi  s,ome  Protomyctophum  and  Eleclrona 

species  (Pisces,  Myctophidae).  Tr.  Inst.  Okeanol.  Akad.  Nauk  SSSR, 
84:222-236.  (In  Russian,  English  transl.  Bur.  Commer.  Fish., 
Southwest  Fish.  Ctr.,  La  Jolla.) 

.    1971.   Unusual  flounder  larvae  (Pleuronectiformes,  Pisces).  J. 

Ichthyol.  (Engl.  Transl.  Vopr.  Ikhtiol.)  1 1:817-822. 


.     1972.    The  larvae  of  lantemfish  (Myctophidae)  collected  by 

Australian  Expeditions  on  H.M.A.S.  Gascoyne  and  H.M.A.S.  Dia- 
mantina.  J.  Ichthyol.  (Engl.  Transl.  Vopr.  Ikhtiol.)  12:634-643. 

.    1974.    New  data  on  lanternfish  larvae  (Myctophidae,  Pisces) 

with  oval  eyes  from  the  Indian  and  Pacific  Oceans.  Tr.  Inst.  Okean- 
ol. Akad.  Nauk  SSSR  96:77-142. 

.  1975.  Larvae  of  the  lanternfish,  D;u/)/h<5  a,?a55/r;  (Myctophi- 
dae, Pisces).  J.  Ichthyol.  (Engl.  Transl.  Vopr.  Ikhtiol.)  15:606-610. 

.    1977.   On  the  development  of  some  species  of  the  genus  Gv"i- 

no.sco/)e/i«(Myctophidae).  J.  Ichthyol.  (Engl.  Transl.  Vopr.  Ikhtiol.) 
17:58-65. 

,  AND  T.  S.  Rass.    1973.    Ichthyoplankton  of  the  southeastern 

Pacific  Ocean.  Trudy  Inst.  Okeanol.  Acad.  Nauk  SSSR  94:7-70. 

Peters,  K.  M.  1981.  Reproductive  biology  and  developmental  os- 
teology of  the  Rorida  blenny.  Chasmodes  saburrae  (Perciformes: 
Blenniidae).  Northeast  Gulf  Sci.  4(2):79-98. 

Peters,  N.  1963.  Embryonale  Anpassungen  oviparerZahnkarpfenaus 
periodisch  austrocknenden  Gewassem.  Int.  Rev.  Gesamlen  Hy- 
drobiol.  48(2):257-313. 

Petersen,  C.  G.  J.  1904.  On  the  larval  and  postlarval  stages  of  the 
long  rough  dab  and  the  genus  Pleuronectes.  Medd.  Komm.  Ha- 
vunders.  Ser.  Fiskeri  1(1). 

.     1905.    On  the  young  stages  of  the  genus  Zeugopterus.  Rep. 

Dan.  Biol.  Stn.  12:25-32. 

.    1906.    On  the  larval  and  post-larval  stages  of  some  Pleuro- 


nectidae  (Pleuronectes.  Zeugopterus).  Medd.  Komm.  Havunders. 
Ser.  Fiskeri  2(1). 

.    1909.    On  the  larval  and  post-larval  stages  of  some  Pleuro- 

nectidae  (Zeugopterus.  .irnoglossus.  Solea).  Medd.  Komm.  Ha- 
vunders., Ser.  Fiskeri  3  (1):1-I8. 

.    1917.    On  the  development  of  our  common  gobies  (G"ofei«) 

from  the  egg  to  the  adult  stages  etc.  Rep.  Dan.  Biol.  Stn.  24:5-16. 

.    1919.   Our  gobies  (Gobiidae).  From  the  egg  to  the  adult  stages. 

Rep.  Dan.  Biol.  Stn.  25:45-66. 

Peterson,  C.  L.  1956.  Observations  on  the  taxonomy,  biology,  and 
ecology  of  the  engraulid  and  clupeid  fishes  in  the  Gulf  of  Nicoya, 
Costa  Rica.  Bull.  Inter-Am.  Trop.  Tuna  Comm.  Bull.  1:139-280. 

Phillips,  J.  B.  1942.  Osteology  of  the  sardine  (Sardinops  caerulea).  J. 
Morphol.  70:463-500. 

Phonlor,  G.  1978.  Morfologia  e  biologia  dos  ovos  de  Engraulidae  do 
Sul  do  Brasil  (Osteichthyes,  Engraulidae).  Unpubl.  thesis.  Univ. 
Sao  Paulo,  Brazil. 

.     1979.    Fecundafao  artificial  e  desenvolvimento  dos  ovos  e 

larvas  de  Parana  signala  em  laboratorio  (Osteichthyes,  Carangi- 
dae).  Atl.  Rio  Grand,  3:47-56. 

PiCQUELLE.S.  J.,  andR.  P.  Hewitt.  1983.  Northern  anchovy  spawning 
biomass  for  the  1982-83  California  fishing  season.  Calif  Coop. 
Oceanic  Fish.  Invest.  Rep.  24. 

Pietri,  R.  1983.  An  ultrastructural  study  of  the  chorion  and  its  fila- 
ments in  Oryzias  latipes.  Unpubl.  MS  thesis.  Rutgers.  State  Univ. 
New  Jersey,  New  Brunswick. 

Pietsch,  T.  W.  1972.  A  review  of  the  monotypic  deepsea  anglerfish 
family  Centrophrynidae:  taxonomy,  distribution,  and  osteology. 
Copeia  1972:17-47. 

.    1974.   Osteology  and  relationships  of  deep-sea  anglerfishes  of 

the  family  Oneirodidae  with  a  review  of  the  genus  Oneirodes  Liit- 
ken.  Nat.  Hist.  Mus.  Los  Ang.  Cty.  Sci.  Bull.  18:1-113. 

.     1975.    Systematics  and  distribution  of  ceratioid  anglerfishes 

ofthe  genus  Chaenophryne (iamWy  Oneirodidae).  Bull.  Mus.  Comp. 
Zool.  Harv.  Univ.  147(2):75-I00. 

.  1976.  Dimorphism,  parasitism  and  sex:  Reproductive  strat- 
egies among  deepsea  ceratioid  anglerfishes.  Copeia  1976:781-793. 

.     1978a.    The  feeding  mechanisms  of  Slylephorus  chordalus 


(Teleostei:  Lampriformcs):  functional  and  ecological  implications. 
Copeia  1978:255-262. 
— .    1978b.   Evolutionary  relationships  ofthe  sea  moths  (Teleostei: 
Pegasidae)  with  a  classification  of  gasterosteiform  families.  Copeia 
1978:517-529. 


LITERATURE  CITED 


729 


.    1979.  Systematics  and  distribution  of  the  ceratioidanglerfishes 

of  the  family  Caulophrynidae.  Contrib.  Sci.  (Los  Ang.)  310. 
.    1981.    The  osteology  and  relationships  of  the  anglerfish  genus 

Telrabrachium  with  comments  on  lophiiform  classification.  U.S. 

Fish.  Bull.  79:387-419. 
.    1984.  The  genera  of  frogfishes  (family  Antennariidae).  Copeia 

1984:27^4. 
,  AND  D.  B.  Grobecker.    1980.    Parental  care  as  an  alternative 

reproductive  mode  in  an  antennariid  anglerfish.  Copeia  1980:551- 

553. 
,  AND .    In  press.    Frogfishes  of  the  world:  Systematics, 

zoogeography  and  behavioral  ecology.  Stanford  University  Press. 

Stanford,  CA. 
PlVNlCKA,  K.    1970.   Morphological  variation  in  the  burbot  (Lora /o/a) 

and  recognition  of  the  subspecies:  a  review.  J.  Fish.  Res.  Board 

Can.  27:1757-1765. 
Platt,  C,  and  a.  N.  Popper.    1981.    Fine  structure  and  function  of 

the  ear,  p.  3-36.  In:  Hearing  and  Sound  Communication  in  Fishes. 

W.  N.  Tavolga.  A.  N.  Popper  and  R.  R.  Fay  (eds.).  Springer-Verlag, 

New  York. 
Podosinnikov,  a.  Y.     1977.    Early  ontogeny  of  the  "stnped  grunt." 

Paraprislipoma  humile  (Pomadasyidae.  Pisces).  J.  Ichlhyol.  (Engl. 

Transl.  Vopr.  Ikhtiol.)  17:683-685. 
Poll,  M.    1953.    Poissons.  III.  Teleosteens  Malacopterygiens.  Result. 

Scient.  Exped.  Oceanogr.  Beige  Eaux  Cot.  Afr.  Atlant.  Sud.  4(2). 
.    1964.    Une  famille  dulcicole  nouvelle  de  poissons  Africains: 

les  Congothrissidae.  Acad.  R.  Sci.  Outre-Mer.  Mem.  N.  S.  15:1- 

40. 
Ponomareva,  L.  A.    1949.    Eggs  and  fry  of  fishes  of  the  Kara  Sea.  Tr. 

Uses.  Nauchno-Issled.  Inst.  Morsk.  Rybn.  Khoz.  Okeanogr.   17: 

189-205. 
Popper,  A.  N.,  and  S.  Coombs.    1980a.   Acoustic  detection  by  fish,  p. 

403-430.  In:  Environmental  Physiology  of  Fish.  M.  A.  Ali  (ed.). 

Plenum  Publ.  Corp.,  New  York. 
,  and  S.  Coombs.    1980b.  Auditory  mechanisms  in  teleost  fishes. 

Am.  Sci.  68:429-440. 
PtippER,  D.,  H.  Gordon  and  G.  W.  Kjssil.    1973.    Fertilization  and 

hatching  of  rabbitfish  Siganus  rivulatus.  Aquaculture  2:37-44. 
Poss,  S.  G.    1975.   Infraorbital  bones  in  cottid  and  hexagrammid  fishes. 

Unpubl.  MA  thesis.  San  Francisco  State  Univ.,  San  Francisco,  CA. 
.    1982.   Taxonomy  and  systematics  of  the  velvetfishes  (Pisces: 

Aploactinidae).  Unpubl.  PhD  dissertation.  Univ.  Michigan,  Ann 

Arbor. 
.  and  W.  N.  Eschmeier.     1975.    The  Indo-Wesl  Pacific  scor- 


pionfish  genus  Ocosia  Jordan  and  Starks  (Scorpaenidae,  Tetraro- 
ginae),  with  descriptions  of  three  new  species.  Matsya  1:1-18. 

,  AND .     1978.    Two  new  Australian  velvetfishes,  genus 

Paraploactis  (Scorpaeniformes:  Aploactinidae),  with  a  revision  of 
the  genus  and  comments  on  the  genera  and  species  of  the  Aploac- 
tinidae. Proc.  Calif  Acad.  Sci.  41:401-426. 

Post,  A.  1976.  Ergebnisse  der  Forschungreisen  des  FFS  "Walther 
Herwig"  nach  Siidamenka.  XLII.  Direlmus  Johnson.  1863  (Beryci- 
formes,  Berycoidei,  Diretmidae).  2.  Morphologie,  Entwicklung, 
Verbreitung.  Arch.  Fischereiwiss.  26(2/31:87-114. 

,  andJ.  QiiERO.    1981.   Revision  des  Diretmidae  (Pisces,  Trach- 

ichthyoidei)  de  L'Atlantique  avec  description  d'un  nouveau  genre 
et  d'une  nouvelle  espece.  Cybium,  3rd  Ser.  5(l):33-60. 

PosTUMA.  K.  H.  1974.  The  nucleus  of  the  herring  otolith  as  a  racial 
character.  J.  Cons.  Int.  Explor.  Mer  35:121-129. 

POTTHOFF,  T.  1974.  Osteological  development  and  variation  in  young 
tunas,  genus  Thunnus  (Pisces,  Scombridae)  from  the  .Atlantic  Ocean. 
Fish.  Bull.  U.S.  72:563-588. 

.    1975.    Development  and  structure  of  the  caudal  complex,  the 

vertebral  column,  and  the  pterygiophores  in  the  blackfin  tuna  I'hun- 
nus  atlanlicus  (Pisces.  Scombndae).  Bull.  Mar.  Sci.  25:205-231. 

.    1980.    Development  and  structure  of  fins  and  fin  supports  in 

dolphin  fishes  Corvphaena  hippuru.^  and  Corvphaena  equisetis 
(Coryphaenidae).  Fish.  Bull.  U.S.  78:277-312. 


,  AND  S.  ICellev.    1982.    Development  of  the  vertebral  column, 

fins  and  fin  supports,  branchiostegal  rays,  and  squamation  in  the 

swordfish,  Xiphias  gladius.  Fish.  Bull.  U.S.  80:161-186. 
, andJ.Javech.    MS.   Comparisons  of  the  development 

of  the  vertebral  column,  fins  and  fin  supports  in  scombroid  fishes. 
, ,  M.  MOE  and  F.  Young.    1984.   Description  of  porkfish 

larvae  (.-inisotremus  virginicus.  Haemulidae)  and  their  osteological 

development.  Bull.  Mar.  Sci.  34:21-59. 
,  W.  J.  Richards  and  S.  Ueyanagi.     1980.    Development  of 

Scombrolabrax  helerolepis  (Pisces,  Scombrolabracidae)  and  com- 
ments on  familial  relationships.  Bull.  Mar.  Sci.  30:329-357. 
PowLES,  H.    1977.    Description  of  larval  Jenkinsia  lamprolaema  (Clu- 

peidae,  Dussumieriinae)  and  their  distribution  off  Barbados,  West 

Indies.  Bull.  Mar.  Sci.  27:788-801. 
.    1980.   Descriptions  of  larval  silver  perch,  Ba;>(//e//ac/irv50Mra. 

banded  drum,  Lanmus  fascialus,  and  star  drum,  Steltifer  lanceo- 

latus  (Sciaenidae).  Fish.  Bull.  U.  S.  78:1 19-136. 
,  and  B.  W.  Stender.    1978.   Taxonomic  data  on  the  early  life 

history  stages  of  Sciaenidae  of  the  South  Atlantic  Bight  of  the  United 

States.  S.  C.  Mar.  Resour.  Cent.  Tech.  Rep.  31. 
Premalatha,  P.    1977.    A  study  of  the  development  and  distribution 

of  the  larvae  of  leather  skin  Chonnemus  sanctipetn  (Cuv.  &  Val.) 

(Carangidae-Pisces)  along  the  southwest  coast  of  India.  Proc.  Symp. 

warm  water  zooplankton  Spec.  Publ.  UNESCO/NIO:450-459. 
Presley,  R.  F.     1970.    Larval  snowy  grouper,  Epincphetus  niveatus 

(Valenciennes,  1828),  from  the  Florida  straits,  Fla.  Dep.  Nat.  Re- 
sour.  Mar.  Res.  Lab.  Leafl.  Ser.  4(18). 
PROKis,  M..  and  M.  Pen.az.    1980.    Early  development  of  the  chub. 

Leuciscus  cephahis.  Acta  Sci.  Nat.  Acad.  Sci.  Bohemoslov  Brno 

14(7):  1-40. 
Qasim,  S.  Z.    1956.   The  spawning  habits  and  embryonic  development 

of  the  shanny  (Blenmus  pholis  L.).  Proc.  Zool.  Soc.  Lond.  127:79- 

93. 
QuAST,  J.  C.     1965.    Osteological  characteristics  and  affinities  of  the 

hexagrammid  fishes,  with  a  synopsis.  Proc.  Calif  Acad.  Sci.  31: 

563-600. 
QuERo,  J.    1978.    Repartition  des  Zeides  (Pisces,  Zeiformes)  captures 

dans  I'Atlantique  oriental  entre  le  Cap  Vert  (15°N)  et  le  Cap  Juby 

(28°N)  au  cours  des  campagnes  de  la  "Thalassa."  Bull.  Off  Natl. 

Pech.  Tunis.  1978,  2(l-2):49-6l. 
.    1979.   Remarques  sur  le  Grammicolepis  hrachiusculus  (Pisces, 

Zeiformes,  Grammicolepididae)  observe  au  port  de  la  Rochelle. 

Ann.  Soc.  Sci.  Nat.  Charente-Marit.  6:573-576. 
QiTNN,  R.  A..  AND  C.  C.  SiGL  (eds.).    1980.    Radiography  in  modem 

industry.  Eastman  Kodak  Co.,  Rochester,  NY. 
Rabalais,  N.  N.,  S.  C.  Rabalais  andC.  R.  Arnold.    1980.  Description 

of  eggs  and  larvae  of  laboratory-reared  red  snapper  (Lutjanus  cam- 

pechanus).  Copeia  1980:704-708. 
Rachow,  a.    1924.    Uber  Cypnnodon  iherus  Valenciennes  und  Fun- 

dulus  hispanicus  (Valenciennes).  Blatt  Aquar.-Terrarienk.  22(21): 

369-371. 
Radtke,  R.  L.    1984.    Otolith  formation  and  increment  deposition  in 

laboratory  reared  skipjack  tuna,  Kalsuwomis  pelanus  larvae.  NOA.^ 

Tech.  Rep.  In  Press. 
,  AND  J.  M.  Dean.    1982.    Increment  formation  in  the  otoliths 

of  embryos,  larvae  and  juveniles  of  the  mummichog,  Fundulu.'i 

heterodilus.  Fish.  Bull.  U.S.  80:201-216. 
,  AND  K.  G.  Waiwood.     1980.    Otolith  formation  and  body 


shrinkage  due  to  fixation  in  larval  cod  (Gadus  morhua).  Can.  Tech. 

Rep.  Fish.  Aquat.  Sci.  929. 
Raffaele,  F.    1888.    Le  uova  galleggianti  e  le  larve  dei  Teleostei  nel 

golfo  di  Napoli.  Mitt.  Zool.  Sta.  Neapel  8. 
R  AjAGOPAL,  P.  K.    1 979.   The  embryonic  development  and  the  thermal 

effects  on  the  development  of  the  mountain  whitefish,  Frnsopium 

williamsoni  (Girand).  J.  Fish  Biol.  15:153-158. 
Raju,  S.  N.    1974.    Three  new  species  of  the  genus  A/ono.gnaf/iK.';  and 

the  leptocephali  of  the  order  Saccopharyngiformes.  Fish  Bull.  U.S. 

72:547-562. 


730 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Randall,  J.  E.  1955a.  An  analysis  of  the  genera  of  surgeon  fishes 
(family  Acanthuridae).  Pac.  Sci.  9:359-367. 

.    1955b.   A  revision  of  the  surgeonfish  genera  Ze/iraiOHia  and 

Paracanthurus.  Pac.  Sci.  9:396-412. 

.  1 955c.  A  revision  of  the  surgeonfish  genus  Ctenochaetus.  fam- 
ily Acanthuridae,  with  descriptions  of  five  new  species.  Zoologica 
(N.Y.)  40:149-166. 

.    1956.    A  revision  of  the  surgeon  fish  genus  Acanthurus.  Pac. 

Sci.  10:159-235. 

.    1961.  A  contribution  to  the  biology  of  the  convict  surgeonfish 

of  the  Hawaiian  Islands,  Acanthurus  triostegus  sandvicensis.  Pac. 
Sci.  15:215-272. 

.    1980.    Revision  of  the  fish  genus  Plectranthias  (Serranidae: 

Anthiinae)  with  descriptions  of  13  new  species.  Micronesica  16: 
101-187. 

.    1981.    Underwater  guide  to  Hawaiian  reef  fishes.  Harrowood 

Books,  Newton  Square,  PA. 

,  AND  R.  Lubbock.    1981.    A  revision  of  the  serranid  fishes  of 

the  subgenus  Mirolabhchthys  (Anthiinae:  Anthias),  with  descrip- 
tions of  five  new  species.  Contrib.  Sci.  (Los  Ang.)  333. 

,  AND  J.  E.  McCosKER.    1975.   The  eels  of  Easter  Island  with  a 


description  of  a  new  moray.  Contrib.  Sci.  (Los  Ang.)  264. 

— ,  AND  H.  Randall.  1963.  The  spawning  and  early  development 
of  the  Atlantic  parrot  fish.  Sparisoma  rubripinne,  with  notes  on 
other  scarid  and  labrid  fishes.  Zoologica  (N.Y.)  48(2):49-60. 

— ,  T.  Shimizu  AND  T.  Yamakawa.    1982.   A  revision  of  the  Hol- 


ocentrid  fish  genus  Ostichthys,  with  descriptions  of  four  new  species 

and  a  related  new  genus.  Jpn.  J.  Ichthyol.  29:1-26. 
Ranzi,  S.    1933.    Famiglia  3:  Sparidae,  p.  332-382.  In:  Uova,  larve  e 

stadi  giovanili  di  Teleostei.  Fauna  Flora  Golfo  Napoli  Monogr.  38. 
Rao,  K.  S.    1973.    Eggs  and  early  developmental  stages  of  Hilsa  kelee 

(Cuvier).  Indian  J.  Fish.  20:250-255. 
Rao,  R.  T.    1977.    Enhancement  of  natural  populations  of  moi  (Po/r- 

dactylus  sextilis)  in  Hawaii  through  the  release  of  hatchery-reared 

juveniles— a  feasibility  study  of  sea  ranching.  Hawaii  Inst.  Mar. 

Biol.  Tech.  Rep.  33. 
Rao,  V.  V.    1970.   Breeding  habits  and  early  development  of  two  blen- 

niid  fishes  Omobranchus  japontcus  (Bleeker)  and  Cruantus  smithi 

(Visweswara  Rao)  from  Godavari  Estuary.  J.  Mar.  Biol.  Assoc. 

India  12(1&2):175-182. 
Rapson,  a.  M.    1940.    The  reproduction,  growth,  and  distnbution  of 

the  lemon  soles  [Pelotretis  ftavilatus  Waite)  of  Tasman  Bay  and 

Marlborough  Sounds.  N.Z.  Mar.  Dep.  Fish.  Bull.  7:1-56. 
Rasquin,  p.    1958.   Ovarian  morphology  and  early  embryology  of  the 

pediculate  fishes  Anlennarius  and  Histno.  Bull.  Am.  Mus.  Nat. 

Hist.  1 14:327-372. 
Rass,  T.  S.    1949.   Composition  of  the  ichthyoplankton  of  the  Barents 

Sea.  In:  T.  S.  Rass,  I.  I.  Kazanova,  S.  P.  Alekseyeva  and  L.  A. 

Ponomareva  (eds.).  Material  on  the  reproduction  and  development 

of  fishes  of  the  Northern  Seas.  Tr.  Vses.  Nauchno-Issled.  Inst.  Morsk. 

Rybn.  Khoz.  Okeanogr.  17:1-68. 
.    1950.    An  unusual  instance  of  a  biological  relation  between 

fish  and  crabs.  Prioroda  39(7):68-69. 
.    1954.    Contribution  to  the  study  of  Pacific  Ocean  Moridae 

(Pisces,  Gadiformes).  Tr.  Inst.  Okeanol.  Akad.  Nauk.  SSSR  1 1:56- 

61. 
.    1 960.   On  the  geographical  distribution  of  bathypelagic  fishes 

of  the  family  Myctophidae  in  the  Pacific  Ocean.  Tr.  Inst.  Okean. 

Akad.  Nauk  SSSR  41:146-152.  (In  Russian.) 
.    1968.    Spawning  and  development  of  polar  cod.  Rapp.  P.  V. 

Reun.  Cons.  Perm.  Int.  Explor.  Mer  158:135-137. 
.    1972.    Ichthyoplankton  and  the  deep-sea  fishes  of  the  world 

ocean.  Trans.  P.  P.  Shirshov.  Inst.  Oceanol.  Acad.  Sci.  USSR  93: 

34-35.  (In  Russian.) 
.    1973.    Systematics  of  pelagic  fish  eggs  and  larvae  (Ichthyo- 
plankton). Unpubl.  manuscript.  FAO,  Rome. 
Rathbun,  R.    1893.    Report  upon  the  inquiry  respecting  food  fishes 

and  the  fishing  grounds.  Rep.  U.S.  Comm.  Fish.  17:97-171. 


Ray,  D.    1950.    The  peripheral  nervous  system  of  Lampanyctus  leu- 

copsarus.  J.  Morphol.  87:61-178. 
Re,  P.    1980.   The  eggs  and  newly  hatched  larvae  of /liu(^e/(/M//!^nrf«i 

(Cuvier,  1 830)  (Pisces:  Pomacentridae)  from  the  Azores.  Arq.  Mus. 

Bocage,  2nd  Ser.  7(8):109-1 16. 
Reay,  P.  J.    1970.   Synopsis  of  biological  data  on  North  Atlantic  sand- 
eels  of  the  genus  Ammodytes.  A.  lobianus,  A.  dubius.  A.  aniericanus. 

and  A.  mannus.  FAO  Fish.  Synop.  82,  Rome. 
Regan,  C.  T.    1902.   On  the  systematic  position  of  Luvarus  imperialis 

Rafinesque.  Ann.  Mag.  Nat.  Hist.,  Ser.  7,  10:278-281. 
.    1903a.    Descriptions  de  poissons  nouveaux  faisant  partie  de 

la  collection  du  Musee  d'histoire  naturelle  de  Geneve.  Rev.  Suisse 

Zool.  11:413-418. 
.    1903b.    LXII.  — On  the  systematic  position  and  classification 

of  the  gadoid  or  anacanthine  fishes.  Ann.  Mag.  Nat.  Hist.,  Ser.  7, 

11:459-466. 
-.    1907.   On  the  anatomy,  classification  and  systematic  position 

of  the  teleostean  fishes  of  the  suborder  Allotriognathi.  Proc.  Zool. 

Soc.  Lond.  1907:634-643. 
.     1909.    On  the  anatomy  and  classification  of  the  scombroid 

fishes.  Ann.  Mag.  Nat.  Hist.,  Ser.  8,  3:66-75. 
.    1910.   The  origin  and  evolution  of  the  teleostean  fishes  of  the 

order  Heterosomata.  Ann.  Mag.  Nat.  Hist.,  Ser.  8,  6:484-496. 
.    1911a.   The  anatomy  and  classification  of  the  teleostean  fishes 

of  the  order  Iniomi.  Ann.  Mag.  Nat.  Hist.,  Ser.  8,  7:120-133. 
.    1911b.   The  classification  of  the  teleostean  fishes  of  the  order 

Synentognathi.  Ann.  Mag.  Nat.  Hist.,  Ser.  8,  7:327-335. 
.    1911c.   The  osteology  and  classification  of  the  gobioid  fishes. 

Ann.  Mag.  Nat.  Hist.,  Ser.  8,  8:729-733. 
.    1912a.   The  classification  of  the  teleostean  fishes  of  the  order 

Pediculati.  Ann.  Mag.  Nat.  Hist..  Ser.  8,  9:277-289. 
.    1912b.    The  classification  of  the  blennioid  fishes.  Ann.  Mag. 

Nat.  Hist.,  Ser.  8,  10:265-280. 
.    1912c.  The  osteology  and  classification  of  the  teleostean  fishes 

of  the  order  Apodes.  Ann.  Mag.  Nat.  Hist.,  Ser.  8,  10:377-387. 
.    1913a.   Osteology  and  classification  of  the  teleostean  fishes  of 

the  order  Scleroparei.  Ann.  Mag.  Nat.  Hist.,  Ser.  8,  1 1:169-184. 
.    1913b.   On  the  classification  of  the  percoid  fishes.  Ann.  Mag. 

Nat.  Hist.,  Ser.  8,  12:111-145. 
.    1 9 1 3c.   Phallostethus  dunckert.  a  remarkable  new  cyprinodont 

fish  from  Johore.  Ann.  Mag.  Nat.  Hist.,  Ser.  8,  12:548-555. 
.    1913d.   The  Antarctic  fishes  of  the  Scottish  National  Antarctic 

Expedition.  Trans.  R.  Soc.  Edinb.  49(2):229-292. 
.    1916.    Larval  and  post-larval  fishes.  Nat.  Hist.  Rep.  Br.  Ant- 
Terra  Nova  Exped.  Zool.  1:125-156. 

1923.    The  classification  of  the  stomiatoid  fishes.  Ann,  Mag. 
Hist.,  Ser.  9,  11:612-614. 
.     1924.    The  morphology  of  a  rare  oceanic  fish,  Stylophorus 

chordatus  Shaw;  based  on  specimens  collected  in  the  Atlantic  by 

the  "Dana"  expeditions,  1920-1922.  Proc.  R.  Soc.  Lond.  96:193- 

207. 
.    1925.   III.  The  fishes  of  the  genus  G/gflWwra,  A.  Brauer;  based 

on  the  specimens  collected  in  the  Atlantic  by  the  "Dana"  Expe- 
ditions, 1920-22.  Ann.  Mag.  Nat.  Hist.  15(9):53-59. 
.     1926.    The  pediculate  fishes  of  the  suborder  Ceratioidea. 

Oceanogr.  Rep.  Dana  Exped.  2. 

.    1929.   Fishes.  Encyclopaedia  Britannica,  14th  ed.  9:305-328. 

,  AND  E.  Trewavas.    1929.   The  fishes  of  the  families  Astrones- 

thidae  and  Chauliodontidae.  Dana-Rep.  Carlsberg  Found.  5. 
,  AND .    1930.    The  fishes  of  the  families  Stomiidae  and 

Malacosteidae.  Dana-Rep.  Carlsberg  Found.  6. 
.AND .    1932.  Deep-sea  angler-fishes  (Ceratioidea).  Dana- 
Rep.  Carlsberg  Found.  2. 
Renzhai,  Z.,  AND  L.  SuiFEN.    1980a.    Studies  on  the  eggs  and  larvae 

development  of  wise  porgy,  Gymnocranius  griseus  (Temminck  et 

Schlegel).  Acta  Zool.  Sin.  26(2):  132-1 35.  (In  Chinese,  English  ab- 
stract.) 
,  and .    1980b.    On  the  eggs  and  larvae  of  the  golden- 


arct. 


Nat. 


LITERATURE  CITED 


731 


thread  Nemiptenis  virgalus  (Houttuyn).  Acta  Zool.  26(2):  136- 139. 

In  Chinese.  English  abstract.) 
Reshetnikov,  Yu.  S.     1975.    The  systematic  position  of  whitefishes. 

Zool.  Zh.  54:1651-1671.  (In  Russian.) 
Richards,  S.  W.    1965.   Description  of  the  posllarvae  of  the  sand  lance 

(Ammodyles)  from  the  east  coast  of  North  America.  J.  Fish.  Res. 

Board  Can.  22:1313-1317. 
.    1982.   Aspects  of  the  biology  of  Ammodyles  americanus  from 

the  St.  Lawrence  River  to  Chesapeake  Bay,  1972-75,  including  a 

comparison  of  the  Long  Island  Sound  postlarvae  with  Ammodytes 

dubius.  J.  Northwest  Atl.  Fish.  Sci.  3:93-104. 
Richards,  W.  J.     1969.    Elopoid  leptocephali  from  Angolan  waters. 

Copeia  1969:515-518. 
.    1974.   Evaluation  of  identification  methods  for  young  billfish- 

es,  p.  62-72.  In:  Proc.  Intematl.  Billfish  Symp.,  Kailua-Kona,  Ha- 
waii, 9-12  August  1972,  Part  2.  Review  and  contributed  papers. 

R.  S.  Shomura  and  F.  Williams  (eds.). 
.    1981.    Kinds  and  abundances  of  fish  larvae  in  the  Caribbean 

Sea.  Rapp.  P.  V.  Reun.  Cons.  Int.  Explor.  Mer  178:240-241. 
.    1984.    Kinds  and  abundances  of  fish  larvae  in  the  Caribbean 

Sea  and  adjacent  areas.  NOAA  Tech.  Rep.  NMFS  SSRF-776. 
,  AND  G.  R.  Dove.    1971.    Internal  development  of  young  tunas 

of  the  genera  Katsuwonus.  Euthynnus.  Auxis.  and  Thunnus  (P'xscti, 

Scombndae).  Copeia  1971:72-78. 
,  AND  W.  L.  Klawe.    1972.    Indexed  bibliography  of  the  eggs 

and  young  of  tunas  and  other  scombrids  (Pisces,  Scombridae)  1 880- 

1970.  NOAA  Tech.  Rep.  NMFS  SSRF-652. 
,  R.  V.  Miller  and  E.  D.  Houde.    1974.    Eggs  and  larval  de- 
velopment of  the  Atlantic  thread  herring,  Opisthonema  ogtimun. 

Fish.  Bull.  U.S.  72:1123-1136. 
,  AND  T.  PoTTHOFF.    1974.  Analysis  of  the  taxonomic  characters 

of  young  scombrid  fishes,  genus  Thunnus.  p.  623-648.  In:  The 

early  life  history  of  fish.  J.  H.  S.  Blaxter  (ed.).  Springer- Verlag, 

Berlin. 
-,  AND  V.  P.  Saksena.     1980.    Description  of  larvae  and  early 


juveniles  of  laboratory-reared  gray  snapper,  Luljanus  griseus  (Lin- 
naeus) (Pisces,  Lutjanidae).  Bull.  Mar.  Sci.  30:515-521. 

Richardson,  S.  L.  1981a.  Current  knowledge  of  larvae  of  sculpins 
(Pisces:  Cottidae  and  allies)  in  northeast  Pacific  genera  with  notes 
on  intergeneric  relationships.  Fish.  Bull.  U.S.  79:103-121. 

.    1981b.    Pelagic  eggs  and  larvae  of  the  deepsea  sole,  Embas- 

sichthys  bathybius  (Pisces:  Pleuronectidae),  with  comments  on  ge- 
neric affinities.  Fish.  Bull.  U.S.  79:163-170. 

.    1981c.    Current  knowledge  of  northeast  Pacific  sculpin  larvae 

(Pisces:  Cottidae  and  allies)  with  notes  on  relationships  within  the 
group.  Rapp.  P.  V.  Reun.  Cons.  Int.  Explor.  Mer  178:603-604. 

,  AND  D.  A.  Dehart.    1975.    Records  of  larval,  transforming, 

and  adult  specimens  of  the  quillfish,  PtiUchlhys  goodei  from  waters 
off  Oregon.  Fish.  Bull.  U.S.  73:681-684. 

,  J.  R.  Dunn  andN.  A.  Naplin.    1980.  Eggs  and  larvae  of  butter 

sole,  Isopsetta  isolepis  (Pleuronectidae),  off  Oregon  and  Washing- 
ton. Fish.  Bull.  U.S.  78:401-417. 

,  and  E.  B.  Joseph.    1973.    Larvae  and  young  of  western  North 

Atlantic  bothid  flatfishes  Etropus  microslomiis  and  Cilhanchlhys 
arctifrons  in  the  Chesapeake  Bight.  Fish.  Bull.  U.S.  71:735-767. 

,  and  W.  A.  Laroche.    1979.    Development  and  occurrence  of 

larvae  and  juveniles  of  the  rockfishes,  Sebastes  crameri.  Sebastes 
pinniger  and  Sebasles  helvomaculalus  (Family  Scorpaenidae)  off 
Oregon.  Fish.  Bull.  U.S.  77:1-41. 

,  and  W.  G.  Pearcy.    1977.   Coastal  and  oceanic  fish  larvae  in 

an  area  of  upwelling  off  Yaquina  Bay,  Oregon.  Fish.  Bull.  U.S.  75: 
125-145. 

,  AND  B.  B.  Washington.    1 980.   Guide  to  identification  of  some 


sculpin  (Cottidae)  larvae  from  marine  and  brackish  waters  off  Or- 
egon and  adjacent  areas  in  the  northeast  Pacific.  NOAA  Tech.  Rep. 
NMFS  Circ.  430. 
Riehl,  R.    1980.    Micropyle  of  some  salmonins  and  coregonins.  En- 
viron. Biol.  Fishes  5:59-66. 


RiLE'i,  J.  D.    1966.    Marine  fish  culture  in  Britain,  VII.  Plaice  {Pteu- 

ronectes plalessa  L.)  post-larval  feeding  on  Anemia  salina  L.  napulii 

and  the  effects  of  varying  feeding  levels.  J.  Cons.  Int.  Explor.  Mer 

30:204-221. 
RiTTER,  W.  E.,  AND  S.  E.  Bailey.    1908.   On  the  weight  of  developing 

eggs.  Univ.  Calif  Publ.  Zool.  6:1-10. 
RivAS,  L.  R.     1972.    Opisthonema  captivai.  a  new  western  Atlantic 

clupeid  fish  from  Columbia.  Copeia  1972:1-3. 
,  AND  S.  M.  Warlen.     1967.    Systematics  and  biology  of  the 

bonefish,  Albula  nemoptera  (Fowler).  U.S.  Fish  Wildl.  Serv.  Fish. 

Bull.  66(2):25 1-258. 
Roberts,  R.  J.,  M.  Bell  and  H.  Young.    1973.    Studies  on  the  skin 

of  plaice  (Plenronectes  platessa  L.).  II.  The  development  of  larval 

plaice  skin.  J.  Fish  Biol.  5:103-108. 
Roberts,  T.  R.    1970.    Description,  osteology  and  relationships  of  the 

Amazonion  cyprinodont  fish  Fhiviphylax  pygmaeus  (Myers  and 

Carvalho).  Breviora  347:1-28. 
.    1971.  The  fishes  of  the  Malaysian  family  Phallostethidae(Ath- 

eriniformes).  Breviora  374:1-27. 
.     1981.    Sundasalangidae,  a  new  family  of  minute  freshwater 

salmoniform  fishes  from  southeast  Asia.  Proc.  Calif.  Acad.  Sci.  42: 

295-302. 
.    1982a.    The  southeast  Asian  freshwater  pufferfish  genus  Cho- 

nerhinos  (Tetraodontidae),  with  descnptions  of  new  species.  Proc. 

Calif  Acad.  Sci.  43:1-16. 
.    1982b.    Unculi  (homy  projections  arising  from  single  cells), 

an  adaptive  feature  of  the  epidermis  of  ostariophysan  fishes.  Zool. 

Scr.  ll(l):55-76. 
Robertson,  D.  A.    1973.    Planktonic  eggs  and  larvae  of  some  New 

Zealand  marine  teleosts.  Unpubl.  PhD  thesis.  Univ.  Otago,  Dune- 
din,  N.Z. 
.    1974.    Developmental  energetics  of  the  southern  pigfish  (Tele- 

ostei:  Congiopodidae).  N.Z.  J.  Mar.  Freshwater  Res.  8:61  1-620. 
.     1975a.    A  key  to  the  planktonic  eggs  of  some  New  Zealand 

marine  teleosts.  N.Z.  Fish.  Res.  Div.  Occas.  Publ.  No.  9. 
.    1975b.    Planktonic  stages  of  the  teleost  family  Carapidae  in 

eastern  New  Zealand  waters.  N.Z.  J.  Mar.  Freshwater  Res.  9:403- 

409. 
.     1976.    Planktonic  stages  of  Mauroticus  muelleri  (Teleostei: 


Stemoptychidae)  in  New  Zealand  waters.  N.Z.  J.  Mar.  Freshwater 

Res.  10:311-328. 
.    1977.   Planktonic  eggs  of  the  lantemfish  Lanzpawctoo'e.s /lec- 

Mw  (family  Myctophidae).  Deep-Sea  Res.  24:1-4. 
.    1978.    Spawning  of  the  tarakihi  (Pisces:  Cheilodactylidae)  in 

New  Zealand  waters.  N.Z.  J.  Mar.  Freshwater  Res.  12:277-286. 
.    1981.   Possible  functions  of  surface  structure  and  size  in  some 

planktonic  eggs  of  manne  fishes.  N.Z.  J.  Mar.  Freshwater  Res.  15: 

147-153. 
,  AND  S.  Mito.     1979.    Sea  surface  ichthyoplankton  off  south- 
eastern New  Zealand,  summer  1977-78.  N.Z.  J.  Mar.  Freshwater 

Res.  13:415-424. 
,  AND  U.  Raj.    1971.  Egg  and  yolk  sac  stages  of  the  sand  flounder 

(Teleostei:  Pleuronectidae).  N.Z.  J.  Mar.  Freshwater  Res.  5:404- 

414. 
,  P.  E.  Roberts  and  J.  B.  Wilson.    1978.    Mesopelagic  faunal 

transition  across  the  Subtropical  Convergence  east  of  New  Zealand. 

N.Z.  J.  Mar.  Freshwater  Res.  12:295-312. 
Robertson,  D.  R.    1973.    Field  observations  on  the  reproductive  be- 
haviour of  a  pomacentrid  fish,  Acanlhochrnmis  polyacanthus.  Z. 

Tierpsychol.  32:319-324. 
Robins,  C.  H.,  and  C.  R.  Robins.    1976.    New  genera  and  species  of 

dysommine  and  synaphobranchine  eels  (Synaphobranchidae)  with 

an  analysis  of  the  Dysomminae.  Proc.  Acad.  Nat.  Sci.  Phila.  127: 

249-280. 
Robins,  C.  R.    1958.    Studies  on  fishes  of  the  family  Ophidiidac.  I.  A 

new  species  of  Lepophidium  from  the  Caribbean  Sea.  Bull.  Mar. 

Sci.  Gulf  Caribb.  8:360-368. 


732 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


.    1959.    Studies  on  fishes  of  the  family  Ophidiidae.  III.  A  new 

species  of  Lepophidium  from  Barbados.  Breviora  104. 
— .    1960.    Studies  on  fishes  of  the  family  Ophidiidae.  V.  Lepo- 


phidium pherornystax,  a  new  Atlantic  species  allied  to  Lepophidium 
jeannae  Fowler.  Bull.  Mar.  Sci.  GulfCaribb.  10:83-95. 

.     1961.    Studies  on  fishes  of  the  family  Ophidiidae.  VI.  Two 

new  genera  and  a  new  species  from  American  waters.  Copeia  1961: 
212-221. 

.    1962.    Studies  on  fishes  of  the  family  Ophidiidae.  VII.  The 

Pacific  species  of  Lepophidium.  Copeia  1962:487-498. 

.     1967.    The  status  of  the  serranid  fish  Liopropoma  aberrans 

with  the  description  of  a  new,  apparently  related  genus.  Copeia 
1967:591-595. 

.     1974.    The  validity  and  status  of  the  roundscale  spearfish. 

Telrapturus  georgei.  p.  54-61.  In:  Proc.  Int.  BiUfish  Symposium 
Kailua-Kona,  Hawaii,  9-12  August  1972.  Part  2.  Review  and  Con- 
tributed Papers.  R.  S.  Shomuraand  F.  Williams  (eds.).  NOAATech. 
Rep.  NMFS  SSRF-675. 

,  R.  M.  Bailey,  C.  E.  Bond,  J.  R.  Brooker,  E.  A.  Lachner,  R. 

N.  Lea  AND  W.  B.  Scott.  1980.  A  list  of  common  and  scientific 
names  of  fishes  from  the  United  States  and  Canada  (Fourth  Edi- 
tion). Am.  Fish.  Soc.  Spec.  Publ.  12. 

,  AND  J.  E.  BoHLKE.  1959.  Studies  On  fishcs  of  the  family  Ophi- 
diidae. IV.  Two  new  dwarf  cusk-eels  (genus  Ophidian)  from  the 
tropical  western  Atlantic.  Not.  Nat.  (Phila.)  325. 

— ,  AND .    1970.  The  first  Atlantic  species  of  the  ammodytid 

fish  genus  Embolichlhys.  Not.  Nat.  (Phila.)  430:1-1  1. 

,  AND  D.  P.  DE  Sylva.    1960.    Description  and  relationships  of 


the  longbill  spearfish,  Tetrapierus  betone.  based  on  western  North 

Atlantic  specimens.  Bull.  Mar.  Sci.  Gulf  and  Caribb.  10:383-413. 
— ,  AND .    1965.    The  Kasidondae.  a  new  family  of  mira- 

pinniform  fishes  from  the  western  Atlantic  Ocean.  Bull.  Mar.  Sci. 

15:189-201. 
— ,  AND  R.  N.  Lea.    1978.    Fishes  of  the  family  Ophidiidae.  IX. 

Analysis  of  geographic  variation  in  Lepophidium  microlepis  with 

descriptions  of  two  new  subspecies.  Bull.  Mar.  Sci.  28:716-727. 
— ,  AND  J.  G.  Nielsen.    1 970.   Snydendia  bothrops,  a  new  tropical, 

amphi-Atlantic  species  (Pisces:  Carapidae).  Stud.  Trop.  Oceanogr. 

(Miami)  4(2):285-293. 
— ,  AND  J.  E.  Randall.    1965.    Three  new  western  Atlantic  fishes 


of  the  blennioid  genus  Chaenopsis.  with  notes  on  the  related  Lu- 
cayablennius  zingaro.  Proc.  Acad.  Nat.  Sci.  Phila.  117:213-277. 

RoFEN,R.R.  1958.  ThemarinefishesofRennell  Island,  p.  149-218. 
In:  The  natural  history  of  Rennell  Island  British  Solomon  Islands. 
Vol.  I.  Danish  Science  Press,  Ltd.,  Copenhagen. 

.    1966a.    Family  Paralepididae,  p.  205-461.  In:  Fishes  of  the 

western  North  Atlantic.  G.  W.  Mead  (ed.).  Mem.  Sears  Found. 
Mar.  Res.  No.  1  (Pt.  5). 

.     1966b.    Family  Omosudidae,  p.  462-481.  In:  Fishes  of  the 

western  North  Atlantic.  G.  W.  Mead  (ed.).  Mem.  Sears  Found. 
Mar.  Res.  No.  1  (Pt.  5). 

.    1966c.    Family  Anotopteridae,  p.  498-410. /«.  Fishes  of  the 

western  North  Atlantic.  G.  W.  Mead  (ed.).  Mem.  Sears  Found. 
Mar.  Res.  No.  1  (Pt.  5). 

.     1966d.    Family  Evermannellidae,  p.  511-565.  In:  Fishes  of 

the  western  North  Atlantic.  G.  W.  Mead  (ed.).  Mem.  Sears  Found. 
Mar.  Res.  No.  I  (Pt.  5). 

Rohde,  F.  C.  1 980.  Dallia  pectoralis.  p.  1 27.  In:  Atlas  of  North  Amer- 
ican freshwater  fishes.  D.  S.  Lee,  C.  R.  Gilbert,  C.  H.  Hocutt,  R. 
E.  Jenkins,  D.  E.  McAllister  and  J.  R.  Stauffer,  Jr.  (eds.).  North 
Carolina  State  Mus.  Nat.  Hist..  Raleigh,  NC. 

Rose,  J.  A.  1961.  Anatomy  and  sexual  dimorphism  of  the  swim  blad- 
der and  vertebral  column  in  Ophidian  holbrooki  (Pisces:  Ophidi- 
idae). Bull.  Mar.  Sci.  Gulf  Canbb.  1  1:280-308. 

Rosen,  D.  E.  1964.  The  relationships  and  taxonomic  position  of  the 
halfbeaks,  killifishes,  silversides,  and  their  relatives.  Bull.  Am.  Mus. 
Nat.  Hist.  127:217-268. 

.    1971.    The  Macristiidae,  a  ctenothrissiform  family  based  on 

juvenile  and  larval  scopelomorph  fishes.  Am.  Mus,  Nov.  2452. 


— .  1973.  Intertelationships  of  higher  euteleostean  fishes,  p.  397- 
513.  In:  Interrelationships  of  fishes.  P.  H.  Greenwood.  R.  S.  Miles, 
and  C,  Patterson  (eds.).  Zool.  J.  Linn.  Soc.  53  (Suppl.  1.). 

.    1974.   Phylogeny  and  zoogeography  of  salmoniform  fishes  and 

relationships  of  Lepidogala.xias  satamandroides.  Bull.  Am.  Mus. 
Nat.  Hist.  153:267-325. 

.  1978.  Vicariant  patterns  and  historical  explanation  in  bio- 
geography.  Syst.  Zool.  27:159-188. 

.    1982.    Teleostean  interrelationships,  morphological  function 

and  evolutionary  inference.  Am.  Zool.  22:261-273. 

.    1983.   Zeoidsare  tetraodontiforms,  acanthuridsare  not.  ASIH 

Program  &  Abstracts,  1983. 

.     1984.    Zeiforms  as  primitive  plectognath  fishes.  Am.  Mus. 

Novit.  2782. 

— ,  and  p.  H.  Greenwood.  1970.  Origin  of  the  Weberian  ap- 
paratus and  the  relationships  of  the  ostariophysan  and  gonoryn- 
chiform  fishes.  Am.  Mus.  Novit.  2428. 

— ,  AND  L.  R.  Parenti.    1981.    Relationships  of  On'r/fli.  and  the 


groups  of  atherinomorph  fishes.  Am.  Mus.  Novit.  2719. 
— ,  AND  C.  Patterson.    1969.   The  structure  and  relationships  of 


the  paracanthopterygian  fishes.  Bull.  Am.  Mus.  Nat.  Hist.  141:357- 

474. 
Rosenblatt,  R.  H.    1959.    A  revisionary  study  of  the  blennioid  fish 

family  Tripterygiidae.  Unpubl.  PhD  dissertation.  Univ.  Calif,  Los 

Angeles. 
.     1960.    The  Atlantic  species  of  the  blennioid  fish  genus  En- 

neanectes.  Proc.  Acad.  Nat.  Sci.  Phila.  1 12(1):  1-23. 
,  AND  M.  A.  Bell.    1976.    Osteology  and  relationships  of  the 

roosterfish,  Nematistius  pectoralis  Gill.  Contrib.  Sci.  (Los  Ang.) 

279:1-23. 
,  and  J.  L.  Butler.    1977.    The  ribbonfish  genus  Z)Mmorfe»ia, 

with  the  description  of  a  new  species  (Pisces,  Trachipteridae).  Fish. 

Bull.  U.S.  75:843-855. 
,  AND  G.  D.  Johnson.    1976.   Anatomical  considerations  of  pec- 
toral swimming  in  the  opah,  Lampns  gutlatus.  Copeia  1976:367- 

370. 
,  AND  T.  D.  Parr.    1969.    The  Pacific  species  of  the  clinid  fish 

genus  Paraclinus.  Copeia  1969:1-20. 
,  AND  J.  S.  Stephens.    1978.  Mccoskerichthys  sandae.  a  new  and 

unusual  chaenopsid  blenny  from  the  Pacific  coast  of  Panama  and 

Costa  Rica.  Contnb.  Sci.  (Los  Ang.)  293:1-21. 
,  AND  L.  R.  Tailor,  Jr.    1971.   The  Pacific  species  of  the  clinid 


fish  tribe  Starksiini.  Pac.  Sci.  25:436-463. 

Rosenthal,  H.  1969.  Verdauungsgeschwindigkeit,  Nahrungswahl  und 
Nahrungsbedarf  bei  den  Larven  des  Herings,  Clupea  harengus  L, 
Ber.  Dtsch.  Wiss.  Komm.  Meeresforsch.  20:60-69. 

,  and  G.  Hempel.    1970.    Experimental  studies  in  feeding  and 

food  requirements  of  herring  larvae  (Clupea  harengus  L.),  p.  344- 
364.  In:  Marine  food  chains.  J.  H.  Steele  (ed.).  Univ.  Calif  Press, 
Berkeley. 

Rostand,  J.  1964.  Biological  and  geological  sciences.  Chapt.  1.  Gen- 
eral biology.  In:  The  beginnings  of  modem  science.  Basic  Books, 
New  York. 

Roule,  L.  1922.  Dtscnplionde  Scombrolabrax  heterolepis  noy .  gen. 
nov.  sp.,  poisson  abyssal  nouveau  de  I'lle  Madere.  Bull.  Inst.  Ocean- 
ogr. (Monaco)  408. 

.    1924.    Etude  sur  Pontogenese  et  al  croissance  avec  hypermet- 

amorphose  de  Luvarus  imperialis  Raf  (poisson  rapporte  a  Tordre 
des  Scombriformes).  Ann.  Inst.  Oceanogr.  Monaco  (n.s.)  1:119- 
157. 

,  AND  F.  Angel.    1 930.   Larves  et  alevins  de  poissons  provenant 


des  croisieres  du  Prince  Albert  I  de  Monaco.  Result.  Camp.  Sci. 
Prince  Albert  I,  79:1-148. 

RoxAS,  H.  A.  1934.  A  review  of  Philippine  Mugilidae.  Philipp.  J.  Sci. 
54(3):393-429. 

Ruck,  J.  G.  1971.  Development  of  the  lumpfish,  Trachelochismus 
melobesia  (Pisces:  Gobiesocidae).  Zool.  Publ.  Victoria  Univ.  Wel- 
lington 57:1-9. 

.    1973a.   HeyeXopmexW.  of  Tripier\-gion  capito  2iTid  T.  robustum 


LITERATURE  CITED 


733 


(Pisces:  Tripterygiidae).  Zool.  Publ.  Victoria  Univ.  Wellington  63: 

1-10. 
.    1973b.    Developmeul  of  ihe  clingfishes,  Diplocrepis  puniceus 

and  Trachelochismiis  pmnii/alus  (Pisces:  Gobiesocidae).  2tool.  Publ. 

Victoria  Univ.  Wellington  64:1-12. 
.    1976.  Studies  on  the  development  and  osteology  of  some  New 

Zealand  inshore  fishes.  Unpubl.  PhD  dissertation.  Victoria  Univ. 

Wellington,  New  Zealand. 
.    1980.    Early  developmenl  of  Forslery-gion  vanum.  Gilloblen- 

nius  decemdigitatus.  and  G.  tnpennis  (Pisces:  Tripterygiidae).  N.  Z. 

J.  Mar.  Freshwater  Res.  14:313-326. 
RuDDELL,  C.  L.    In  press.   Initiating  polymerization  of  glycol  methac- 

rylate  with  cyclic  diketo  carbon  acids.  Stain  Technol. 
RuDOMETKiNA,  G.  P.   1980.   Larvae  oflizardfish  (family  Synodontidae) 

from  the  eastern  tropical  Atlantic.  J.  Ichthyol.  (Engl.  Transl.  Vopr. 

Ikhtiol.)  20:(3):147-152. 
.    1981.    Postembryonic  development  of  Diplophos  taenia  Gun- 

ther,  1873  (Gonostomatidae)  from  the  .Atlantic  Ocean.  J.  Ichthyol. 

(Engl.  Transl.  Vopr.  Ikhtiol.)  21(4):I06-1 13. 
RuNYAN,  S.    1961.   Early  development  of  clingfish,  Gobiesox  strumosus 

Cope.  Chesapeake  Sci.  2:113-141. 
RupLE,  D.   MS.   Larvae  of  some  gobioid  fishes  from  the  Western  North 

Atlantic. 
Russell,  B.  C.    1980.    Revision  of  the  fish  genus  Pseudolabrus  and 

allied  genera  (family  Labridae)  with  a  phylogenetic  analysis  of  re- 
lationships. Unpubl.  PhD  dissertation.  Macquarie  Univ.,  Sydney. 
Russell,  F.  S.    1976.   The  eggs  and  planktonic  stages  of  British  marine 

fishes.  Academic  Press,  London. 
Russo,J.  L.    1983.   Interrelationshipsof  the  gempylid  fishes  (Teleostei. 

Scombroidei).  Unpubl.  PhD  dissertation.  George  Washington  Univ., 

Washington,  D.C. 
RuTENBERG,  E.  P.    1962.    Obzor  ryb  semeistva  terpugovykh  (Hexa- 

grammidae)  (Survey  of  the  fishes  of  family  Hexagrammidae).  Tr. 

Inst.  Okeanol.  Akad.  Nauk  SSSR  59:3-100.  (Transl.  Isr.  Prog.  Sci. 

Transl.,  1970,  p.  1-103.  In:  Greenlings:  Taxonomy,  biology,  in- 

teroceanic  transplantation.  T.  S.  Rass  (ed).  Avail.:  NTIS.  Spring- 
field, VA;  TT69-55097.) 
Ryder,  J.  A.    1883.    Observations  on  the  absorption  of  the  yolk,  the 

food,  feeding,  and  development  of  embryo  fishes,  comprising  some 

investigations  conducted  at  the  central  hatchery.  Armory  Building, 

Washington  D.C.  in  1882.  Bull.  U.S.  Fish.  Comm.  2(  1882):  179- 

205. 
.    1884.    A  contribution  to  the  embryography  of  osseous  fishes, 

with  special  reference  to  the  cod(Gadus  morhua).  Rep.  U.S.  Comm. 

Fish.  10(1882):455-605. 
.    1885.    On  the  development  of  viviparous  osseous  fishes  and 

of  the  Atlantic  salmon.  Proc.  U.S.  Natl.  Mus.  8:128-162. 
.    1887.   On  the  development  of  osseoiis  fishes,  including  marine 

and  freshwater  forms.  Rep.  U.S.  Comm.  Fish.  13(1885):489-603. 
.    1888.    Development  of  the  sea-bass  (Serra/iiw  a?rani«).  Am. 


Nat.  22:755. 
Saf'Yanova,  T.  Y.,  and  N.  I.  Revina.    1960.   The  biology  of  the  large 

horse  mackerel  and  fishing  for  it.  Tr.  Azovo-Chemomorsk.  Nauch- 

no-Issled.  Inst.  Mosrsk.  Rybn.  Khoz.  Okeanogr.  18. 
Saioal,  B.  N.,  AND  M.  P.  MoTWANL    1962.   Studies  on  the  fishery  and 

biology  of  the  commercial  cat-fishes  of  the  Ganga  River  system.  1. 

Early  life-history,  bionomics  and  breeding  of  Mystus  (Osteobagrus) 

seenghala  Sykes.  Indian  J.  Fish.  8(I):60-74. 
Sakamoto,  T.,  AND  K.  Suzuki.    1978.    Spawning  behavior  and  early 

life  history  of  the  porcupine  puffer,  Diodon  holacanthus.  in  aquaria. 

Jpn.  J.  Ichthyol.  24:261-270. 
Saksena,  V.  P.,  AND  W.  J.  Richards.    1975.    Description  of  eggs  and 

larvae  of  laboratory-reared  white  grunt,  Haemulon  plumien  (La- 

cepede)  (Pisces:  Pomadasyidae).  Bull.  Mar.  Sci.  25:523-526. 
Saldanha,  L.,  and  N,  C.  Pereira.    1977.    Lophoius  lacepedei  Gioma, 

1809  (Pisces,  Lophotidae)  signale  pour  la  premiere  fois  sur  la  cote 

portugaise.  Arq.  Mus.  Bocage  6(4):51-59. 
Salekhova,  L.  p.    1959.  O  razvitii  lufarya  (Pomawmus  saltalrix  Linne). 


Tr.  Sevastop.  Biol.  Stn.  Im.  A.  D.  Kovalenskogo  Acad.  Nauk  Ukr. 

SSR  11:182-188. 
Sanchez  Delgacxj,  F.     1981.    Contribucion  al  conocimiento  de  los 

Labridos  (Familia  Labridae)  de  la  costas  Ibericas.  Parte  1.  Descrip- 

cion  de  las  especies.  Bol.  Inst.  Esp.  Oceanogr.  6(3):  19-57. 
Sandercock,  F.  K.,  and  N.  J.  Wilimovsky.    1968.    Revision  of  the 

cottid  genus  Enophrys.  Copeia  1968:832-853. 
Santander,  H.,  and  O,  S.  de  Castillo.    1969.    Desarrollo  y  distri- 

bucion  de  huevos  y  larvas  de  merluza,  Merluccius  gayi  (Guichenot) 

en  la  costa  peruana.  Bol.  Inst.  Mar  Peru  (Callao)  2(3):79-107. 
,  and .   1971.  Desarrollo  y  distribucion  de  huevos  y  larvas 

de  jurel  ( Trachums  symmetricus  murphyi  Nichols)  en  la  costa  peru- 
ana. Inst.  Mar  Peru  (Callao)  Inf.  36:1-22. 
Santerre,  M.  T.,  AND  R.  C.  May.    1977.   Some  effects  of  temperature 

on  laboratory-reared  eggs  and  larvae  of  Polydactylus  sextilis  (Pisces: 

Polynemidae).  Aquaculture  10:341-351. 
Sanzo,  L.   1911.  Distnbuzione  delle  papille  cutanee  (Organi  ciatiformi) 

e  suo  valore  sistematice  nei  Gobi.  Mitt.  Zool.  Sta.  Neapel  20:249- 

328. 
.    1912a.   Larva  di  S'rom/a.sioa  Risso.  Mem.  R.  Com.  Talassogr. 

Ital.  10. 
.    1912b.   Uova  di  "Stemoptychidae"  Gthr.  R.  Com.  Talassogr. 

Ital.  Boll.  Bimest.  19:384-386. 
.    19 1 5a.   Stadi  larvali  di  Bathophilus  nigerriinus  Gigl.  Mem.  R. 

Com.  Talassogr.  Ital.  48. 
.    1915b.    Stylopthalmoides  lobiancoi  Mazzarelli  e  S.  mediter- 

raneus  Mazzarelli  sono  le  rispettive  forme  larvali  di  Scopelus  can- 

imamis  C.  e  V.  e  5.  humbohi  Rissoi.  R.  Com.  Talassogr.  Ital.  Boll. 

44:1-24. 
.    1917.    Stadi  larvali  di /'ara/e;7«  M'a//>Ja  C.V.  Mem.  R.  Com. 

Talassogr.  Ital.  59. 
.    1918.    Uova  e  larve  di  Trachipterus  cristatits  Bp.  Mem.  R. 

Com.  Talassogr.  Ital.  64. 
.    1925.    \Jo\a.e\a.r\e  (\\  Regalecus  glesne  Asc.  Mem.  K.  Com. 

Talassogr.  Ital.  1 18. 
.     1928.    Uova  e  larve  di  Remora  remora  L.  Mem.  R.  Com. 


Talassogr.  Ital.  138. 
— .    1930a.   Contributo  alia  conoscenza  di  uova,  larve  e  stadi  gio- 

vanili  in  Echeneis  naucrales  Linn.  Ann.  Idrogr.  Genova  11:201- 

209. 
— .    1930b.   Uova,  sviluppo  embrionale.  stadi  larvali,  post  larvali 

e  giovanili  di  Stemoptychidae  e  Stomiatidae.  Stemoptychidae.  2. 

Ichthyococcus  ovalus  Cocco.  R.  Com.  Talassogr.   Ital.   Monogr. 

2:69-119. 
— .    1930c.    Contributo  alia  consoscenza  della  sviluppo  nei  Car- 

ancidi:  Seriola  dumehlii  Risso.  Boll.  Zool.  1:33-34. 
— .    1930d.    Ricerche  biologiche  su  materiale  raccolti  dal  Prof.  L. 

Sanzo  nella  campagna  idiografica  nei  Mar  Rosso  della  R.N.  Am- 

miraglio  Magnaghi  1923-1924.  VII.  Plectognati.  Mem.  R.  Com. 

Talassogr.  Ital.  Monogr.  167. 
— .    193  la.   Uova  stadi  embrionali  e  post-embnonali  di  Naucrates 

duclorL.  Mem.  R.  Com.  Talassogr.  Ital.  185. 
— .    1931b.    Uova  e  larve  di  Zeus  faber  L.  Arch.  Zool.  Ital.  15: 


475. 
.    1931c.    Uova  e  primi  stadi  larvali  di  Myctophum  gemellari 

Cocco  (=Scopelus  gemellari  C.  e  V.).  Atti  Accad.  Naz.  Lincei  14: 

515-519. 
— .    193 Id.   Sottordine:  Salmonoidei,  p.  21-92.  In:  Uova,  larve  e 


stadi  giovanili  di  Teleostei.  Fauna  Flora  Golfo  Napoli  Monogr.  38. 
.    1932a.   \Jo\ae  X)nm\  ^ladWarvaM  <X\  Trachurus  meditenaneus. 

Boll.  Zool.  3:73. 
— .    1932b.    Uova,  stadi  larvali  e  giovanili  di  Centrolophus  pom- 

pilus  C.V.  Mem.  R.  Com.  Talassogr.  Ital.  196. 
.    1933a.    Famiglia  2:  Macruridae,  p.  255-265.  //;.•  Uova,  larve 

e  stadi  giovanili  di  Teleostei.  Fauna  Flora  Golfo  Napoli  Monogr. 

38:255-265. 
— .    1933b.   Uova  larve  e  stadi  giovanili  di  5er(o/a  rfMmer//// Risso. 

Mem.  R.  Com.  Talassogr.  Ital.  205. 


734 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


— .    1 933c.  Uova  stadi  larvali  e  giovanili  di  Dactylopterus  voUtans 

L.  Mem.  R.  Com.  Talassogr.  Ital.  207. 
— .    1936.   Contributi  alia  conoscenza  dello  sviluppo  embrionovio 

nei  Mugilidi.  I.  Uova  e  larve  di  Mugil  cephalus  Cuv.  Ottenute  per 

tecondazione  anificiale.  II.  Uova  e  larve  di  Mugil  chelo  Cuv.  Mem. 

R.  Com.  Talassogr.  Ital.  230. 
— .    1 938a.   Uova,  stadi  embrionali,  prelarve  e  larve  di  Saurenche- 

lys  canchvora  Peters.  Mem.  R.  Com.  Talassogr.  Ital.  249. 
— .    1938b.    Uova,  ovariche  e  sviluppo  larvale  di  Aulopus  fila- 

mentosus  Cuv.  Mem.  R.  Com.  Talassogr.  Ital.  254:3-7. 
— .    1939a.    Nuovo  contributo  alia  conoscenza  dello  sviluppo  di 

Myctophum  rissoi  (Cocco).  Atti  Accad.  Gioenia  Sci.  Nat.  Catanias 

(6)3(25):  1-8. 
— .    1939b.    Rarissimi  stadi  larvali  di  Teleostei.  Arch.  Zool.  Ital. 

26:121-151. 
— .    1939c.   Uova  e  larve  di  Smara  (>!i/rf/arorC. V.  Mem.  R.  Com. 

Talassogr.  Ital.  262. 
.    1940.    Uova  e  larve  appena  achiusa  di  Lophotes  cepediamts 


(Gioma).  Mem.  R.  Com.  Talassogr.  Ital.  272. 

1956.    Divisione:  Zeomorphi  Regan,  p.  461-470.  In:  Uova, 


larve  e  stadi  giovanili  di  Teleostei.  Fauna  Flora  Golfo  Napoli 
Monogr.  38. 

Sardou,  J.  1966.  Oeufet  developpement  embryonnairede  Trac/np- 
terus  taenia  Bloch  (ordre  des  Lampridiformes  =  AUotriognathes 
famille  des  Trachypteridae)  Vie  Milieu,  Fasc.  1-A,  15:199-215. 

.    1974.   Contribution  a  la  connaissance  de  la  fauna  ichthyolo- 

gique  Malgache:  decouverte  de  poissons  de  la  famille  des  Schind- 
leriidae  dans  le  canal  de  Mozambique,  a  Nosy-Be  et  etude  d'une 
collection  de  Schindleria.  Cah.  ORSTOM,  Ser.  Oceanogr.  12(1):3- 
15. 

Sarojini,  K.  E.,  AND  J.  C.  Malhotra.  1952.  The  larval  development 
of  the  so-called  Indian  salmon,  Eleutheronema  letradaclylum  (Shaw). 
J.  Zool.  Soc.  India  4(l):63-72. 

Sasaki,  T.  1974.  On  the  larvae  of  three  species  of  rockfish  (Genus: 
Sebastes)  in  Hokkaido.  Bull.  Fac.  Fish  Hokkaido  Univ.  25:169- 
173. 

Sato,  G.  1983.  Identifica?ao,  distribuicao  e  desenvolvimento  larval 
de"Lanceta'"  Thyrsitops  lepidopoides  (Cuvier,  1931)  (Pisces:  Gem- 
pylidae)  da  regiao  comprendida  entre  Cabo  Frio  (23°S)  e  Cabo  de 
Sta.  Marta  Grande  (29°S).  Unpubl.  MS  thesis.  Instituto  Oceano- 
grafico  da  Universidade  de  Sao  Paulo,  Sao  Paulo,  Brazil.  (In  Por- 
tuguese.) 

Sato,  M.,  and  F.  Yasuda.  1980.  Metamorphosis  of  the  leptocephali 
of  the  ten-pounder,  Elops  hawaiensis.  from  Ishigaki  Island,  Japan. 
Jpn.  J.  Ichthyol.  26:315-324. 

Sauskan,  V.  I.,  AND  V.  P.  Serebryakov.  1968.  Reproduction  and 
development  of  the  silver  hake  {Merluccius  biltnearis  Mitchill).  J. 
Ichthyol.  (Engl.  Trans.  Vopr.  Ikhtiol.)  8:398-414. 

Saville,  A.  1964.  Clupeoidae.  Identification  sheets  on  eggs  and  larvae. 
Sheet  1 .  Cons.  Int.  Explor.  Mer. 

.    1981.  The  estimation  of  spawning  stock  size  from  fish  egg  and 

larval  surveys.  Rapp.  P.  V.  Reun.  Cons.  Int.  Explor.  Mer  178:268- 
278. 
-,  I.  G.  Baxter  and  D.  W.  McKay.    1974.    Relations  between 


egg  production,  larval  production  and  spawning  stock  size  in  Clyde 
herring,  p.  1 29-1 38.  In:  The  early  life  history  offish.  J.  H.  S.  Blaxter 
(ed.).  Springer-Verlag,  Berlin. 

Sawada,  Y.  1982.  Phylogeny  and  zoogeography  of  the  superfamily 
Cobitoidea  (Cyprinoidei,  Cypriniformes).  Mem.  Fac.  Fish.  Hok- 
kaido Univ.  28(2):65-223. 

Sawyer,  P.  J.  1967.  Intertidal  life-history  of  the  rock  gunnel,  f/!ofa 
gunnellus  in  the  western  Atlantic.  Copeia  1967:55-61. 

Sazima,  I.,  AND  V.  S.  Uieda.  1979.  Adaptacoes  defensivas  em  jovens 
de  Oligoplites  palometa  (Pisces,  Carangidae).  Rev.  Bras.  Biol.  39: 
687-694. 

ScHEEL,  J.  J.    1961.   Nigerian  lampeye.  Aquarium  J.  32(9):41 1-416. 

.    1962.   Factors  governing  development  of  killie  eggs.  Tropicals 

7(2):9-15. 

.    1968.   Rivuhns  of  the  Old  World.  T.F.H.  Publ.  Neptune  City, 


ScHiNDLER,  O.  1932.  Sexually  mature  larval  Hemirhamphidae  from 
the  Hawaiian  Islands.  Bull.  Bemice  P.  Bishop  Mus.  197:1-28. 

ScHLESiNGER,  G.  1909.  Zur  Phylogenie  und  Ethologie  der  Scombre- 
sociden.  Verh.  Zool.  Bot.  Ges.  Wien  59:302-339. 

Schlotterbeck,  R.  E.,  and  D.  W.  Connally.  1982.  Vertical  strati- 
fication of  three  nearshore  Southern  California  larval  fishes  (En- 
graulis  mordax.  Genyonemus  linealus.  and  Seriphus  politus).  Fish. 
Bull.  U.S.,  80:895-902. 

Schmidt,  J.  1904.  On  the  pelagic  post-larval  halibut  {Hippoglossus 
vulgaris  Flem.  and  H.  hippoglossoides  (Walb).  Medd.  Komm.  Ha- 
vunders.  Ser.  Fiskeri  1(3). 

.    1905a.    The  pelagic  post-larval  stages  of  the  Atlantic  species 

ofGadus.  Part  I.  Medd.  Komm.  Dan.  Havunders.  Ser.  Fiskeri  1(4). 

.    1905b.  On  the  larval  and  post-larval  stages  of  the  torsk  (Broi- 

mius  hrosme (,Ascan.).  Medd.  Komm.  Havunders.  Ser.  Fiskeri  1(8). 

.    1 906a.    The  pelagic  post-larval  stages  of  the  Atlantic  species 

of  Gadus.  Part  II.  Medd.  Komm.  Havunders.  Ser.  Fiskeri  2(2). 

.    1 906b.  The  pelagic  post-larval  stages  of  the  ling  (Moha  moha 

(Linne)  and  Moha  byrelange  (Walbaum).  Medd.  Komm.  Havun- 
ders. Ser.  Fiskeri  2(3). 

.    1906c.   On  the  larval  and  post-larval  development  of  the  ar- 

gentines  Argentina  silus  (Ascan.)  and  Argentina  sphyraena  (Linne) 
with  some  notes  on  Maliotus  villosus  (O.  F.  Muller).  Medd.  Komm. 
Havunders.  Ser.  Fiskeri  2(4). 

.    1907a.   On  the  post-larval  development  of  the  hake  (Afer/uc- 

cii4S  vulgaris  Flem.).  Medd.  Komm.  Havunders.  Ser.  Fiskeri  2(7). 

.    1907b.  On  the  post-larval  development  of  some  north  Atlantic 

gadoids  (Raniceps  raninus  Linne)  and  Moha  elongata  (Risso).  Medd. 
Komm.  Havunders.  Ser.  Fiskeri  2(8). 

.    1908.    On  the  post-larval  stages  of  the  John  Dory  (Zei«ya/)er 


L.)  and  some  other  Acanthopterygian  fishes.  Medd.  Komm.  Ha- 
vunders. Ser.  Fiskeri  2(9):  1-1 2. 

— .  1913.  On  the  identification  of  muraenoid  larvae  in  their  eariy 
("pre-leptocephaline")  stages.  Medd.  Komm.  Havunders.  Ser.  Fis- 
keri, 4(2):  1-1 3. 

— .  1915.  Respiratory  adaptations  of  pleuronectids.  Bull.  Acad. 
Imp.  Sci.  Petrograd  1915:421^44. 

— .  1917.  Racial  investigations.  I.  Zoarces  viviparus  L.  and  local 
races  of  same.  C.-R.  Trav.  Lab.  Carisberg  13(3):279-396. 

— .  1918.  Argentinidae,  Microstomidae,  Opisthoproctridae.  Med- 
iterranean Odontostomidae.  Rep.  Dan.  Oceanogr.  Exped.  Mediterr. 
2(A.5). 

1921.   Contributions  to  the  knowledge  of  the  young  of  the  sun- 


N.J. 


fishes  (Mola  and  Ranzania).  Medd.  Komm.  Havunders.  Ser.  Fis- 
keri 6(6):3-13. 
Schmidt,  P.  Y.    1950.    Fishes  of  the  Sea  of  Okhtsk.  Tr.  Zool.  Inst. 

Akad.  Nauk  SSSR  6: 1-392.  (Translated  by  Isr.  Program  Sci.  Transl. 

No.  1263). 
ScHNACK,  D.,  AND  H.  Rosenthal.    1978.   Shrinkage  of  Pacific  herring 

larvae  due  to  formalin  fixation  and  preservation.  Ber.  Dtsch.  Wiss. 

Komm.  Meeresforsch.  26:222-226. 
Schnakenbeck,  W.    1928.    Beitrag  zur  Kenntnis  der  Entwicklung  ei- 

niger  Meeresfische.  1.  Ber.  Dtsch.  Wiss.  Komm.  Meeresforsch.,  N.F. 

4(4):  119-229. 
.     1931.    Carangidae.  Rep.  Dan.  Oceanogr.  Exped.  Mediterr. 

2(A.14):l-20. 
Schonevelde,  S.     1624.    Ichthyologia  et  nomenclaturae  animalium 

marinorum,  quae  in  florentissimus  ducatibus.  Slesvici  et  Holstatiae, 

Hamburg. 
ScHULTZ,  L.  P.    1945.    Emmelichthyops  atlanticus.  a  new  genus  and 

species  of  fish  (family  Emmelichthyidae)  from  the  Bahamas,  with 

a  key  to  related  genera.  J.  Wash.  Acad.  Sci.  35:132-136. 
.    1 946.   A  revision  of  the  genera  of  mullets,  fishes  of  the  family 

Mugilidae.  with  descriptions  of  three  new  genera.  Proc.  U.S.  Natl. 

Mus.  96(3204):377-395. 
.    1953.   Sphyraenidae. /«.  Fishes  of  the  Marshall  and  Marianas 

Islands.  Vol.  1 .  Families  from  Asymmetrontidae  through  Siganidae. 

Schultz,  L.  P.,  E.  S.  Herald,  E.  A.  Lachner,  A.  D.  Welander  and  L. 

P.  Woods  (eds.).  U.  S.  Natl.  Mus.  Bull.  202. 


LITERATURE  CITED 


735 


— .    1957.   The  frogfishes  of  the  family  Antennariidae.  Proc.  U.S. 

Natl.  Mus.  107:47-105. 
— .    1958.    Review  of  the  parrotfishes  family  Scaridae.  U.S.  Natl. 

Mus.  Bull.  214. 

1967.   A  new  genus  and  species  of  zoarcid  fish  from  the  North 


Pacific  Ocean.  Proc.  U.S.  Natl.  Mus.  122:1-5. 

ScHWARZHANS,  W.  1978.  Otolith-morphology  and  its  usagc  for  higher 
systematic  units,  with  special  reference  to  the  Myctophiformes. 
Meded.  Werkgr.  Ten.  Kwart.  Geol.  15(4):  167-1 85. 

Scott,  A.  P.,  and  C.  Middleton.  1979.  Unicellular  algae  as  a  food 
for  turbot  (Scophthalmus  maximus  L.)  larvae— the  importance  of 
dietary  long-chain  polyunsaturated  fatty  acids.  Aquaculture  18:227- 
240. 

Scott,  E.  O.  G.  1955.  Observations  on  some  Tasmanian  fishes,  part 
7.  Pap.  Proc.  R.  Soc.  Tasmania  89:131-146. 

.    1966.   Observations  on  some  Tasmanian  fishes,  part  14.  Pap. 

Proc.  R.  Soc.  Tasmania  100:93-1 15. 

.    1967.   Observations  on  some  Tasmanian  fishes,  part  15.  Pap. 

Proc.  R.  Soc.  Tasmania  101:189-220. 

Scott,  J.  S.  1972.  Eggs  and  larvae  of  northern  sand  lance  (Ammodytes 
dubius)  from  the  Scotian  Shelf  J.  Fish.  Res.  Board  Can.  29:1667- 
1671. 

.    1982.    Distribution  of  juvenile  haddock  around  Sable  Island 

on  the  Scotian  Shelf  J.  Northwest  Atl.  Fish.  Sci.  3:87-90. 

Scott,  T.  D.  1962.  The  marine  and  freshwater  fishes  of  South  Aus- 
tralia. W.  L.  Harves,  Gov.  Printer,  Adelaide. 

,  C.  J.  M.  Glover  and  R.  V.  Southcott.    1974.    The  marine 

and  freshwater  fishes  of  South  Australia.  2nd  ed.  A.  B.  James,  Gov. 
Printer,  South  Australia. 

, ,  AND .    1980.    The  marine  and  freshwater  fishes 

of  South  Australia.  3rd  ed.  Gov.  Printer,  South  Australia. 

Scott,  W.  B.,  and  E.  J.  Grossman.  1973.  Freshwater  fishes  of  Canada. 
Bull.  Fish.  Res.  Board  Can.  184. 

Scotten,  L.  M.,  R.  E.  Smith,  N.  S.  Smith,  K.  S.  Price  and  D.  P. 
DE  SvLVA.  1973.  Pictorial  guide  to  fish  larvae  of  Delaware  Bay. 
Univ.  Del.,  Del.  Bay  Rep.  Ser.  7. 

Scrimshaw,  N.  S.  1945.  Embryonic  development  in  poeciliid  fishes. 
Biol.  Bull.  (Woods  Hole)  88(3):233-246. 

.    1946.    Egg  size  in  poeciliid  fishes.  Copeia  1946:20-23. 

Scura,  E.  D.,  and  C.  W.  Jerde.  1977.  Various  species  of  phytoplank- 
ton  as  food  for  larval  northern  anchovy,  Engraulis  mordax.  and 
relative  nutritional  value  of  the  dinoflagellates  Gymnodinium 
splendens  and  Gonyaulax  polyedra.  Fish.  Bull.  U.S.  75:577-583. 

Seaton,  D.  D.,  and  R.  S.  Bailev.  1971.  The  identification  and  de- 
velopment of  the  eggs  and  larvae  of  the  blue  whiting  Micromesislhis 
poulassou  (Risso).  J.  Cons.  Int.  Explor.  Mer  34:76-83. 

Senta.  T.  1958.  Young  of  Daicocus  pelerseni  i^yslTom).  In:  Studies 
on  the  eggs,  larvae  and  juveniles  of  Japanese  fishes.  Ser.  1.  Uchida 
et  al.  (eds.).  Fac.  Agric.  Kyushu  Univ.  Fukuoka,  Japan.  (In  Jap- 
anese.) 

.    1965.   The  buoyancy  of  sandeel  eggs  and  their  distribution  in 

theSeto  Inland  Sea.  Bull.  Jpn.  Soc.  Sci.  Fish.  31:511-517. 

,  and  S.  KuMAOAi.    1977.   Variation  in  the  vertebral  number  of 


milkfish  Chanos  chanos.  collected  from  various  localities.  Bull.  Fac. 
Fish.  Nagasaki  Univ.  43:35-40. 

Serebryakov,  V.  P.  1978.  Development  of  the  spotted  hake,  Uro- 
phycis  regius.  from  the  Northwestern  Atlantic.  J.  Ichthyol.  (Engl. 
Transl.  Vopr.  Ikhtiol.)  18:793-799. 

.    1 980.   Fish  eggs  and  larvae  from  the  Flemish  Cap  area.  NAFO 

SCR  Doc,  No.  90,  Senal  No.  N145. 

Seshappa,  G.,  AND  B.  S.  Bhimachar.  1955.  Studies  on  the  fishery  and 
biology  of  the  Malabar  sole,  Cvnoglossus  semifasciatus  Day.  Indian 
J.  Fish.  2:180-230. 

Sette,  O.  E.  1943.  Biology  of  the  Atlantic  mackerel  (5co/nfter  jcom- 
brns)  of  North  Amenca:  Part  I.  Early  life  history,  including  growth, 
drift  and  mortality  of  the  egg  and  larval  populations.  U.S.  Fish 
Wildl.  Serv.  Fish.  Bull.  50:149-237. 

,  AND  E.  H.  Ahlstrom.    1948.   Estimation  of  abundance  of  the 

eggs  of  the  Pacific  pilchard  Sardinops  caerulea  off  southern  Cali- 
fornia during  1940  and  1941.  J.  Mar.  Res.  7:511-542. 

Sha,  X.,  AND  H.  RuAN.    1981.    The  habits  and  morphological  devel- 


opment of  early  stages  of  IHsha  elongata  (Bennett).  Trans.  Chin. 

Ichthyol.  Soc.  2:81-88. 
, ,  AND  G.  He.    1 98 1 .    The  development  of  the  egg  and 

larval  stages  of  the  lumpfish,  Immicusjaponicus(C.  and  V.).  Ocean- 

ol.  Limnol.  Sin.  12(4):365-371. 
Shakxee,  J.  B.,  AND  C.  S.  Tamaru.    1981.    Biochemical  and  morpho- 
logical evolution  of  Hawaiian  bonefishes  (Albula).  Syst.  Zool.  30: 

125-146. 
,  AND  G.  S.  Whitt.    1981.    Lactate  dehydrogenase  isozymes  of 

gadiform  fishes:  divergent  patterns  of  gene  expression  indicate  a 

heterogeneous  taxon.  Copeia  1981:563-578. 
Shaw,  E.  S.    1955.    The  embryology  of  the  sergeant  major, /46w/e/i/«/ 

saxatilis.  Copeia  1955:85-99. 
Shelbourne,  J.  E.    1964.    The  artificial  propagation  of  marine  fish. 

Adv.  Mar.  Biol.  2:1-83. 
.   1975.  Marine  fish  cultivation:  pioneering  studies  on  the  culture 

of  the  larvae  of  the  plaice  (Pleuronectes  platessa  L.)  and  the  sole 

{Solea  solea  L.).  Fish.  Invest.  Ser.  II  Mar.  Fish.  G.  B.  Minist.  Agric. 

Fish.  Food,  Vol.  9. 
Shelton,  W.  L.    1978.    Fate  of  the  follicular  epithelium  in  Dorosoma 

petenense  {Pisct%:  Clupeidae).  Copeia  1978:237-244. 
,  AND  R.  Stephens.    1980.   Comparative  embryogeny  and  early 

development  of  threadfin  and  gizzard  shad.  Prog.  Fish-Cult.  42: 

34-41. 
Shen,  S.  C.    1971.    A  new  genus  of  clinid  fishes  from  the  Indo-West 

Pacific,  with  a  redescription  of  Clinus  nematopterus.  Copeia  1971: 

697-706. 
Shiganova,  T.  A.    1974.    Postembryonic  development  oi  Hygophum 

benoili  (Myctophidae,  Pisces).  J.  Ichthyol.  (Engl.  Transl.  Vopr.  Ikh- 
tiol.) 14:746-754. 
.    1975.    Postembryonic  development  of //v^o/i/jKW  macroc/!(> 

(Myctophidae,  Pisces).  J.  Ichthyol.  (Engl.  Transl.  Vopr.  Ikhtiol.)  15: 

429^37. 
.    1977.   Larvae  and  juveniles  of  the  lantemfishes  (Myctophidae, 

Pisces)  of  the  Atlantic  Ocean.  Tr.  Inst.  Okeanol.  Akad.  Nauk  SSSR 

109:42-12. 
Shiooaki,  M.    1980.    Redescription  of  the  stichaeid  fish  C/»>o/op>i/i 

saitone.  Jpn.  J.  Ichthyol.  28:129-134. 
.    1981.    Life  history  of  the  stichaeid  fish,  Opisthocentrus  ocel- 

lalus.  Jpn.  J.  Ichthyol.  29:77-85. 
.    1 982.    Larvae,  juveniles,  and  a  gravid  female  of  the  blennioid 

Cryptacanthodes  bergi,  collected  from  Aomori,  Japan.  Jpn.  J.  Ich- 
thyol. 29:102-104. 
.    1983.    On  the  life  history  of  the  stichaeid  fish  Chirolophis 

japonicus.  Jpn.  J.  Ichthyol.  29:446-455. 
,  AND  Y.  DoTsu.    1971a.    The  hfe  history  of  the  clingfish, -4.S- 

pasma  minima.  Jpn.  J.  Ichthyol.  18:76-84. 
,  AND .    1971b.    Larvae  and  juveniles  of  the  clingfishes. 


Lepadichlhys frenatus  and  Aspasmichlhys  ciconiae.  Jpn.  J.  Ichthyol. 

18:85-89. 
— ,  AND .    1971c.    Egg  development  and  hatched  larva  of 

the  clingfish,  Lepadichlhys  frenatus.  Bull.  Fac.  Fish.  Nagasaki  Univ. 

32:1-5. 
— ,  AND .    197  Id.   The  life  history  of  the  clingfish,  Con/rfe/w 

laticephalus.  Bull.  Fac.  Fish.  Nagasaki  Univ.  32:7-16. 
,  AND .    I971e.  The  life  history  of  the  gobiid  fish,  f.vperf/o 

parvulus.  Bull.  Fac.  Fish.  Nagasaki  Univ.  32:17-25. 
— ,  AND .    1972a.   Life  history  of  the  blennioid  fish,  Dictyo- 

soma  burgeri.  Bull.  Fac.  Fish.  Nagasaki  Univ.  33:21-38.  (In  Jap- 
anese.) 
— ,  AND .    1 972b.   The  life  history  of  the  blenniid  fish  Neo- 

clinns byrope.  Bull.  Fac.  Fish.  Nagasaki  Univ.  34:1-8.  (In  Japanese.) 
,  AND .    1 972c.    The  life  history  of  the  gobiid  fish,  Lucio- 

gobius  elongalus.  Bull.  Fac.  Fish.  Nagasaki  Univ.  34:9-18. 
— ,  AND .    I972d.    The  spawning  and  the  larva  rearing  of 

the  clingfish,  Aspasma  ciconiae  and  the  spawning  of  the  clingfish, 

Aspasma  minima.  Bull.  Fac.  Fish.  Nagasaki  Univ.  34:39-33. 
— ,  AND .    1973.   The  egg  development  and  larva  rearing  of 


the  tripterygiid  blenny,  Trlplerygion  etheosloma.  Jpn.  J.  Ichthyol. 
20:42-46.  (In  Japanese,  English  summary.) 


736 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Shiokawa,  T.,  AND  H.  TsuKAHARA.  1961.  Studies  On  habits  ofcoastal 
fishes  in  the  Amakusa  Islands.  I.  Early  life  history  of  the  purple 
rockfish,  Sebastes  pachycephalus  pachycephalus  Temminck  et 
Schlegel.  Rec.  Oceanogr.  Works  Jpn.,  Spec.  No.  5:123-127. 

Shojima,  Y.  1962.  On  the  postlarvae  and  juveniles  of  carangid  fishes 
collected  together  with  jelly-fishes.  Bull.  Seikai  Reg.  Fish.  Res.  Lab. 
27:49-58. 

,  S.  FujiTA  AND  K.  UcHiDA.    1957.  On  the  egg  development  and 

prelarval  stages  of  a  kind  of  barracuda,  Sphyraena  pinguis  Gunther. 
Sci.  Bull.  Fac.  Agric.  Kyushu  Univ.  16:313-318. 

Shores,  D.  L.  1969.  Postlarval  Sudis  (Pisces:  Paralepididae)  in  the 
Atlantic  Ocean.  Breviora  334. 

Shuleikjn,  V.  V.  1958.  How  the  pilot  fish  moves  with  the  speed  of 
the  shark.  Dokl.  Akad.  Nauk  SSSR  1 19(106):  140-143.  (Transl.  by 
the  Amer.  Inst,  of  Biol.  Sci.). 

Shust,  K.  V.  1978.  On  the  distribution  and  biology  of  members  of 
the  genus  Micromesislius  (family  Gadidae).  J.  Ichthyol.  (Engl.  Transl. 
Vopr.  Ikhtiol.)  18:490^*93. 

SiBLEV,  C.  G.,  AND  J.  E.  Ahlquist.  1981.  The  phylogeny  and  rela- 
tionships of  the  ratite  birds  as  indicated  by  DNA-DNA  hybridiza- 
tion, p.  301-335.  In:  Evolution  today.  G.  G.  E.  Scudder  and  J.  J. 
Reveal  (eds.).  Proc.  Sec.  Int.  Congr.  Syst.  Evol.  Biol. 

Silverman,  M.J.  1975.  Scale  development  in  the  bluefish.Po  wa/owM.s 
saltatrix.  Trans.  Am.  Fish.  Soc.  104:773-774. 

Simons,  E.  v.,  AND  J.  R.  VAN  Horn.  1971.  A  new  procedure  for  whole- 
mount  alcian  blue  staining  of  the  cartilaginous  skeleton  of  chicken 
embryos,  adapted  to  the  clearing  procedure  in  potassium  hydroxide. 
Acta  Morphol.  Need.  Scand.  8(1970/1971):281-292. 

Simonsen,  M.  1978.  The  use  of  discriminant  analysis  in  the  identi- 
fication of  two  species  of  larval  smelt.  Spinnchus  ihaleichthys,  and 
Hypomesus  I.  transpacificus  in  the  Sacramento-San  Joaquin  estu- 
ary. Unpubl.  MS  thesis.  Univ.  of  Pacific,  Stockton,  CA. 

Simpson,  A.  C.  1959.  The  spawning  of  plaice  in  the  North  Sea.  Fish. 
Invest.  Ser.  II  Mar.  Fish.  Gr.  Brit.  Minist.  Agric.  Fish.  Food  22(7). 

Simpson,  J.  G.  1959.  Identification  of  the  egg.  early  life  history  and 
spawning  areas  of  the  anchoveta,  Cetengraulis  mysticetus  (Gun- 
ther), in  the  Gulf  of  Panama.  Inter-Amer.  Trop.  Tuna  Comm.  Bull. 
3:437-580. 

Singer,  C.    1950.   A  history  of  biology.  Henry  Schuman,  New  York. 

.    1959.    A  short  history  of  scientific  ideas  to  1900.  Clarendon 

Press,  Oxford. 

Slipp,  J.  W.,  and  A.  C.  DeLacy.  1952.  On  the  distribution  and  habits 
of  the  wattled  eelpout,  Lycodes  palearis.  Copeia  1952:201-203. 

Smigielskj,  A.  1975a.  Hormone-induced  spawnings  of  the  summer 
flounder  and  rearing  of  the  larvae  in  the  laboratory.  Prog.  Fish- 
Cuh.  37:3-8. 

.    1975b.    Hormonal-induced  ovulation  of  the  winter  flounder, 

Pseudopleuronectes  amencanus.  Fish.  Bull.  U.S.  73:431-438. 

Smith,  C.  L.  1957.  Two  new  clinid  blennies  (Atalacoclenus)  from 
Pueno  Rico.  Univ.  Mich.  Mus.  Zool.  Occas.  Pap.  585:1-15. 

.    1971.   A  revision  of  the  American  groupers:  ii^j/Vifp/je/Mi  and 

allied  genera.  Bull.  Am.  Mus.  Nat.  Hist.  146:67-241. 

,  and  R.  M.  Bailey.    1961.   Evolution  of  the  dorsal-fin  supports 

of  percoid  fishes.  Pap.  Mich.  Acad.  Sci.  Arts  Lett.  46:345-361. 

,  J.  C.  Tyler  AND  M.  N.  Feinberg.    1981.    Population  ecology 


and  biology  of  the  pearlfish  (Carapus  bermudensis)  in  the  lagoon 
at  Bimini,  Bahamas.  Bull.  Mar.  Sci.  31:876-902. 

Smith,  D.  G.  1969.  Xenocongrid  eel  larvae  in  the  western  North 
Atlantic.  Bull.  Mar.  Sci.  19:377-408. 

.  1970.  Notacanthiform  leptocephali  in  the  western  North  At- 
lantic. Copeia  1970:1-9. 

.    1971.    Osteology  and  relationships  of  the  congrid  eels  of  the 

western  North  Atlantic  (Pisces,  Anguilliformes).  Unpubl.  PhD  dis- 
sertation. Univ.  of  Miami,  Coral  Gables,  FL. 

.    1979.   Guide  to  the  leptocephali  (Elopiformes,  Anguilliformes 

and  Notacanthiformes).  NOAA  Tech.  Rep.  NMFS  Circ.  424. 

.  1980.  Early  larvae  of  the  tarpon,  Megalops  atlantica  Valen- 
ciennes (Pisces:  Elopidae),  with  notes  on  spawning  in  the  Gulf  of 
Mexico  and  the  Yucatan  Channel.  Bull.  Mar.  Sci.  30:136-141. 


,  and  P.  H.  J.  Castle.    1972.    The  eel  genus  A'eocon^er  Girard: 

Systematics,  osteology,  and  life  history.  Bull.  Mar.  Sci.  22: 196-249. 

,  AND .  1982.  Larvae  of  the  nettastomatid  eels:  system- 
atics and  distribution.  Dana-Rep.  Carlsberg  Found.  90:1-44. 

,  AND  M.  M.  Leiby.    1980.    The  larva  of  the  congrid  eel  Acro- 

mycler  alcocki  {Pisces:  Anguilliformes),  and  the  distinction  between 
congrid  and  ophichthid  larvae.  Proc.  Biol.  Soc.  Wash.  93(2):388- 
394. 

,  and  T.  Potthoff.    1975.    Leptocephali  taxonomy,  p.  19-20. 

In:  Ichthyoplankton  Rep.  CICAR  Ichthyoplankton  Workshop, 
Mexico  City,  17-26  July  1974.  UNESCO  Tech.  Pap.  Mar.  Sci.  20. 

Smith,  H.  M.  1907.  The  fishes  of  North  Carolina.  N.C.  Geol.  Econ. 
Surv.  2:1-453. 

Smith,  J.  L.  B.  1933a.  An  interesting  new  myctophid  fish  from  South 
Africa.  Trans.  R.  Soc.  S.  Afr.,  21(2):I25-127. 

.    1933b.  The  South  African  species  of  the  genus //e«i;rAa«?p/!Mj 

Cuv.  Trans.  R.  Soc.  S.  Afr.  2 1(2):  129-1 50. 

.    1935.   The  fishes  of  the  family  Mugilidae  in  South  Africa.  Ann. 

S.  Afr.  Mus.  30:587-644. 

.    1938.   An  interesting  early  post-larval  stage  of  the  "Galjoen." 

Trans.  R.  Soc.  S.  Afr.  25:389-390. 

.    1947.   A  generic  revision  of  the  mugilid  fishes  of  South  Africa. 


Ann.  Mag.  Nat.  Hist.,  Ser.  II,  14:833-843. 
— .    1949.  The  sea  fishes  of  southern  Africa.  Central  News  Agency, 

Ltd.,  Johannesburg. 
— .    1956a.    A  new  dealfish  from  South  Africa.  Ann.  Mag.  Nat. 

Hist.  12:449^53. 
— .    1956b.    The  fishes  of  the  family  Sphyraenidae  in  the  western 

Indian  Ocean.  Rhodes  Univ.  Dep.  Ichthyol.  Bull.  3:37-46. 
— .    1958a.    The  gunnellichthid  fishes  with  description  of  two  new 

species  from  East  Africa  and  of  Gunnellichthys  (Ctarkichthys)  bi- 

Imeatus  (C\2ir^),  1936.  Rhodes  Univ.  Dep.  Ichthyol.  Bull.  9:123- 

129. 
— .    1958b.   Fishes  of  the  family  Tetrarogidae,  Caracanthidae,  and 

Synanciidae  from  the  Western  Indian  Ocean  with  further  notes  on 

scorpaenid  fishes.  Rhodes  Univ.  Dep.  Ichthyol.  Bull.  12:167-181. 
— .    1961.   The  sea  fishes  of  southern  Africa.  Central  News  Agency, 

Ltd.,  South  Africa. 
— .    1963.  Fishes  of  the  families  Draconettidae  and  Callionymidae 

from  the  Red  Sea  and  the  western  Indian  Ocean.  Rhodes  Univ. 

Dep.  Ichthyol.  Bull.  28:547-564. 
— .    1966a.    Fishes  of  the  sub-family  Nasinae  with  a  synopsis  of 


the  Prionurinae.  Rhodes  Univ.  Dep.  Ichthyol.  Bull.  32:635-681. 
.    1966b.    Hitherto  unknown  early  developmental  larval  stadia 

of  the  west  African  albulid  fish,  Pterothrissus  belloci  Cadenat,  1937. 

Rhodes  Univ.  Dep.  Ichthyol.  Occas.  Pap.  (6):57-61. 
.    1968.    A  new  liparine  fish  from  the  Red  Sea.  J.  Nat.  Hist.  2: 

105-109. 
Smith,  P.  E.    1972.    The  increase  in  spawning  biomass  of  northern 

anchovy  Engraults  mordax.  Fish.  Bull.  U.S.  70:849-874. 
,  and  S.  L.  Richardson.    1977.   Standard  techniques  for  pelagic 

fish  egg  and  larval  surveys.  FAO  Fish.  Tech.  Pap.  No.  175. 
,  AND .    1979.    Selected  bibliography  on  pelagic  fish  egg 

and  larva  surveys.  FAO  Fish.  Circ.  No.  706. 
Smith,  S.    1957.   Early  development  and  hatching,  p.  323-359. /«.■  The 

Physiology  of  Fishes.  Vol.  1.  M.  E.  Brown  (ed).  Academic  Press, 

New  York. 
Smith,  W.  G.,  AND  M.  P.  Fahay.    1970.   Description  of  eggs  and  larvae 

of  the  summer  flounder,  Paralichthys  dentatus.  U.S.  Fish  Wildl. 

Serv.  Res.  Rep.  75. 
Smith-Vaniz,  W.  F.    1975.   Supplemental  description  of  a  rare  blenniid 

fish  Phenablennius  heyligen  (Bleeker).  Proc.  Acad.  Nat.  Sci.  Phila. 

l27(6):53-55. 
.    1976.    The  saber-toothed  blennies,  tribe  Nemophini  (Pisces: 

Blenniidae).  Proc.  Acad.  Nat.  Sci.  Phila.  Monogr.  19:1-196. 
,  AND  F.  J.  Palacio.    1974.    Atlantic  fishes  of  the  genus  Acan- 

themblemaria.  with  descriptions  of  three  new  species  and  com- 
ments on  Pacific  species  (Clinidae:  Chaenopsinae).  Proc.  Acad.  Nat. 

Sci.  Phila.  125(  1  l):197-224. 


LITERATURE  CITED 


737 


— ,  AND  V.  G.  Springer.    1971.    Synopsis  of  the  tribe  Salariini, 
with  description  of  five  new  genera  and  three  new  species  (Pisces: 
Blenniidae).  Smithson.  Contrib.  Zool.  73:1-72. 
-,  AND  J.  C.  Staiger.    1973.   Comparative  revision  oi  Scomber- 


oides,  Oligoplttes.  Parana,  and  Hypacanthus  with  comments  on  the 
phylogenetic  position  of  Campo^ramma  (Pisces:  Carangidae).  Proc. 
Calif.  Acad.  Sci.  39:185-256. 

Smitt,  F.  A.  1895.  A  history  of  Scandinavian  fishes,  II.  Stockholm 
andLondon:567-1240. 

Smoker,  W.,  AND  W.  C.  Pearcv.  1970.  Growth  and  reproduction  of 
the  lantemfish  Stenobrachius  lecopsarus.  J.  Fish.  Res.  Board  Can. 
27:1265-1275. 

Smol'Yanov,  I.  I.  1961.  The  development  of  Brachymyslax  lenok 
(Pallas).  Vopr.  Ikhtiol.  18:136-148.  (In  Russian.) 

Sneath,  P.,  and  R.  SoKAL.  1973.  Numerical  taxonomy.  W.  H.  Free- 
man, San  Francisco. 

Snyder,  D.  E.  1976.  Terminologies  for  intervals  of  larval  fish  devel- 
opment, p.  4 1-58.  In:  Great  Lakes  fish  egg  and  larvae  identification: 
Proceedings  of  a  workshop.  J.  Boreman(ed.).  U.S.  FishWildl.  Serv. 
Biol.  Serv.  Prog.  FWS/OBS-76/23. 

.  1979.  Burbot:  larval  evidence  for  more  than  one  North  Amer- 
ican species,  p.  204-219.  In:  Proc.  3rd  Symp.  Larval  Fish.  West. 
Kentucky  Univ.  Dep.  Biol.,  Bowling  Green. 

.    1981.   Contributions  to  a  guide  to  the  cypriniform  fish  larvae 

of  the  upper  Colorado  River  system.  U.S.  Dep.  Inter..  Bur.  Land 
Manage.,  Colo.  Off.,  Biol.  Sci.  Sen  3. 

,  AND  S.  C.  Douglas.    1978.    Description  and  identification  of 

mooneye,  Hiodon  tergtsus.  protolarvae.  Trans.  Am.  Fish.  Soc.  107: 
590-594. 

,  M.  B.  M.  Snyder  and  S.  C.  Douglas.    1977.    Identification 


of  golden  shiner,  Notemigonus  crysoleucas,  spotfin  shiner,  Nolropis 

spilopterus.  and  fathead  minnow,  Pimephales  promelas.  larvae.  J. 

Fish.  Res.  Board  Can.  34:1397-1409. 
Soin,  S.  C.    1966.    Development  types  of  structures  and  phylogeny  of 

vascular  systems  of  the  vitelline  sac  in  fish  embryos,  performing  a 

respiratory  function.  Zool.  Zh.  45:1382-1397.  (In  Russian.) 
.    1968.  Some  features  in  the  development  of  the  blenny[ZoarcM 

viviparus  (L.)]  in  relation  to  viviparity.  Prob.  Ichthyol.  (Eng.  Transl. 

Vopr.  Ikhtiol.)  8:222-229. 

1980.    Types  of  development  of  salmoniform  fishes  and  their 


taxonomic  importance.  J.  Ichthyol.  (Engl.  Transl.  Vopr.  Ikhtiol.) 

20(l):49-56. 
Sokal,  R.,  AND  Sneath,  P.    1963.    Principles  of  numerical  taxonomy. 

W.  H.  Freeman,  San  Francisco. 
SoLDATOv,  V.  K.,  and  G.  U.  LiNDBERG.    1930.   A  review  of  the  fishes 

of  the  seas  of  the  Far  East.  Izv.  Tikhookean.  Nauchno-Issled.  Inst. 

Rybn.  Khoz.  Okeanogr.  5. 
SOLTZ,  D.  L.,  AND  M.  F.  Hirschreld.    1981.    Genetic  differentiation 

of  pupfishes  (Genus  Cyprinodon)  in  the  American  southwest,  p. 

291-333.  In:  Fishes  in  North  American  Deserts.  R.  J.  Naiman  and 

D.  L.  Soltz  (eds.).  John  Wiley  &  Sons,  New  York. 
SooNG,  M.  H.  H.    1968.    Aspects  of  reproduction  in  freshwater  half- 
beaks.  Malay.  Nat.  J.  21(suppl.):xxxiii-xxxiv. 
Sparta,  A.    1926.    Uova  e  larve  di  Fierasfer  dentatus  Cuv.  Mem.  R. 

Com.  Talassogr.  Ital.  122. 
.    1 928.  Stadi  larvali  di  Uraleptus  maraldi Risso.  Mem.  R.  Com. 

Talassogr.  Ital.  141. 
.     1932.    Contributo  alia  conoscenza  dello  sviluppo  post-em- 

brionale  in  Anthias  sacer  Bloch.  Mem.  R.  Com.  Talassogr.  Ital. 

195. 
.    1933.    Ordine:  Allotriognathi,  p.  266-279.  In:  Uova.  larve  e 

stadi  giovanili  di  Teleostei.  Fauna  Flora  Golfo  Napoli  Monogr.  38. 
.    1934.   Uova  e  larve  di  Gobiidae.  I.  GoW«ipaga/!c//u5  L.  Mem. 

R.  Com.  Talassogr.  Ital.  211. 
.    1935.    Contributo  alia  conoscenze  dello  sviluppo  nei  Percidi. 

Uova  ovariche  mature  di  Epinephelus  giia:a  L.  e  stadi  post-em- 

brionali  e  larvali  di  Epinephelus  alexandrinus  Cuv.  e  Val.  Mem. 

R.  Com.  Talassogr.  Ital.  224. 
.    1936.   Contributo  alia  conoscenza  di  uova,  stadi  embrionali  e 


post-embrionali  in  Macrorhamphosus  scolopax  L.  Mem.  R.  Com. 

Talassogr.  Ital.  225. 
— .    1937.   Contributo  alia  conoscenza  dello  sviluppo  embrionale 

e  post-embnonale  nei  Murenoidi.   \.  —  Coecu!a  imberbis  (De  la 

Roche).  2.  —  Ophichthys  remicaudus  (Kaup).  Mem.  R.  Com.  Tal- 
assogr. Ital.  244. 
— .    1938a.   Contributo  alia  conoscenza  dello  sviluppo  embrionale 

e  post-embrionale  nei  Murenoidi.  i.  —  Coecula  coeca  (L.).  Mem.  R. 

Com.  Talassogr.  Ital.  248. 
— .    1938b.   Contributo  alia  conoscenza  dello  sviluppo  embrionale 

e  post-embrionale  nei  Murenoidi.  4.  —  Ophisoma  balearicum  De  la 

Roche.  Mem.  R.  Com.  Talassogr.  Ital.  252. 
— .    1939a.   Contributo  alia  conoscenza  dello  sviluppo  nei  Percidi. 

Uova,  stadi  embrionali  e  post-embrionali  di  Polyprion  cernium 

(Val.)  Mem.  R.  Com.  Talassogr.  Ital.  259. 
— .    1939b.  Contributo  alia  conoscenza  dello  sviluppo  embrionale 

e  post-embrionale  nei  Murenoidi.  5.  — Nuova  specie  di  uovo  di 

Murenoidi  (Conger  congerl).  Mem.  R.  Com.  Talassogr.  Ital.  264. 
— .    1 939c.   Contributo  alia  conoscenza  dello  sviluppo  embrionale 

e  post-embrionale  nei  Murenoidi.  (>.  —  Ophisurus  serpens  L.  Mem. 

R.  Com.  Talassogr.  Ital.  267. 
— .    1 939d.  Contributo  alia  conoscenza  dello  sviluppo  embrionale 

e  post-embrionale  nei  Murenoidi.  1  .  —  Chlopsis  bicolor('R.ai.).  Mem. 

R.  Com.  Talassogr.  Ital.  268. 
— .    1 940a.   Contributo  alia  conoscenza  dello  sviluppo  embrionale 


e  post-embrionale  nei  Murenoidi.  %.  —  Echelus  myrus  L.  Mem.  R. 

Com.  Talassogr.  Ital.  273. 
— .    1 940b.   Contributo  alia  conoscenza  dello  sviluppo  embrionale 

e  post-embrionale  nei  Murenoidi.  9.  — Seconda  nuova  specie  di  uova 

di  Murenoidi.  Mem.  R.  Com.  Talassogr.  Ital.  279. 
— .    1 942a.  Uova,  stadi  embrionali,  prelarve  e  larve  di  Nettastoma 

melanurum  Raf.  Arch.  Oceanogr.  Limnol.  2:151-158. 
.    1942b.    Uova  e  larvae  di  Aiherina  rissoi  C.  V.  ottenute  da 

fecondiazione  artificiale,  e  stadi  post-embrionali  raccolti  nei  planc- 

ton.  Arch.  Oceanogr.  Limnol.  2:161-168. 
.     1946.    Contributo  alia  conoscenza  dello  sviluppo  post-em- 
brionale in  Cubiceps  gracilis  Lowe.  Boll.  Pesca  Piscic.  Idrobiol. 

22(1). 
.    1948.   Uova  ovariche,  uova  fecondate  tenute  in  coltura,  larva 

alia  schiusa,  stadi  larvali  e  giovanili  di  Cristiceps  argentalus  Risso. 

Arch.  Oceanogr.  Liminol.,  Mem.  315,  5(l-3):79-84. 
.    1951.  Rarissimi  stadi  larvali  e  giovanili  di  Myctophum  glaciale 

(Reinhardt).  Boll.  Pesca  Piscic.  Idrobiol.  6(2):101-107. 
.     1952.    Contributo  alia  conoscenza  dello  sviluppo  larvale  di 

Myctophum  melopoclampum  Cocco.  Boll.  Pesca  Piscic.  Idrobiol. 

7(i):5-10. 
.    1953.    Uova  e  larve  di  Tetrapturus  belone  Raf  (Aguglia  Im- 


perial). Boll.  Pesca  Piscic.  Idrobiol.  8(l):58-62. 
.    1954.   Desstadesjeunesraresde  teleosteens  marins((7ra/ep/M.s 

maraldii  Risso,   Temnodon  saltalor  C.  V.,  Lophotes  cepedianus 

Gioma).  Rapp.  P.  V.  Reun.  Comm.  Int.  Explor.  Sci.  Mer.  Mediterr. 

Monaco  12:79-84. 
— .    1956a.   Sottordine:  Pharyngognathi  J.  Miiller,  p.  576-598.  In: 

Uova,  larve  e  stadi  giovanili  di  Teleostei.  Fauna  Flora  Golfo  Napoli 

Monogr.  38. 
.    1956b.    Sottordine:  Scleroparei.  Famiglia  1:  Scorpaenidae,  p. 

599-626.  In:  Uova.  larve  e  stadi  giovanili  de  Teleostei.  Fauna  Flora 

Golfo  Napoli  Monogr.  38. 
— .    1962.    Uova,  larve  alia  schiusa  ed  al  4  giomo  de  vita,  Po- 


matomus  sallalrix  GiW.  (Syn.  Temnodor  sallator.  C.V.).  Boll.  Pesca 

Piscic.  Idrobiol.  l7(l):5-8. 
Sponaugle,  D.  L,  and  J.  P.  Wourms.    1979.   Secondary  egg  envelope 

pre-patteming  by  follicle  cells  during  fish  oogenesis.  Am.  Zool.  19: 

946. 
Springer,  V.  G.    1954.  Western  Atlantic  fishes  of  the  genus /"aracZ/wMi:. 

Tex.  J.  Sci.  4:422-441. 
.    1955.   The  taxonomic  status  of  the  fishes  of  the  genus  5;ar/i- 

monotus.  including  a  review  of  the  Atlantic  species.  Bull.  Mar.  Sci. 

Gulf  Caribb.  5:66-80. 


738 


ONTOGENY  AND  SYSTEM ATICS  OF  FISHES -AHLSTROM  SYMPOSIUM 


— .  1958.  Systematics  and  zoogeography  of  the  clinid  fish  of  the 
subtribe  Labrisomini.  Tex.  Inst.  Mar.  Sci.,  Hubbs  Publ.  5:4 1 8-492. 

— .  1959.  A  new  species  of  LaZimowiis  from  the  Caribbean  Sea, 
with  notes  on  other  fishes  of  the  subtribe  Labrisomini.  Copeia  1959: 
289-292. 

— .  1967.  Revision  of  the  circumtropical  shorefish  genus  ifmo- 
macro^^iis  (Blenniidae:  Salariinae).  Proc.  U.S.  Natl.  Mus.  1 22(3582): 
1-150. 

— .  1968.  Osteology  and  classification  of  the  fishes  of  the  family 
Blenniidae.  U.S.  Natl.  Mus.  Bull.  284. 

— .  1970.  The  western  south  Atlantic  clinid  fish  Ribeiroclinus 
eigenmanni.  with  discussion  of  the  intrarelationships  and  zooge- 
ography of  the  Clinidae.  Copeia  1970:430-436. 

— .  1971.  Revision  of  the  fish  genus  Ecsenius  (Blenniidae.  Blen- 
niinae,  Salariini).  Smithson.  Contrib.  Zool.  72. 

— .  1972a.  Synopsis  of  the  tribe  of  Omobranchini  with  descrip- 
tions of  three  new  genera  and  two  new  species  (Pisces:  Blenniidae). 
Smithson.  Contrib.  Zool.  130. 

— .    1972b.   Additions  to  revisions  of  the  blenniid  fish  genera  £c- 


senius  and  Enlomacrodus,  with  descriptions  of  three  new  species 
of  Ecsenius.  Smithson.  Contrib.  Zool.  134. 

— .  1976.  C;>ma/ur;i«  iMwarei.  new  genus  and  species  of  blenniid 
fish  from  the  Indian  Ocean.  Proc.  Biol.  Soc.  Wash.  89(3):  199-203. 

— .  1978.  Synonymization  of  the  family  Oxudercidae,  with  com- 
ments on  the  identity  of  Apocryptes  cantons  Day  (Pisces:  Gobiidae). 
Smithson.  Contrib.  Zool.  270. 

— .  1982.  Pacific  plate  biogeography,  with  special  reference  to 
shorefishes.  Smithson.  Contrib.  Zool.  367. 

— .  1983.  Tyson  belos,  new  genus  and  species  of  western  Pacific 
fish  (Gobiidae,  Xenisthminae),  with  discussions  of  gobioid  osteol- 
ogy and  classification.  Smithson.  Contrib.  Zool.  390. 

— .  MS.  A  new  genus  and  species  of  western  Pacific  fish  (Gobiidae, 
Xenisthminae),  with  discussions  of  gobioid  osteology  and  classi- 
fication. 

— ,  AND  T.  H.  Eraser.  1976.  Synonymy  of  the  fish  families  Chei- 
lobranchidae  (=Alabetidae)  and  Gobiesocidae,  with  descriptions  of 
two  new  species  of  Alabes.  Smithson.  Contrib.  Zool.  234. 

— ,  AND  M.  F.  GoMON.  1975a.  Revision  of  the  blenniid  fish  genus 
Omobranchus  with  descriptions  of  three  new  species  and  notes  on 
other  species  of  the  tribe  Omobranchini.  Smithson.  Contrib.  Zool. 
177. 

— ,  AND .    1975b.    Variation  in  the  western  Atlantic  clinid 


fish  Malacoctenus  iriangulatus  with  a  revised  key  to  the  Atlantic 

species  of  Malacoctenus.  Smithson.  Contrib.  Zool.  200. 
,  C.  L.  Smith  and  T.  H.  Fraser.    1977.    New  species  of  pro- 

togynous  hermaphroditic  fish,  and  synonymy  of  Anisochromidae, 

Pseudoplesiopidae,  and  Pseudochromidae.  Smithson.  Contrib.  Zool. 

252. 
,  AND  W.  F.  Smith-Vaniz.    1972.   A  new  iribe  (Phenablennitni) 

and  genus  (Phenablennius)  of  blenniid  fishes  based  on  Petroscirtes 

heyligeri  (Bleeker).  Copeia  1972:64-71. 
,  AND  A.  E.  Spreitzer.    1978.   Five  new  species  and  a  new  genus 

of  Indian  Ocean  blenniid  fishes,  tribe  Salariini,  with  a  key  to  genera 

of  the  tribe.  Smithson.  Contrib.  Zool.  268. 
Staioer,  J.  C.    1965.    Atlantic  flying  fishes  of  the  genus  Cypselurus, 

with  descriptions  of  the  juveniles.  Bull.  Mar.  Sci.  15:672-725. 
Stark,  W.  A.,  II.    1971.   Biology  of  the  gray  snapper,  Li^f/a/iMjgr/ieiis 

(Linnaeus),  in  the  Florida  Keys,  p.  1 1-150.  In:  Investigations  on 

the  gray  snapper,  Lutjanus  griseus.  W.  A.  Stark,  II,  and  R.  E.  Schroe- 

der  (eds.).  Stud.  Trop.  Oceanogr.  (Miami)  10. 
Starks,  E.  C.    1 900.    The  osteological  characters  of  the  fishes  of  the 

Suborder  Percesoces.  Proc.  U.S.  Natl.  Mus.  22:1-10. 
.    1902.   The  shoulder  girdle  and  characteristic  osteology  of  the 

hemibranchiate  fishes.  Proc.  U.S.  Natl.  Mus.  25:619-634. 
.    1911.   The  osteology  and  relationships  of  fishes  belonging  to 

the  family  Carangidae.  Leland  Stanford  Jr.  Univ.  Publ.  Univ.  Ser. 

5:27-49. 
Steedman,  H.  F.    1976.    General  and  applied  data  on  formaldehyde 

fixation  and  preservation  of  marine  zooplankton,  p.  103-154.  In: 


Zooplankton  fixation  and  preservation.  H.  F.  Steedman  (ed.). 
UNESCO  Press,  Paris. 

Steftensen,  E.  1980.  Daily  growth  increments  observed  in  otoliths 
from  juvenile  East  Baltic  cod.  Dana  1:29-37. 

Stehmann,  M.,  and  W.  Lenz.  1973.  Ergebnisse  der  Forschimgsreisen 
des  FFS  "Walther  Herwig"  nach  Sudamerika.  XXVI.  Systematik 
und  Verbreitung  der  Artengruppe— 5er/o/e//a  punctata  (Schneider, 
1801),  S.  porosa  Guichenot,  1848,  5.  dobula  Gunther,  1869)- 
sowie  taxonomische  Bemerkungen  zu  Hyperoglyphe  Gunther,  1859 
und  Schedophilus  Cocco,  1839  (Osteichthyes,  Stromateoidei,  Cen- 
trolophidae).  Arch.  Fischereiwiss.  23:179-201. 

Stehr,  C.  M.  1979.  The  development  of  the  hexagonal  surface  struc- 
ture of  the  C-O  sole  {Pleuronichthys  coenosus).  Am.  Zool.  19:976. 

.    1982.    The  development  of  the  hexagonally  structured  egg 

envelope  of  the  C-O  sole  (Pleuronichthys  coenosus).  Unpubl.  MS 
thesis,  Univ.  Washington,  Seattle. 

,  and  J.  W.  Hawkes.    1979.   The  comparative  ultrastructure  of 

the  egg  membrane  and  associated  pore  structures  in  the  starry  floun- 
der, Platichthys  stellatus  (Pallas),  and  pink  salmon,  Oncorhynchus 
gorbuscha  (Walbaum).  Cell  Tissue  Res.  202:347-356. 

Stein,  D.  L.  1978.  A  review  of  the  deepwater  Liparidae  (Pisces)  from 
the  coast  of  Oregon  and  adjacent  waters.  Occ.  Pap.  Calif  Acad. 
Sci.  No.  127. 

.  1 980a.  Aspects  of  reproduction  of  liparid  fishes  from  the  con- 
tinental slope  and  abyssal  plain  off"  Oregon,  with  notes  on  growth. 
Copeia  1980:687-699. 

.    1980b.    Description  and  occurrence  of  macrourid  larvae  and 

juveniles  in  the  northeast  Pacific  Ocean  off  Oregon,  U.S.A.  Deep- 
Sea  Res.  27A:889-900. 

,  AND  C.  E.  Bond.    1978.   A  new  deep-sea  fish  from  the  eastern 

North  Pacific  Psychrolutes  phrictus  (Pisces:  Cottidae  [Psychroluti- 
dae]).  Nat.  Hist.  Mus.  Los  Ang.  Cty.  Contrib.  Sci.  296:1-9. 

,  AND  W.  G.  Pearcy.    1982.   Aspects  of  reproduction,  early  life 

history,  and  biology  of  macrourid  fishes  off"  Oregon,  U.S.A.  Deep- 
SeaRes.  29:1313-1300. 

Stein,  R.  1972.  Identification  of  some  larval  Pacific  cottids.  Unpubl. 
MS  Thesis.  Humboldt  State  Coll.,  Areata,  CA. 

.    1973.    Description  of  laboratory-reared  larvae  of  0%ocoHitf 

macutosus  GmYd  (Pisces:  Cottidae).  Copeia  1973:373-377. 

Stephens,  J.  S.  1963.  A  revised  classification  of  the  blennioid  fishes 
of  the  American  family  Chaenopsidae.  Univ.  Calif  Publ.  Zool.  68. 

.    1970.   Seven  new  chaenopsid  fishes  from  the  western  Atlantic. 


Copeia  1970:280-309. 
— ,  E.  S.  Hobson  and  R.  K.  Johnson. 


1 966.   Notes  on  distribution. 


behavior,  and  morphological  variation  in  some  chaenopsid  fishes 
from  the  tropical  eastern  Pacific,  with  descriptions  of  two  new 
species,  Acanthemblemaria  castroi  and  Coralliozetus  springeri.  Co- 
peia 1966:424-438. 
— ,  AND  V.  G.  Springer.    1974.    Clinid  fishes  of  Chile  and  Peru, 


with  description  of  a  new  species,  Myxodes  ornattis  from  Chile. 
Smithson.  Contrib.  Zool.  1 59. 

Stephenson,  A.  B.  1977.  Metavelifer  mulliradiatus  (Regan,  1907) 
(Pisces:  Veliferidae),  a  new  record  from  New  Zealand  waters.  Rec. 
Auckl.  Inst.  Mus.  14:143-144. 

Sterba,  G.  1962.  Die  Schmerlenartigen  (Cobitidae),  p.  201-234.  In: 
Handbuch  der  Binnenfischerei  Mitteleuropas.  3B(9).  R.  Demoll  and 
H.  N.  Maier  (eds.).  E.  Schweizerbart'sche  Verlagsbuchhandlung, 
Stuttgart,  West  Germany. 

,  andH.  MOller.    1962.   Elektronenmikroskopische  untersuch- 

ungen  iiber  Bildung  und  Struktur  der  Eihiillen  bei  Knochenfischen, 
I.  Die  Hiillen  junger  Oocyten  von  Cynolebias  bellotttt.  Steindachner 
(Cyprinodontidae).  Zool.  Jb.  Abt.  Anat.  80(l):65-80. 

Stevens,  E.  G.,  and  H.  G.  Moser.  1982.  Observations  on  the  early 
life  history  of  the  mussel  blenny,  Hypsoblennius  jenkinsi.  and  the 
bay  blenny,  Hypsoblennius  gentilis,  from  specimens  reared  in  the 
laboratory.  Calif.  Coop.  Oceanic  Fish.  Invest.  Rep.  23:269-275. 

Steyskal,  G.  C.  1980.  The  grammar  of  family-group  names  as  ex- 
emplified by  those  of  fishes.  Proc.  Biol.  Soc.  Wash.  93:168-177. 

Stiassny,  M.  L.  J.    1981.    The  phyletic  status  of  the  family  Cichlidae 


LITERATURE  CITED 


739 


(Pisces,  Perciformes):  A  comparative  anatomical  investigation.  Neth. 

J.  Zool.  31(2);275-314. 
Strasburg,  D.  W.    1 960.  A  new  Hawaiian  engraulid  fish.  Pac.  Sci.  14: 

395-399. 
.    1961.   Larval  carapid  fishes  from  Hawaii,  with  remarks  on  the 

ecology  of  adults.  Copeia  1961:478-480. 
.    1962.    Pelagic  stages  of  Za«c/«i  canMce^w  from  Hawaii.  Co- 
peia 1962:344-345. 
.    1964.    Postlarval  scombroid  fishes  of  the  genera  ^camAocyt- 

ium.  Nealotus  and  Diplospinus  from  the  Central  Pacific  Ocean.  Pac. 

Sci.  18:174-185. 
-.    1965.    Description  of  the  larva  and  familial  relationships  of 


the  fish  Snyderidia  camna.  Copeia  1965:20-24. 

Stroud,  G.  J.,  B.  Goldman,  F.  H.  Talbot  and  W.  Gladstone.  MS. 
Development,  growth  and  age  determination  of  the  short-nosed 
pufferfish  Canthigaster  valenlini  (Pisces:  Tetraodontidae). 

Struhsaker,  P.,  and  J.  H.  Uchiyama.  1976.  Age  and  growth  of  the 
nehu  Slolephorus  purpureus  (Pisces:  Engraulidae)  from  the  Hawai- 
ian Islands  as  indicated  by  daily  growth  mcrements  of  sagittae. 
Fish.  Bull.  U.S.  74:9-17. 

Suarez,  S.  S.  1975.  The  reproductive  biology  of  Og//Wa  cayoruwj,  a 
viviparous  brotulid  fish.  Bull.  Mar.  Sci.  25:143-173. 

Sudarsan,  D.  1968a.  On  the  early  development  of  the  pipe-fish  Sy«- 
gnathoides  biaculealus  (Bloch).  J.  Mar.  Biol.  Assoc.  India  8:222- 
224. 

Sudarsan,  D.  1968b.  Eggs  and  larvae  of  a  hemirhamphid  fish  from 
Mandapam.  J.  Mar.  Biol.  Assoc.  India  8:342-346. 

SuLAK,  K.  J.  1977.  The  systematics  and  biology  of  Sar/!yp?eroK  (Pisces, 
Chlorophthalmidae)  with  a  revised  classification  of  benthic  myc- 
tophiform  fishes.  Galathea  Rep.  14:49-108. 

SuLYA,  L.  L.,  B.  E.  Box  AND  G.  GuNTER.  1960.  Distribution  of  some 
blood  constituents  in  fishes  from  the  Gulf  of  Mexico.  Am.  J.  Phys- 
iol. 199:1177-1180. 

SuMiDA,  B.,  E.  H.  Ahlstrom  and  H.  G.  Moser.  1979.  Early  devel- 
opment of  seven  flatfishes  of  the  eastern  North  Pacific  with  heavily 
pigmented  larvae  (Pisces.  Pleuronectiformes).  Fish.  Bull.  U.S.  77: 
105-145. 

,  AND  H.  G.  MosER.    1980.    Food  and  feeding  of  Pacific  hake 

larvae,  Merluccius  productus.  off' southern  California  and  northern 
Baja  California.  Calif  Coop.  Oceanic  Fish.  Invest.  Rep.  21:161- 
171. 

Sunny,  K.  G.  1971.  Morphology  of  the  vertebral  column  of  Mugil 
macrolepis  (Smith).  Bull.  Dep.  Mar.  Biol.  Oceanogr.  Univ.  Cochin 
5:101-108. 

Suzuki,  K.  1962.  Anatomical  and  taxonomical  studies  on  the  carangid 
fishes  of  Japan.  Rep.  Fac.  Fish.  Prefect.  Univ.  Mie  4(2):43-232. 

,  AND  S.  HioKL    1978.    Spawning  behavior,  eggs,  and  larvae  of 

the  sea  bream,  Gymnocranius  griseus,  iiran  aquarium.  Jpn.  J.  Ich- 
thyol.  24:271-277. 

,  AND .    1979a.    Morphology  and  phyletic  aspects  of  the 

newly  hatched  larvae,  with  special  references  to  those  of  family 
Serranidae.  Kaiyo  Kaguku  (Mar.  Sci.)  11:1 17-125.  (In  Japanese). 

,  AND .    1979b.    Spawning  behavior,  eggs,  and  larvae  of 

the  lutjanid  fish,  Liiljanus  kasmira.  in  an  aquarium.  Jpn.  J.  Ich- 
thyol.  26:161-166. 

,  ,  K.  KoBAYASHi  AND  Y.  Tanaka.     1981.    Developing 

eggs  and  early  larvae  of  the  wxasses,  Cirrhilabrus  temminckii  and 
Labroides  diinidiatus.  with  a  note  on  their  spawning  behaviors.  J. 
Fac.  Mar.  Sci.  Technol.  Tokai  Univ.  14:369-372.  (In  Japanese, 
English  Abstr.) 

, ,  Y.  Tanaka  andK.  IwASA.    1979.   Spawning  behaviour, 

eggs,  larvae,  and  sex  reversal  of  two  Pomacanthine  fish,  Genican- 
ihus  leinarck  and  G.  semifascialus.  in  the  aquarium.  J.  Fac.  Mar. 
Sci.  Technol.  Tokai  Univ.  12:149-165. 

,  K.  KoBAYASHi,  S.  HioKj  AND  T.  SAKAMOTO.    1974.   Ecological 

studies  of  the  anthiine  fish  Sacura  margahtacea  in  Suruga  Bay. 
Jpn.  J.  Ichthyol.  21:21-33. 

, , AND .    1978.    Ecological  studies  of  the 


anthiine  fish,  Franzia  squamipinnis  in  Suruga  Bay,  Japan.  Jpn.  J. 
Ichthyol.  25:124-140.  (In  Japanese,  English  abstract.) 

— , , AND .   In  press.   Ecological  studies  of  the 

anthiine  fish,  Pseudanthias  elongatus.  in  Suruga  Bay,  Japan. 

— ,  H.  Syozo,  Y.  Tanaka  and  H.  Kjtazawa.  1983.  Spawning 
and  early  life  history  of  Hapalogenys  mucronatus  (Eydoux  et  Sou- 
leyet)  (Pisces:  Pomadasyidae)  in  an  aquarium.  J.  Fac.  Mar.  Sci. 
Technol.  Tokai  Univ.  16:183-191. 

— ,  Y.  Tanaka  and  S.  Hioki.  1 980.  Spawning  behaviour,  eggs, 
and  larvae  of  the  butterflyfish,  Chaetodon  nippon,  in  an  aquarium. 
Jpn.  J.  Ichthyol.  26:334-341. 

— ,  ,  AND  Y.  Shiobara.  1980.  Studies  on  repro- 
duction and  larval  rearing  of  coastal  marine  fishes,  p.  53-82.  In: 
Research  in  a  large-scale  culture  of  marine  fisheries  resources.  G. 
Yamamoto  (ed.).  Inst.  Oceanic  Res.  and  Dev.,  Tokai  Univ.,  Shim- 
izu.  (In  Japanese.) 

— , , AND  S.  Ueno.    1981.    Spawning  behavior  and 


early  life  history  of  Hypodytes  rubripinnis  (Pisces:  Congiopodidae) 
in  an  aquarium.  J.  Fac.  Oceanogr.  Tokai  Univ.  14:357-367. 

Suzuki,  R.  1955.  Note  on  the  interfamiliar  hybrid  larvae  between 
mud  loach  (Misgitrnus  anguitlicaudatus)  and  Crucian  carp  (Car- 
assius  carassius)  or  goldfish  (Carassius  auratus).  Jpn.  J.  Ichthyol. 
4:50-58. 

.  1968.  Hybridization  experiments  in  cyprinid  fishes.  XI.  Sur- 
vival rate  of  F,  hybrids  with  special  reference  to  the  closeness  of 
taxonomical  position  of  combined  fishes.  Bull.  Freshwater  Fish. 
Res.  Lab.  (Tokyo)18(2):l  13-155. 

Sverdrup,  H.  U.,  M.  W.  Johnson  and  R.  H.  Fleming.  1942.  The 
Oceans.  Their  physics,  chemistry,  and  general  biology.  Prentice- 
Hall,  Inc.,  Englewood  Cliffs,  NJ. 

SvETOviDOv,  A.  N.  1937.  Uber  die  Klassifikation  der  Gadiformes  oder 
Anacanthini.  Bull.  Acad.  Sci.  URSS,  p.  1281-1288. 

.   1948.  Treskoobraznye  (Gadiformes).  Akad.NaukSSSR,  Zool. 

Inst.,  Fauna  SSSR  9(4),  N.S.  34.  (In  Russian,  transl.  by  Isr.  Program 
Sci.  Transl.,  1962.  Avail,  from:  NTIS,  Springfield,  VA.,  OTS  63- 
11071.) 

.     1956.    Morphological  principles  of  the  classification  of  the 

Gadidae.  Proc.  Int.  Congr.  Zool.  14(1953):535-540. 

.    1963.    Fauna  of  U.S.S.R.  Fishes.  Vol.  2,  No.  1.  Clupeidae. 

Transl.  from  Russian  by  Isr.  Program  Sci.  Transl.  OTS  61-1 1435. 
(First  published  in  1952,  Zool.  Inst.  Akad.  Nauk  SSSR,  N.S.  48.) 

.    1969.   On  the  taxonomical  position  of  the  genus  £Mc/;c/i;/iyi 

(Pisces,  Gadiformes).  Zool.  Zh.  48:1824-1831. 
-.    1982.    Taxonomic  differences  between  red  hake.  Urophycis 


chuss,  and  U.  tenuis  (Gadidae).  J.  Ichthyol.  (Engl.  Transl.  Vopr. 

Ikhtiol.)  22(3):  150-1 53. 
SwARUP,  H.    1958.   Stages  in  the  development  of  the  stickleback  Gar- 

terosteus  oculealus  (L.).  J.  Embryol.  Exp.  Morphol.  6(3):373-383. 
Takagi,  K.    1950.    On  the  glossohyal  bone  of  the  gobioid  fishes  of 

Japan,  with  some  phylogenetic  considerations.  Jpn.  J.  Ichthyol. 

1:37-52.  (In  Japanese.) 
.    1953.    A  study  on  the  scale  of  the  gobiid  fishes  of  Japan.  J. 

Tokyo  Univ.  Fish.  39:231-253. 
-.    1966.    Distributions  and  ecology  of  the  gobioid  fishes  in  the 


Japanese  waters.  J.  Tokyo  Univ.  Fish.  52:83-127. 

Takai,  T.,  and  T.  Fukunaoa.  1971.  The  life  history  of  a  ovo- vivip- 
arous scorpaenoid  fish,  Sebasles  longispinis  (Matsubara).  1 .  Egg  and 
larval  stages.  J.  Shimonoseki  Univ.  Fish.  20:91-95. 

,  and  a.  Mizokami.   1959.  On  the  reproduction,  eggs,  and  larvae 

of  the  pipefish,  Synagnathus  schlegeli  Kaup.  J.  Shimonoseki  Coll. 
Fish.  8:85-89. 

AND  T.  YosHioKA.    1979.    The  life  history  of  the  dragonet. 


CaUionymus  beniteguri  in  the  Seto  Inland  Sea.  I.  Egg,  larva  and 
juvenile.  J.  Shimonoseki  Univ.  Fish.  27:147-154. 

Takita,  T.  1966.  Egg  development  and  larval  stages  of  the  small 
clupeoid  fish,  Harengula  -unasi  Bleeker  and  some  information  about 
the  spawning  and  nursery  in  Ariake  Sound.  Bull.  Fac.  Fish.  Na- 
gasaki Univ.  21:171-179. 

.    1967.   The  spawning  and  early  hfe  history  of  the  engraulid  fish 


740 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Coilia  sp.  distributed  in  Ariake  Sound.  Bull.  Fac.  Fish.  Nagasaki 

Univ.  23:107-122. 
— .    1980.   Embryonicdevelopment  and  larvae  of  three  dragonets. 

Bull.  Jap.  See.  Sci.  Fish.  46:1-7. 
— .    1983.    Embryonic  and  larval  development  of  the  callionymid 


fish,  Callionymus  calUste.  Jpn.  J.  Ichthyol.  29:441-445. 
— ,  AND  S.  FujiTA.    1964.    Egg  development  and  prolarval  stages 

of  the  turbot,  Pleuronichthys  cornulus  (Temminck  et  Schlegel).  Bull. 

Jpn.  Soc.  Sci.  Fish.  30:613-618.  (In  Japanese,  English  summary.) 
— , andY.  DoTSu.    1967.   Egg  development  and  pro-larval 

stages  of  flatfish,  Verasper  variegalus  (Temminck  et  Schlegel).  Bull. 

Fac.  Fish.  Nagasaki  Univ.  23:101-106. 
,  AND  E.  Okamoto.    1979.    Spawning  behavior  of  the  two  dra- 


gonets, Callionvmus  flagris  and  C  rkhardsoni  in  the  aquarium. 

Jpn.  J.  Ichthyol.  26:282-288. 
Taliev,  D.  N.    1955.    Bullheads  and  sculpins  (Cottoidei)  from  Lake 

Baikal.  Akad.  Nauk.  SSSR.  (In  Russian.) 
Talwar,  p.  K.    1968.   Studies  on  the  biology  of  Hemirhamphus  mar- 

ginatus  (Forskal)  (Hemirhamphidae-Pisces).  J.  Mar.  Biol.  Assoc. 

India  9:61-69. 
TAning,  a.  V.     1918.     Mediterranean  Scopelidae  (Saurus  aulopus. 

Chlorophthalmus.  and  Myctophum).  Rep.  Dan.  Oceanogr.  E.xped. 

Mediterr.  2(A.7). 

.    1923.   LopA/us.  Rep.  Dan.  Oceanogr.  Exped.  Mediterr.  2( A.  10). 

.   1931.  PostlarvalstagesofBa//i.v/agi«  from  the  North  Atlantic. 

Vidensk.  Medd.  Dann.  Naturhist.  Foren.  Khobenhavn  92:269-274. 
.    1936.    On  the  eggs  and  young  stages  of  the  halibut.  Medd. 

Komm.  Havunders.  Ser.  Fiskeri  10(4). 
.    1952.  Experimental  study  of  meristic  characters  in  fishes.  Biol. 

Rev.  27:169-193. 
.    1955.  On  the  breeding  areas  of  the  swordfish  (Xiphias).  Deep- 

Sea  Res.  3  (Suppl.):438-450. 
.    1961.    Larval  and  postlarval  stages  of  Sebastes  species  and 

Heticolenus daciylopterus.  Rapp.  P.V.  Reun.  Cons.  Int.  Explor.  Mer 

150:234-240. 
Taranets,  S.  Y.    1941.   On  the  classification  and  ongm  of  the  family 

Cottidae.   Izv.   Akad.   Nauk  SSSR  Otd.   Biol.    194l(3):427^47. 

(Translated  from  Russian,  Univ.  Brit.  Columbia  Mus.  Contrib.  5.) 
Taranetz,  A.  Y.    1935.   Briefsurvey  of  species  of  the  genus  forofo»M.s. 

Bull.  Far  Eastern  Branch  Acad.  Sci.  USSR  14:177-180. 
.    1937.   Handbook  for  identification  of  fishes  of  Soviet,  Far  East 

and  adjacent  waters.  Bull.  Pac.  Sci.  Inst.  Fish.  Oceanogr.  11(11):!- 

200. 
Tarp,   F.   H.     1952.     A  revision  of  the   family  Embiotocidae  (the 

surfperches).  Calif  Dep.  Fish  Game  Fish  Bull.  88:1-99. 
Ta VOLGA,  W.  N.     1949.     Embryonic  development  of  the  platyfish 

(Platypoecilius)  and  swordtail  (Xiphophorus)  and  their  hybnd.  Bull. 

Am.  Mus.  Nat.  Hist.  94:165-229. 
Taylor,  J.  N.    1983.    Field  observations  on  the  reproductive  ecology 

of  three  species  of  armored  catfishes  (Loricariidae:  Loricariinae)  in 

Paraguay.  Copeia  1983:257-259. 
Taylor,  M.  N.,  L.  DiMichele  and  G.  J.  Leach.    1977.    Egg  stranding 

in  the  life  cycle  of  the  mummichog.  Fundulus  heterocUtus.  Copeia 

1977:397-399. 
Taylor,  W.  R.    1964.    Fishes  of  Arhnem  Land.  Rec.  Am.  Aust.  Sci. 

Exped.  Amhem  Land.  4:45-307. 
.    1967.   An  enzyme  method  of  clearing  and  staining  small  ver- 
tebrates. Proc.  U.S.  Natl.  Mus.  122(3596):  1-1 7. 
.    1977.    Observations  on  specimen  fixation.  Proc.  Biol.  Soc. 


Wash.  90:753-763. 

,  AND  G.  C.  Van  Dyke.  MS.  Staining  and  clearing  small  ver- 
tebrates for  bone  and  cartilage  study. 

Templeman,  W.  1 948.  The  life  history  of  the  capelin,  Malhtus  villosus 
in  Newfoundland  waters.  Res.  Bull.  Div.  Fish.  Res.  Newfoundland 
17:1-151. 

.    1968.   A  review  of  the  morid  fish  genus //a/a/-g3T««  with  first 

records  from  the  western  North  Atlantic.  J.  Fish.  Res.  Board  Can. 
25:877-901. 

Terner,  C.    1979.   Metabolism  and  energy  conversion,  p.  261-278. /«. 


Fish  Physiology.  Vol.  8.  W.  S.  Hoar,  D.  J.  Randall  and  J.  R.  Brett 

(eds.).  Academic  Press,  New  York. 
Thangaraja,  M.,  and  K.  Ramamoorthl    1980.    Early  life  history  of 

Anadontostoma  chacunda  (Ham.-Buch.)  from  Porto  Novo.  Indian 

J.  Mar.  Sci.  9:136-139. 
Theilacker,  G.  H.    1978.    Effect  of  starvation  on  the  histological  and 

morphological  characteristics  of  jack  mackerel   Trachurus  sym- 

metncus  larvae.  Fish.  Bull.  U.S.  76:403-414. 
.     1980a.    Changes  in  body  measurements  of  larval  northern 

anchovy  Engraulis  mordax  and  other  fishes  due  to  handling  and 

preservation.  Fish.  Bull.  U.S.  78:685-692. 
.    1980b.    Rearing  container  size  affects  morphology  and  nutri- 
tional condition  of  larval  jack  mackerel,  Trachurus  svmmetricus. 

Fish  Bull.  U.S.  78:789-791. 
,  AND  K.  Dorsey.    1980.    Larval  fish  diversity,  a  summary  of 

laboratory  and  field  research.  IOC  Workshop  Rep.  No.  28:105- 

142. 
,  andM.  F.  McMaster.    1971.   Mass  culture  of  the  rotifer  Brac-/;- 


ionus  plicatiHs  and  its  evaluation  as  a  food  for  larval  anchovies. 
Mar.  Biol.  10:183-188. 

Thibault,  R.  E.,  and  R.  J.  ScHiiLTZ.  1978.  Reproductive  adaptation 
among  viviparous  fishes  (Cyprinodontiformes:  Poeciliidae).  Evo- 
lution 32:320-332. 

Thomopoulos,  A.  1954.  Sur  quelques  oeufs  planctoniques  de  teleos- 
teens  de  la  Baie  de  Villefranche.  Bull.  Inst.  Oceanogr.  (Monaco) 
1043:1-15. 

.    1956.    Sur  quelques  oeufs  planctoniques  de  teleosteens  de  la 

Baie  de  Villefranche.  II.  Peches  de  mois  de  Septembre.  Bull.  Inst. 
Oceanogr.  Monaco  1072. 

Thompson,  W.  1840.  On  a  new  genus  of  fishes  from  India.  Charles- 
worth's  Mag.  Nat.  Hist.  4:184-187. 

Thompson,  W.  F.  1916.  Fishes  collected  by  the  United  States  Bureau 
of  Fisheries  steamer  "Albatross"  during  1888,  between  Montevi- 
deo, Uruguay,  and  Tome,  Chile,  on  the  voyage  through  the  Straits 
of  Magellan.  Proc.  U.S.  Natl.  Mus.  50(2 133):40 1^76. 

,  and  J.  B.  Thompson.    1919.  The  spawning  of  the  grunion  {Leu- 

resthes  tenuis).  Calif  Fish  Game  Fish  Bull.  3:1-29. 

,  and  R.  Van  Cleve.    1936.    Life  history  of  the  Pacific  halibut. 

2.  Distribution  and  early  life  history.  Rep.  Int.  Fish.  Comm.  9. 

,  F.  H.  Bell,  H.  A.  Dunlop,  L.  P.  Schultz  and  R.  Van  Cleve. 


1936.  The  spawning  of  the  silver  smelt  Hypomesus  pretiosus.  Ecol- 
ogy 17:158-168. 

Thomson,  G.  M.  1906.  The  Portobello  marine  fish-halchery  and  bi- 
ological station.  Trans.  N.Z.  Inst.  38  (Art.  6 1):  1-31. 

,  AND  T.  Anderton.    1921.    History  of  the  Portobello  Marine 

Fish-Hatchery  and  Biological  Station.  Bull.  Board  Sci.  Art  N.Z. 
No.  2. 

Thomson,  J.  M.  1954.  The  Mugilidae  of  Australia  and  adjacent  seas. 
Aust.  J.  Mar.  Freshwater  Res.  5:70-131. 

.    1966.   The  grey  mullets.  Oceanogr.  Mar.  Biol.  Annu.  Rev.  4: 

301-355. 

.  1975.  The  dentition  of  grey  mullets.  Aquaculture  5:100.  (Ab- 
stract.) 

.    1982.  The  taxonomy  of  grey  mullets,  p.  1-63. /«;  Aquaculture 

of  grey  mullets.  O.  H.  Oren  (ed.).  Cambridge  Univ.  Press,  Cam- 
bridge. 

,  AND  A.  E.  Bennett.    1953.    The  oyster  blenny.  Omobranchus 


anclius  (Valenciennes)  (Blenniidae).  Aust.  J.  Mar.  Freshwater  Res. 

4:227-233. 
Thorp,  C.  H.     1969.    A  new  species  of  mirapinniform  fish  (family 

Kasidoridae)  from  the  western  Indian  Ocean.  J.  Nat.  Hist.  3:61- 

70. 
Thorsteinsson,  v.     1981.    The  aging  validation  of  the  lumpsucker 

(Cycloplerus  lumpus)  and  the  age  composition  of  the  lumpsucker 

in  Icelandic  lumpsucker  fisheries.  Int.  Counc.  Explor.  Sea  CM. 

I98I/G58. 
Thresher,  R.  E.,  AND  E.  B.  Brothers.   In  press.   Reproductive  ecology 

and  biogeography  of  Indo-west  Pacific  angelfishes.  (Pisces:  Poma- 

canthidae).  Evolution. 


LITERATURE  CITED 


741 


Tin,  H.  T.  1982a.  Family  Clupeidae,  herrings,  p.  64-73.  In:  Identi- 
fication of  larval  fishes  of  the  Great  Lakes  basin  with  emphasis  on 
the  Lake  Michigan  drainage.  N.  A.  Auer  (ed.).  Great  Lakes  Fish. 
Comm.  Spec.  Publ.  82-3,  Ann  Arbor,  MI. 

.  1982b.  Family  Osmendae,  smelts,  p.  146-151.  In:  Identifi- 
cation of  larval  fishes  of  the  Great  Lakes  basin  with  emphasis  on 
the  Lake  Michigan  drainage.  N.  A.  Auer  (ed.).  Great  Lakes  Fish. 
Comm.  Spec.  Publ.  82-3,  Ann  Arbor.  MI. 

.    1982c.    Family  Ictaluridae,  bullhead  catfishes,  p.  436-457. /?!.- 

Identification  of  larval  fishes  of  the  Great  Lakes  basin  with  emphasis 
on  the  Lake  Michigan  drainage.  N.  A.  Auer  (ed.).  Great  Lakes  Fish. 
Comm.  Spec.  Publ.  82-3,  Ann  Arbor,  Ml. 

Tokuv.«,,  K.,  and  K,  Amaoka.  1980.  Studies  on  larval  and  juvenile 
blennies  in  the  coastal  waters  of  southern  Hokkaido.  Bull.  Fac. 
Fish.  Hokkaido  Univ.  3 1(1):  16-49. 

ToMiNAGA,  Y.  1965.  The  internal  morphology  and  systematic  position 
of  Leptohrama  mulleri.  formerly  included  in  the  family  Pempher- 
ididae.  Jpn.  J.  Ichthyol.  12:33-56. 

TONG,  L.  J.,  AND  R.  Saito.  1977.  Further  postlarvae  of  the  tarakihi 
in  New  Zealand  waters.  N.Z.  J,  Mar.  Freshwater  Res.  1 1:159-162. 

ToRNO,  A.  E.,  A.  P.  ToMO  and  E.  Marschoff.  1977.  Una  nueva 
especie  del  genero  Oidiphorus  McAllister  y  Rees  (Pisces,  Zoarcidae) 
capturado  en  la  adyacencia  de  las  Islas  Sandwich  del  Sur.  Contrib. 
Inst.  Antarct.  Argent.  223:1-10. 

Tortonese,  E.  1956.  Plectognathi,  p.  960-977. /n.Uova,  larvee  stadi 
giovanili  di  Teleostei.  Fauna  Flora  Golfo  Napoli  Monogr.  38. 

.    1973.    Serranidae,  p.  355-362.  In:  Checklist  of  the  fishes  of 

the  north-eastern  Atlantic  and  of  the  Mediterranean.  J.  C  Hureau 
and  Th.  Monod  (eds.).  UNESCO,  Pans. 

Townsend,  D.  W.,  andJ.  J.  Graham.  1981.  Growth  and  age  structure 
of  larval  Atlantic  herring,  Clupea  harengus.  in  the  Sheepscot  River 
estuary,  Maine,  as  determined  by  daily  growth  increments  in  oto- 
liths. Fish.  Bull.  U.S.  79:123-130. 

ToYOSHiMA,  M.  1981.  Redescription  of  the  zoarcid  fish  Lycozoarces 
regam.  Jpn.  J.  Ichthyol.  27:296-300. 

Tranter,  D.J.  (ed.).  1968.  Zooplankton  sampling.  UNESCO  Monogr. 
Oceanogr.  Methodol.  2. 

Trewavas,  E.  1932.  A  contnbution  to  the  classification  of  the  fishes 
of  the  order  Apodes,  based  on  the  osteology  of  some  rare  eels.  Proc. 
Zool,  Soc.  Lond.  1932:639-659. 

Troemner.  J.  L.  1932.  Recent  importations-Fundulus  cubensis. 
Aquarium,  Philadelphia  1(3):70,  82. 

.  1941.  Cubanichlhys  cubensis  Eigenmann.  Aquarium,  Phila- 
delphia 10(81:129-130. 

Trott,  L.  B.  1970.  Contributions  to  the  biology  of  carapid  fishes 
(Paracanthopterygii:  Gadiformes).  Univ.  Calif  Publ.  Zool.  89:1- 
60. 

.    1981.   A  general  review  of  the  pearlfishes  (Pisces,  Carapidae). 

Bull.  Mar.  Sci.  31:623-629. 

Trunov,  I.  A.  1981.  Parakumba  maculisquama  (New  genus  and 
species)  family  Macrouridae  from  the  South  Atlantic.  J.  Ichthyol, 
(Engl,  Trans.  Vopr.  Ikhtiol.)  21(l):27-34. 

.    1982.   Species  of  the  family  Regalecidae  (Lampnformes)  from 

the  Southeastern  Atlantic  Ocean.  J.  Ichthyol.  (Engl.  Transl.  Vopr. 
Ikhtiol.)  22(1):  I -6. 

Tsinovskv,  V.  D.,  AND  I.  A.  Mel'nikov.  1980.  On  the  discovery  of 
Liparis  koefoedi  (Liparidae,  Osteichthyes)  in  the  waters  of  the  cen- 
tral Arctic  Basin,  p.  211-214.  In:  Biology  of  the  Central  Arctic 
Basin.  Nauka  Publ.,  Moscow. 

Tsokur,  A.  G.  1975.  Data  on  the  development  of  the  Malayan  Head- 
light fish,  Diaphus  malayanus  (Myctophidae,  Pisces).  J.  Ichthyol. 
(Engl.  Transl.  Vopr.  Ikhtiol.)  15:611-622. 

.     1977.    Larvae  of  carangid  fishes  (Carangidae.  Pisces)  of  the 

Arabian  Sea.  In:  Ichthyoplankton  of  tropical  waters  of  the  world 
ocean.  Acad.  Sci.  U.S.S.R.,  Proc.  P.  P.  Shirshov  Inst.  Oceanol.  109: 
149-227.  (In  Russian.) 

.    1981.   The  \arvae  of  Benthosema  pterota  (McoQk.  \%9\)Myc- 

tophidae,  from  the  Arabian  Sea.  J.  Ichthyol.  (Engl.  Transl.  Vopr. 
Ikhtiol.)  21(3):38-53. 


Tsukahara,  H.  1961.  Biology  of  the  cutlassfish,  Trichiurus  lepturus 
Linaeus.  Pan  1 .  Early  life  history.  Rec.  Oceanogr.  Works  Jpn.  Spec. 
No,  5:117-121. 

.     1962.    Studies  on  habits  of  coastal  fishes  in  the  Amakusa 

Islands.  Part  2.  Early  life  history  of  the  rockfish.  Sebastiscus  mar- 
moratus  (Cuvier  and  Valenciennes).  Rec.  Oceanogr.  Works  Jpn., 
Spec.  No.  6:49-55. 

,  and  T.  Shiokawa.    1957.    Studies  on  the  flying-fishes  of  the 

Amakusa  Islands.  Pari  2.  The  life  history  and  Yvab\l%of  Parexocoetus 
mento  (Cuvier  et  Valenciennes).  Sci.  Bull.  Fac.  Agri.  Kyushu  Univ. 
16(2):275-286.  (In  Japanese,  English  summary.) 

Tucker,  D.W.  1954.  Report  on  the  fishes  collected  by  S.Y."Rosaura" 
in  the  North  and  Central  Atlantic,  1937-38.  Part  IV.  Families 
Carcharhinidae,  Torpedinidae,  Rosauridae  (nov.),  Salmonidae, 
Alepocephalidae,  Searsidae,  Clupeidae.  Bull.  Br.  Mus.  (Nat.  Hist.) 
2(6):  163-2 14. 

.  1956.  Studies  on  the  Trichiuroid  fishes.  3.  A  preliminary- 
revision  of  the  family  Trichiuridae.  Bull.  Brit.  Mus.  (Nat.  Hist.) 
Zool.  4(3):73-130. 

Tucker,  J.  W.,  Jr.  1982.  Larval  development  of  Citharichthys  cor- 
nulus,  C.  gymnorhinus.  C.  spilopterus.  and  Elropuscrossotus(SoXh- 
idae),  with  notes  on  larval  occurrence.  Fish.  Bull.  U.S.  80:35-73. 

,  and  A.  J.  Chester.  In  press.  Effects  of  salinity,  formalin  con- 
centration, and  buffer  on  quality  of  preservation  of  southern  floun- 
der {Paralichlhys  lethostigma)  larvae.  Copeia. 

Tung,  I.-H.  1973.  On  the  egg  development  and  larval  stages  of  the 
grey  mullet,  Mugil  cephalus  Linnaeus.  Rep.  Inst.  Fish.  Biol.  Taipei 
3(1):187-210.  (In  Chinese,  English  summary.) 

Turner,  C  H.,  and  E.  E.  Ebert.  1962.  The  nesting  of  C/jrom/i  ;7m«c- 
tipinnis  (Cooper)  and  a  description  of  their  eggs  and  larvae.  Calif 
Fish  Game  48(4):243-248. 

Turner,  C.  L.  1937.  Reproductive  cycles  and  superfetation  in  poe- 
ciliid  fishes.  Biol.  Bull.  (Woods  Hole)  72:145-164. 

.    1940a.   Superfetation  in  viviparous  cyprinodont  fishes.  Copeia 

1940:88-91. 

.     1940b.    Pseudoamnion,  pseudochorion  and  follicular  pseu- 

doplacenta  in  poeciliid  fishes.  J.  Morphol.  67:59-89. 

.    1940c.    Follicular  pseudoplacenta  and  gut  modifications  in 

anablepid  fishes.  J.  Morphol.  67:91-105. 

.  1940d.  Adaptations  for  viviparity  in  jenynsiid  fishes.  J.  Mor- 
phol. 67:291-297. 

Tyler,  J.  C  1968.  A  monograph  on  plectognath  fishes  of  the  super- 
family  Tnacanthoidea.  Monogr.  Acad.  Nat.  Sci.  Phila.  16. 

.    1970.   Osteological  aspects  of  interrelationships  of  surgeon  fish 

genera  (Acanthuridae).  Proc.  Acad.  Nat.  Sci.  Phila.  122(2):87-124. 
1980.    Osteology,  phylogeny,  and  higher  classification  of  the 


fishes  of  the  order  Plectognathi  (Tetradontiformes).  NOA.^  Tech. 

Rept.  NMFS  Circ.  434. 
,  I.  Nakamura  and  B.  B,  Collette.    MS.   Osteology  and  rela- 
tionships of  the  oceanic  fish  Luvanis  imperialis  (Luvaridae):  an 

acanthuroid,  not  a  scombroid. 
UcHiDA,  K.    1933.    Studies  on  the  early  life  history  of  Japanese  Het- 

erosomata.  II.  Verasper  variegalus.  Zool.  Soc.  Jpn.  45:268-277. 
.    1936.    A  note  on  the  pelagic  postlarval  stages  of  Aniigonia 

rubescens  (Pisces,  Capriformes).  Zool.  Mag.  (Tokyo)  48:935-939. 
,  Y.  DoTsu,  S.  MiTo  AND  K.  Nakahara.    1958.   The  spawning 

and  early  life  history  of  "Buri"  Japanese  yellowtail,  Seriola  quin- 

queradiata  Temminck  et  Schlegel.  Sci.  Bull.  Fac.  Agric.  Kyushu 

Univ.  16(3):329-342. 
,  S.  Imai,  S.  Mito.  S.  FujiTA.  M.  Ueno,  Y.  Shojima,  T.  Senta. 

M.  Tahuka  and  Y.  Dotsu.    1958.    Studies  on  the  eggs,  larvae, 

and  juveniles  of  Japanese  fishes.  Series  I.  Second  Laboratory  of 

Fisheries  Biology,  Fish.  Dep.,  Fac.  Agric.  Kyushu  Univ..  fukuoka, 

Japan.  (In  Japanese.) 
Ueno,  M.,  and  S.  Fujita.    1954.    On  the  development  of  the  egg  of 

Sillago  sihama{¥orsV.a\).  ix>n.  J.  Ichthyol.  3:1 18-120.  (In  Japanese, 

English  summary.) 
.AND .    1958.    Eggs,  larvae,  and  juveniles  of /'/aO'f<7'^- 

alus  indicus  (Linne)  (Platycephalidae),  p.  86.  In:  Studies  on  the 


742 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES -AHLSTROM  SYMPOSIUM 


eggs,  larvae  and  juveniles  of  Japanese  fishes.  K.  Uchida  et  al.,  (eds.). 
Fac.  Agric.  Kyushu  Univ.,  Fukuoka,  Japan.  (In  Japanese.) 

Ueno,  T.  1970.  Fauna  Japonica:  Cyclopteridae  (Pisces).  Academic 
Press  of  Japan,  Tokyo. 

Ueyanagi,  S.  1963a.  Methods  for  identification  and  discrimination 
of  the  larvae  of  five  istiophorid  species  distributed  m  the  Indo- 
Pacific.  Rep.  Nankai  Reg.  Fish.  Res.  Lab.  (I7):137-151.  (In  Jap- 
anese, English  summary.) 

.    1963b.    A  study  of  the  relationships  of  the  Indo-Pacific  isti- 

ophorids.  Rep.  Nankai  Reg.  Fish.  Res.  Lab.  (17):151-165.  (In  Jap- 
anese, English  summary.) 

.  1 974a.  On  an  additional  diagnostic  character  for  the  identi- 
fication of  billfish  larvae  with  some  notes  on  the  variations  in 
pigmentation,  p.  73-78.  In:  Proc.  Intematl.  Billfish  Symp..  Kailua- 
Kona,  Hawaii,  9-12  August  1972,  Part  2.  Review  and  contributed 
papers.  R.  S.  Shomura  and  F.  Williams  (eds.).  NOAA  Tech.  Rep. 
NMFS  SSRF  675. 

.    1974b.    Present  state  of  billfish  larval  taxonomy,  p.  649-654. 

In:  The  early  life  history  of  fish.  J.  H.  S.  Blaxter  (ed.).  Springer- 
Verlag,  Berlin. 

,  Y.  NisHiKAWA  AND  T.  Matsuoka.   1974.  Artificial  fertilization 

and  larval  development  of  skipjack  tuna,  Katsuwonus  pelamis.  Bull. 
Far  Seas  Fish.  Res.  Lab.  (Shimizu)  10:179-188. 

AND  M.  Okiyama.    1979.    Morphology  and  phylogeny  of  fish 


— .  1958.  Studies  on  fish  eggs  and  larvae  of  Madras  Coast.  Univ. 
Madras,  G.  S.  Press. 

— .  1973.  Studies  on  fish  eggs  and  larvae  from  Indian  waters.  1. 
Development  of  egg  and  larvae  of  Hirundichthys  (Hinmdichthys) 
coromandelensis  (WomeW).  Indian  J.  Fish.  20(1):108-137. 

— .    1975.   Studies  on  fish  eggs  and  larvae  from  Indian  waters.  3. 


larvae  -  scombroids.  Kaiyo  Kaguku  (Mar.  Sci.)  1 1:93-99.  (In  Jap- 
anese.) 
Ukawa,  M.,  M.  HioucHi,  AND  S.  MiTO.    1966.    Spawning  habits  and 

early  life  history  of  a  serranid  fish,  Epinephelus  akaara  (Temminck 

et  Schlegel).  Jpn.  J.  Ichthyol.  13:156-161. 
Umeda,  S.,  and  a.  Ochiai.    1975.    On  the  histological  structure  and 

function  of  digestive  organs  of  the  fed  and  starved  larvae  of  the 

yellowtail,  Seriola  quinqueradiata.  Jpn.  J.  Ichthyol.  21:213-219. 
Utter,  F.  M.    1981.    Biological  criteria  for  definition  of  species  and 

district  intraspecific  populations  of  anadromous  salmonids  under 

the  U.S.  Endangered  Species  Act  of  1973.  Can.  J.  Fish.  Aquat.  Sci. 

38:1626-1635. 
VAN  der  Veer,  J.    1982.   Simple  and  reliable  methods  for  the  fixation, 

mounting  and  staining  of  small  and  delicate  marine  plankton  for 

light  microscopic  identification.  Mar.  Biol.  (Berl.)  66:9-14. 
Varikul,  v.,  and  C.  Boonsom.    1969.    Illustrations  of  the  gross  egg 

and  larval  development  of  Pla  Swai  Pangasius  sutchi  Fowler.  Indo- 

Pac.  Fish.  Counc.  Occas.  Pap.  14:1-4. 
Vatanachi,  S.  N.    1972.   The  identification  offish  eggs  and  fish  larvae 

obtained  from  the  survey  cruises  in  the  South  China  Sea.  Proc.  1 5 

Session  Indo-Pac.  Fish.  Counc,  FAO,  Bangkok. 
Venkataramanujan,  K.,  AND  K.  Ramanoorthi.    1981.    Fish  eggs  of 

Porto  Novo  coastal  waters,  east  coast  of  India.  Rapp.  P.  V.  Reun. 

Cons.  Int.  Explor.  Mer  178:597-599. 
Vergara,  R.    1972.    Analisis  taxonomico  y  consideraciones  filogene- 

ticas  sobre  las  especies  Cubanas  del  genero  Caranx  (Teleostei,  Per- 

ciformes,  Carangidae).  Cent.  Invest.  Pesq.  Cuba,  Contrib.  34. 
.    1974.    Estudio  fenetico  de  los  Carangidos  Cubanos.  (Pisces, 

Teleostei).  Acad.  Cienc.  Cuba  Ser.  Oceanogr.  23:1-30. 
ViALLi,  M.    1937.   Famiglia  2:  Mugilidae,  p.  433-456.  In:  Uova,  larve 

e  stadi  giovanili  di  Teleostei.  Fauna  Flora  Golfo  Napoli  Monogr. 

38. 
.    1956.   Famiglia  3:  Sphyraenidae.  p.  457^61. //I.  Uova,  larve 

e  stadi  giovanili  di  Teleostei.  Fauna  Flora  Golfo  Napoli  Monogr. 

38. 
Victor,  B.  C.    1982.   Daily  growth  increments  and  recruitment  in  two 

coral-reef  wrasses,  Thalassoma  bifasciatum  and  Halichoeres  bi- 

vittatus.  Mar.  Biol.  (Bert.)  71:203-208. 
.    1983.    Recruitment  and  population  dynamics  of  a  coral-reef 

fish.  Science  (Wash.  D.C.)  219:419-420. 
,  AND  E.  B.  Brothers.    1982.    Age  and  growth  of  the  fallfish 

Semotilus  corporalis  with  daily  otolith  increments  as  a  method  of 

annulus  verification.  Can.  J.  Zool.  60:2543-2550. 
VijAYARAGHAVAN,  P.    1957.   Studies  On  fish-cggs  and  larvae  of  Madras 

coast.  PhD  thesis  submitted  to  Univ.  Madras,  India. 


Development  of  egg  and  early  larva  of  Cypselurus  spiloplerus  (Cu- 
vier  and  Valenciennes).  Indian  J.  Fish.  2I(1):21 1-219. 

ViLELA,  M.  H.,  AND  J.  J.  ZijLSTRA.  1971.  On  the  condition  of  herring 
larvae  in  the  central  and  southern  North  Sea.  Rapp.  P.  V.  Reun. 
Cons.  Int.  Explor.  Mer  160:137-141. 

ViLLADOLiD,  D.  v.,  AND  P.  R.  Manacop.  1934.  The  PhiUipine  Phal- 
lostethidae,  a  description  of  a  new  species,  and  a  report  on  the 
biology  of  Gullaphallus  mirabilis  Herre.  Phillipp.  J.  Sci.  55:193- 
220. 

ViLLwocK,  W.  1960.  Uber  Bastarde  zwischen  Angehorigen  zweier 
Subfamilien  oviparer  Zahnkarpfen  (Microcyprini,  Cyprinodonti- 
dae).  Zool.  Anz.  165(9-10):349-352. 

Vinagradov,  K.  A.  1950.  Biology  of  the  Pacific  lumpfishes  in  Kam- 
chatka waters.  Priroda  39(3):69-70. 

Vladykov,  V.  D.  1970.  Pearl  tubercles  and  certain  cranial  pecularities 
useful  in  the  taxonomy  of  coregonid  genera.,  p.  167-193.  In:  Bi- 
ology of  coregonid  fishes.  C.  C.  Lindsey  and  C.  S.  Woods  (eds.). 
Univ.  Manitoba  Press,  Winnipeg. 

Vlymen,  W.  J.  1977.  A  mathematical  model  of  the  relationship  be- 
tween larval  anchovy  (Engraulis  mordax)  growth,  prey,  microdis- 
tribution,  and  larval  behavior.  Environ.  Biol.  Fishes  2:21 1-233. 

Vodyanitsky,  V.  a.,  and  I.  I.  Kazanova.  1954.  Key  to  the  pelagic 
fish  eggs  and  larvae  of  the  Black  Sea.  Tr.  Vses.  Nauchno-Issled. 
Inst.  Morsk.  Rybn.  Khoz.  Okeanogr.  28:240-345.  (In  Russian.) 

VON  Ihering,  R.,  J.  DE  Camargo  B.  and  N.  Planet.  1928.  Os  ovulos 
e  a  desova  dos  peixes  d'agua  doce  do  Brasil.  Bol.  Biol,  14:97-109. 

VON  Westernhagen,  H.  1974.  Observations  on  the  natural  spawning 
of  Alectis  indicus  (Ruppell)  and  Caranx  tgnobilis  (Forsk.)  (Caran- 
gidae). J.  Fish  Biol.  6:513-516. 

,  AND .    1975.    Rearing  and  spawning  siganids  (Pisces: 

Teleostei)  in  a  closed  seawater  system.  Helgol.  Wiss.  Meeresuniers. 
27:1-18. 

,  AND  H.  Rosenthal.    1976.  Induced  multiple  spawning  of  reared 

Siganus  oramin  (Schneider)  (=5.  canaliculatus  Park).  Aquaculture 
7:193-196. 

VooREN,  C.  M.  1972.  Postlarvae  and  juveniles  of  the  tarakihi  (Tel- 
eostei: Cheilodactylidae)  in  New  Zealand.  N.Z.  J.  Mar.  and  Fresh- 
water Res.  6:602-618. 

Voss,  N.  A.  1954.  The  postlarval  development  of  the  fishes  of  the 
family  Gempylidae  from  the  Florida  Current.  I.  Nesiarchus  nasutus 
Johnson  and  Gempylus  Cuv.  and  Val.  Bull.  Mar.  Sci.  Gulf  Caribb. 
4:120-159. 

.  1957.  Fishes  of  the  family  Gempylidae  collected  by  the  Ber- 
muda Oceanographic  Expeditions  1929,  1930,  1931  and  1934. 
Copeia  1957:304-305. 

Vrat,  V.  1949.  Reproductive  behavior  and  development  of  eggs  of 
the  three-spined  stickleback  (Gasleroslens  oculeatus)  of  California. 
Copeia  1949:252-260. 

Wade,  C.  B.  1946.  New  fishes  in  the  collections  of  the  Allan  Hancock 
Foundation.  Allan  Hancock  Found.  Pac.  Exped.  9(8):2 15-237. 

Wade,  R.  A.  1962.  The  biology  of  the  tarpon,  Megalops  atlanticus, 
and  the  ox-eye,  Megalops  cyprmoides.  with  emphasis  on  larval 
development.  Bull.  Mar.  Sci.  Gulf  Caribb.  12:545-622. 

Wake,  M.  H.  1979.  Hyman's  comparative  anatomy.  3rd  ed.  Univ. 
Chicago  Press,  Chicago. 

Wakjya,  Y.,  AND  N.  Takahashi.  1913.  Study  on  fishes  of  the  family 
Salangidae.  J.  Coll.  Agric.  Univ.  Tokyo  14:265-295. 

Waldron,  K.  D.  1968.  Early  larvae  of  the  canary  rockfish,  Se/ia.s/o^/t'.y 
pinniger.  J.  Fish.  Res.  Board  Can.  25:801-803. 

,  AND  B.  M.  ViNTER.  1978.  Ichthyoplankton  of  the  eastern  Ber- 
ing Sea.  Unpubl.  Rep.,  Northwest  Alaska  Fish.  Cent.,  Seattle. 

Walker,  B.  W.  1952.  A  guide  to  the  grunion.  CaHf  Fish  Game  38: 
409-420. 


LITERATURE  CITED 


743 


Wallace,  J.,  and  R.  van  der  Elst.  1973.  Identification  of  juvenile 
mullet.  Bull.  S.  Afr.  Assoc.  Mar.  Biol.  Res.  10:26-27. 

Wallus,  R.  1981.  Early  development  of  the  mooneye,  Hiodon  tergisus 
LeSueur,  with  discussion  of  larval  hiodontid  taxonomy.  Rapp.  P.V. 
Reun.  Cons.  Int.  Explor.  Mer  178:583-584. 

,  AND  K.  L.  Grannemann.    1979.   Spawning  behavior  and  early 

development  of  the  banded  sculpin  Cottus  carolinae  (Gill),  1 99- 
235.  In:  Proceedings  of  a  workshop  on  freshwater  larval  fishes.  R. 
Wallus  and  C.  W.  Voigtlander  (eds.).  Tennessee  Valley  Authority, 
Norris,  TN. 

Walters,  V.  1960.  Synopsis  of  the  lampriform  suborder  Veliferoidei. 
Copeia  3:245-247. 

.    1961.   A  contribution  to  the  biology  of  the  Giganturidae  with 

the  description  of  a  new  genus  and  species.  Bull.  Mus.  Comp.  Zool. 
125(10):297-319. 

.    1963.    The  trachipterid  integument  and  an  hypothesis  on  its 

hydrodynamic  function.  Copeia  1963:260-270. 

.     1964.    Order  Giganturoidei,  p.  566-577.  In:  Fishes  of  the 

western  North  Atlantic.  Mem.  Sears  Found.  Mar.  Res.  No.  l(Pt. 
4):566-577. 

,  AND  J.  E.  Fitch.    1960.   The  families  and  genera  of  the  Lam- 

pridiform  (Allotriognath)  suborder  Trachipteroidei.  Calif  Fish  Game 
46:441-451. 

,  AND .    1964.    The  ribbonfishes  (family  Trachipteridae) 

of  the  eastern  Pacific  Ocean  with  a  description  of  a  new  species. 
Calif  Fish  Game  50:228-240. 

Wang,  J.  C.  1981.  Taxonomy  of  the  early  life  stages  of  fishes  — fishes 
of  the  Sacramento-San  Joaquin  Estuary  and  Moss  Landing  Harbor- 
Elkhom  Slough,  California.  Ecological  Analysts,  Inc.,  Concord,  CA. 

Wang,  J.  C.  S.,  AND  R.  J.  Kernehan.  1979.  Fishes  of  the  Delaware 
estuaries,  a  guide  to  the  early  life  histories.  Ecological  Analysts, 
Inc.,  Towson,  MD. 

Ward,  J.  W.  1957.  The  reproduction  and  early  development  of  the 
sea  catfish,  Galeichthysfelis.  in  the  Biloxi  (Mississippi)  Bay.  Copeia 
1957:295-298. 

Washington,  B.  B.  1981.  Identification  and  systematics  of  larvae  of 
Artedius,  Ctinocottus.  and  Oligocoltus  (Scorpaeniformes:  Cottidae). 
Unpubl.  MS  thesis.  Oregon  State  Univ.,  Corvallis. 

,  AND  S.  L.  Richardson.   MS.   Systematic  relationships  among 

cottid  fishes  based  on  developmental  osteology  with  notes  on  their 
allies.  Gulf  Coast  Res.  Lab.,  Ocean  Springs,  MS. 

Wassersuo,  R.  J.,  AND  R.  K.  Johnson.  1976.  A  remarkable  pyloric 
caecum  in  the  evermannellid  genus  Coccorella  with  notes  on  gut 
structure  and  function  in  the  alepisauroid  fishes  (Pisces:  Mycto- 
phiformes).  J.  Zool.  (Lond.)  179:273-289. 

Watanabe,  H.,  M.  Yukinawa,  S.  Nafcazawa  and  S.  Uevanagi.  1 966. 
On  the  larvae  probably  referable  to  slender  tuna.  Allot hunnusfallai 
Servently.  Rep.  Nankai  Reg.  Fish.  Res.  Lab.  23:85-94.  (In  Japanese, 
English  summary.) 

Watanabe,  M.  1960.  Fauna  Japonica  Cottidae.  Biogeographical  So- 
ciety of  Japan,  Tokyo. 

.    1976.   Studies  on  the  sculpin  of  Japan  and  its  adjacent  waters. 

Bull.  Watanabe  Ichthyol.  Inst.  No.  1 . 

Watling,  H.,  AND  C.  J.  D.  Brown.  1955.  The  embryological  devel- 
opment of  the  American  grayling  (7"Avma//i«  signifer  tricolor)  from 
fertilization  to  hatching.  Trans.  Am.  Microsc.  Soc.  74:85-93. 

Watson,  M.  E.,  and  F.  J.  Mather,  III.  MS.  Species  identification  of 
juvenile  tunas  (genus  Thunnus)  from  the  Straits  of  Messina,  north- 
western Atlantic  and  the  Gulf  of  Mexico.  Manuscript  distributed 
for  information  in  connection  with  the  informal  Pacific  Tuna  Bi- 
ology Conference,  Honolulu,  August  1961. 

Watson,  W.  1974.  Diel  changes  in  the  vertical  distributions  of  some 
common  fish  larvae  in  southern  Kaneohe  Bay,  Oahu.  Hawaii.  Un- 
publ. MS  thesis.  Univ.  Hawaii,  Honolulu. 

.    1982.    Development  of  eggs  and  larvae  of  the  white  croaker, 

Genyonemus  linealus  Ayres  (Pisces;  Sciaenidae)  off  the  southern 
California  coast.  Fish.  Bull.  U.S.  80:403-417. 

,  AND  J.  M.  Leis.    1974.   Ichthyoplankton  of  Kaneohe  Bay,  Ha- 


waii: a  one-year  study  of  the  fish  eggs  and  larvae.  Univ.  Hawaii 


Sea  Grant  College  Program,  Honolulu,  Tech.  Rep.  UNIHI-SEA- 
GRANT-TR-75-01. 

Weber,  M.    1913.    Die  Fische.  Siboga  Exped.  57:1-710. 

,  AND  L.  F.  DE  Beaufort.    1922.    The  fishes  of  the  Indo-Aus- 

tralian  archipelago.  Vol.  4.  E.  J.  Brill,  Leiden. 

,  AND .  1936.  The  fishes  of  the  Indo-Australian  Archi- 
pelago. Vol.  7.  E.  J.  Brill,  Leiden. 

Weihs,  D.,  AND  H.  G.  Moser.  1981.  Stalked  eyes  as  an  adaptation 
towards  more  efficient  foraging  in  marine  fish  larvae.  Bull.  Mar. 
Sci.  31:31-36. 

Weiler,  W.  1968.  Fossilium  Catalogus  I:  Animalia,  Otolithi  piscium 
117. 

Weisel,  G.  F.  1967.  Early  ossification  in  the  skeleton  of  the  sucker 
(Catostomus  macrocheilus)  and  the  guppy  (Poecilia  reticulata).  J. 
Morphol.  121:1-18. 

Weiss,  G.  1974.  Hallazgo  y  descripcion  de  larvas  de  las  polaca  Mi- 
cromesistius  australis  en  aguas  del  sector  Patagonico  Argentino 
(Pisces,  Gadidae)  (Finding  and  description  of  larvae  of  Microme- 
sistius  australis  in  Patagonic  waters  of  Argentina  (Pisces,  Gadidae). 
Physis  Secc.  A  Oceanos.  Org.  33(87):537-542. 

.    1975.    Hallazgo,  descripcion  y  distribucion  de  las  postlarvas 

del  bacalao  criollo,  Salilota  australis  y  del  pez  sable  Lepidopus 
caudatusen  aguas  de  la  plataforma  argentina.  Physis  Secc.  A  Ocean- 
os Org.  34(89):3 19-325. 

Weitzman,  S.  H.  1967.  The  origin  of  the  stomiatoid  fishes  with  com- 
ments on  the  classification  of  salmoniform  fishes.  Copeia  1967: 
507-540. 

.    1974.    Osteology  and  evolutionary  relationships  of  the  Ster- 

noptychidae,  with  a  new  classification  of  stomiatoid  families.  Bull. 
Am.  Mus.  Nat.  Hist.  153(3):327-478. 

,  AND  W.  L.  Fink.    1983.    Relationships  of  the  neon  tetras,  a 

group  of  South  American  freshwater  fishes  (Teleostei,  Characidae), 
with  comments  on  the  phylogeny  of  new  world  characiforms.  Bull. 
Mus.  Comp.  Zool.  150:339-395. 

Wellings,  S.  R.,  and  G.  A.  Brown.  1969.  Larval  skin  of  the  flathead 
sole,  Hippoglossoides  elassodon.  Z.  Zellforsch.  100:167-179. 

Welsh,  W.  W.,  AND  C.  M.  Breder.  1921.  A  contribution  to  the  life 
history  of  the  puffer,  Spheroides  maculatus  (Schneider).  Zoologica 
(N.Y.)  2:261-276. 

Wenner,  C.  a.  1978.  Making  a  living  on  the  continental  slope  and 
in  the  deep-sea:  life  history  of  some  dominant  fishes  of  the  Norfolk 
Canyon  area.  Unpubl.  PhD  dissertation.  Coll.  William  and  Mary, 
Wilhamsburg,  VA. 

.    1979.   Notes  on  fishes  of  the  genus  Paraliparis  (Cyclopteridae) 

on  the  middle  Atlantic  continental  slope.  Copeia  1979:145-146. 

Werner,  R.  C,  and  J.  H.  S.  Blaxter.  1 980.  Growth  and  survival  of 
larval  herring  (Clupea  harengus)  in  relation  to  prey  density.  Can. 
J.  Fish.  Aquat.  Sci.  37:1063-1069. 

Westrheim,  S.  J.  1975.  Reproduction,  maturation,  and  identification 
of  larvae  of  some  Sebastes  (Scorpaenidae)  species  in  the  northeast 
Pacific  Ocean.  J.  Fish.  Res.  Board  Can.  32:2399-241 1. 

Wheeler,  A.  1969.  The  fishes  of  the  British  Isles  and  north-west 
Europe.  Macmillan,  London. 

,  AND  J.  Dunne.    1975.    Triplerygion  atlanticus  sp.  nov.  (Tel- 

eostei-Tripterygiidae)  the  first  record  of  a  tripterygiid  fish  in  north- 
western Europe.  J.  Fish  Biol.  7:639-649. 

,  J.  QuERO  AND  G.  K.REFFT.    1973.   Zcomorphi  (Zeiformes),  p. 

349-354.  In:  Check  list  of  the  fishes  of  the  north-eastern  Atlantic 
and  of  the  Mediterranean  (Clofnam).  Vol.  1.  J.  C.  Hureau  and  Th. 
Monod  (ed.).  Paris,  UNESCO. 

White,  D.  S.  1977.  Early  development  and  pattern  of  scale  formation 
in  the  spotted  sucker,  Minytrema  melanops  (Catostomidae).  Copeia 
1977:400-403. 

White,  H.  C.  1939.  The  nesting  and  embryo  of  Zoarcra  a/igw/Z/am. 
J.  Fish.  Res.  Board  Can.  4:337-338. 

White,  M.  W.,  A.  W.  North,  E.  L.  Twelves  and  S.  Jones.  1982. 
Early  development  of  Notothenia  neglecta  from  the  Scotia  Sea, 
Antarctica.  Cybium,  3rd  Ser.,  6(1);43-51. 

Whitehead,  P.  J.  P.   1 964.  A  redescription  of  the  holotype  of  Clupalosa 


744 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


bulan  Bleeker,  and  notes  on  the  genera  Herklotsichthys.  Sardinella 

and  Escuatosa  (Pisces:  Clupeidae).  Ann.  Mag.  Nat.  Hist.,  Ser.  13, 

7:33-47. 
.    1965.    A  new  genus  and  subgenus  of  clupeid  fishes  and  notes 

on  the  genera  Clupea,  Spratrus  and  Clupeonella.  Ann.  Mag.  Nat. 

Hist.  Ser.  13,  7(78):32 1-330. 
.    1972.   A  synopsis  of  the  clupeoid  fishes  of  India.  J.  Mar.  Biol. 

Assoc.  India  14:160-256. 
.    1973.   The  clupeoid  fishes  of  the  Guianas.  Bull.  Br.  Mus.  (Nat. 

Hist.)Zool.  Suppl.  5:1-227. 
.     1981.    Clupeidae.  In:  FAO  species  identification  sheets  for 

fishery  purposes.  Eastern  Central  Atlantic,  Vol.  2;  Fishing  Areas 

34,  47  (in  part).  W.  Fischer,  G.  Bianchi  and  W.  B.  Scott,  (eds.). 

Canada  Funds-in-Trust.  Ottawa,  Department  of  Fisheries  and 

Oceans  Canada,  by  arrangement  with  FAO. 
Whitley,  G.  P.    1 926.  The  biology  of  north-west  Islet,  Capricorn  Group. 

(C.)  Fishes.  Aust.  Zool.  4:227-236. 
.    1941.   Ichthyological  notes  and  illustrations.  Aust.  Zool.  10(1): 

1-50. 
.    1945.    New  sharks  and  fishes  from  Western  Australia.  Part  2. 

Aust.  Zool.  11:43^5. 
.    1948.    Studies  in  ichthyology.  13.  Rec.  Aust.  Mus.  22(1):70- 


94. 
,  andA,  N.CoLEFAX.   1938.  Fishes  from  Nauru,  Gilbert  Islands, 

Oceania.  Proc.  Linn.  Soc.  N.S.W.  43(3-4):282-304. 
WiCKJ-ER,  W.    1957.    Die  Larve  von  Blennius  fluviatilis  (Asso  1784). 

Biol.  Zentralbl.  76:453-466.  (In  German.) 
.    1959.    Weitere  utersuchungen  iiber  Haftfaden  an  Teleosteer- 

eiem,  speziell  an  Cyprinodon  variegalus  Lacepede  1 803.  Zool.  Anz. 

163:90-107. 
Wiley,  E.  O.    1981,    Phylogenetics,  the  theory  and  practice  of  phylo- 

genetic  systematics.  John  Wiley  &  Sons,  New  York. 
Williams,  F.    1959.   The  barracudas  (genus  Sp/nrae^a)  in  British  East 

African  waters.  Ann.  Mag.  Nat.  Hist.,  Ser.  2,  13:92-128, 
Williams,  G.  C,  S.  W.  Richards  AND  E.  G.  Farnsworth.    1964.  Eggs 

of  Ammodytes  hexapterus  from  Long  Island,  New  York.  Copeia 

1964:242-243. 
Williams,  J.  T.    1984.    Synopsis  of  the  pearlfish  subfamily  Pyramo- 

dontinae  (Pisces:  Carapidae).  Bull.  Mar.  Sci.  33:846-854. 
,  and  R.  L.  Shipp.    1982.   A  new  species  of  the  genus  £c/i/o</on 

(Pisces:  Carapidae)  from  the  eastern  Gulf  of  Mexico.  Copeia  1982: 

845-851. 
Wilson,  D.  E.   1973.   Revision  of  the  cottid  genus  Gymnocanthus.  with 

a  description  of  their  osteology.  Unpubl.  MS  thesis.  Univ.  British 

Columbia. 
Wilson,  D.  F,    1972.   Diel  migration  of  sound  scatterers  into,  and  out 

of,  the  Cariaco  Trench  anoxic  water,  J.  Mar.  Res.  30:168-176. 
Wilson,  D.  P.    1956.   Nesting  of  the  black  sea-bream.  Neptune  (April): 

15,  17. 
.    1958,   Notes  from  the  Plymouth  aquarium.  III.  J.  Mar.  Biol. 

Assoc.  U.K.  37:299-307. 
.    1978.    Territorial  behavior  of  male  dragonets  (Catlionymns 

lyra).  J.  Mar.  Biol.  Assoc.  U.K.  58:731-734. 
Wilson,  H.  V.    1891.   The  embryology  of  the  sea  bass  (Serraniua/rar- 

ius).  Bull.  U.S.  Fish.  Comm.  9(1889):209-277. 
Wilson,  K.  H.,  AND  P.  A.  Larkin.    1980.    Daily  growth  increments  in 

the  otoliths  of  juvenile  sockeye  salmon  (Oncorhynchus  nerka).  Can. 

J.  Fish.  Aquat.  Sei.  37:1495-1498. 
Wilson,  M.  V.  H.,  and  P.  Veilleux.    1982.    Comparative  osteology 

and  relationships  of  the  Umbridae  (Pisces:  Salmoniformes).  Zool. 

J.  Linn.  Soc.  76:321-352. 
Winn,  H.  E.,  and  J.  E.  Bardach.    1960.    Some  aspects  of  the  com- 
parative biology  of  parrot  fishes  at  Bermuda.  Zoologica  (N.Y.)  45(1 ): 

29-34. 
Winterbottom,  R.    1974a.   The  familial  phylogeny  of  the  Tetraodon- 

tiformes  (Acanthopterygii:  Pisces)  as  evidenced  by  their  compar- 
ative myology.  Smithson.  Contrib.  Zool.  155. 
.    1974b.   A  descriptive  synonymy  of  the  striated  muscles  of  the 

Teleostei.  Proc.  Acad.  Nat.  Sci.  Phila.  125:255-317. 


,  and  J.  C.  Tyler.    1983.   Phylogenetic  relationships  of  aracanin 

genera  of  boxfishes  (Ostraciidae,  Tetraodontiformes).  Copeia  1983: 

902-917. 
Wirtz,  p.   1978.  The  behavior  of  the  Mediterranean  Tripterygion  species 

(Pisces,  Blennioidei).  Z.  Tierpsychol.  48:142-174. 
WisNER,  R.  L.    1963.   A  new  genus  and  species  of  myctophid  fish  from 

the  south-central  Pacific  Ocean,  with  notes  on  related  genera  and 

the  designation  of  a  new  tribe,  Electronini.  Copeia  1963:24-28. 
.    1976.   The  taxonomy  and  distribution  of  lantemfishes  (family 

Myctophidae)  in  the  eastern  Pacific  Ocean.  Navy  Oceanographic 

Research  Development  Activity,  Washington  D,C, 
Wollam,  M.  B.    1969.    Larval  wahoo,  Acanthocybium  solandeh  (Cu- 

vier),  (Scombridae)  from  the  Straits  of  Yucatan  and  Florida.  Ba. 

Dep.  Nat.  Resour.  Leafl.  4  (Pt.  1,  No.  12):  1-7. 
WoNGRATANA,  T.    1980.    Systematics  of  clupeoid  fishes  of  the  Indo- 

Pacific  region.  Unpubl.  PhD  dissertation.  Univ.  London. 
.     1983.    Diagnoses  of  24  new  species  and  proposal  of  a  new 

name  for  a  species  of  Indo-Pacific  clupeoid  fishes.  Jpn.  J.  Ichthyol. 

29:385-407. 
Woodland,  D.  J.    1983.    Zoogeography  of  the  Siganidae  (Pisces);  an 

interpretation  of  distribution  and  richness  patterns.  Bull.  Mar.  Sci. 

33:713-717. 
Woods,  L.,  and  P.  Sonoda.     1973.    Order  Berycomorphi  (Beryci- 

formes),  p.  263-396.  In:  Fishes  of  the  western  North  Atlantic.  D. 

M.  Cohen  (ed.).  Mem.  Sears  Found.  Mar.  Res.  No.  1  (Pt.  6). 
WouRMs,  J.  P.     1972a.    Developmental  biology  of  annual  fishes.  I. 

Stages  in  the  normal  development  of  Austrofundulus  myersi  Dahl. 

J.  Exp.  Zool.  182:143-168. 
.    1972b.   Developmental  biology  of  annual  fishes.  II.  Naturally 

occurring  dispersion  and  reaggregation  of  blastomeres  during  the 

development  of  annual  fish  eggs.  J.  Exp.  Zool.  182:169-200. 
.    1972c.    Developmental  biology  of  annual  fishes.  III.  Pre-em- 

bryonic  and  embryonic  diapause  of  variable  duration  in  the  eggs 

of  annual  fishes.  J.'  Exp.  Zool.  182:389-414. 
.    1981.  Viviparity:  the  maternal-fetal  relationship  in  fishes.  Am. 

Zool.  21:473-515. 
,  andO.  Bayne.    1973.  Development  of  the  viviparous  brotulid 

fish  Dinematichthys  ilucoeteoides.  Copeia  1973:32-40. 
,  and  D.  M.  Cohen.    1975.    Trophotaeniae,  embryonic  adap- 
tations, in  the  viviparous  ophidioid  fish,  Oligopus  longhursli:  a 

study  of  museum  specimens.  J.  Morphol.  147:385-402. 
,  and  D.Evans.    1974.  The  embryonic  development  of  the  black 

prickleback,  Xiphister  alropurpureus.  a  Pacific  coast  blennioid  fish. 

Can.  J.  Zool.  52:879-887. 
,  AND  G.  S.  Whitt.    1981.    Future  directions  of  research  on  the 


developmental  biology  of  fishes.  Am.  Zool.  21:597-604. 
Wu,  H.-W.    1932.    Contribution  a  Tetude  morphologique,  biologique 

et  systematique  des  poissons  heterosomes  (Pisces  Heterosomata) 

de  la  Chine.  Thesis,  Univ.  Paris. 
WvANSKj,  D.  S.,  AND  T.  E.  Targett.    1981.    Feeding  biology  of  fishes 

of  the  endemic  Antarctic  family  Harpagiferidae.  Copeia  1981 :686- 

693. 
Wyman,J.    1859.  On  some  unusual  modes  of  gestation.  Am.  J.  Science 

and  Arts  27:5-13. 
Yabe,  H.,  S.  Ueyanagi,  S.  KiicAWA  AND  H.  Watanabe.    1959.    Study 

of  the  life-history  of  the  sword-fish  Xiphias  gladins  Linnaeus.  Rep. 

Nankai  Reg.  Fish.  Res.  Lab.  10:107-150.  (In  Japanese,  English 

summary.) 
Yabe,  M.    1981.   Osteological  review  of  the  family  Icelidae  Berg  1 940, 

(Pisces;  Scorpaeniformes),  with  comment  on  the  validity  of  this 

family.  Bull.  Fac.  Fish.  Hokkaido  Univ.  32(4):293-315. 
,  D.  M.  Cohen,  K.  Wakabayashi  AND T.  IwAMOTO.   1981.  Fish- 
es new  to  the  eastern  Bering  Sea.  Fish.  Bull.  U.S.  79:353-356. 
-,  K.  Tsumura  AND  M.  Katayama.    1980.   Descnption  ofa  new 


cottid  fish,  Icelinus  japomcus  from  Japanese  waters.  Jpn.  J.  Ich- 
thyology 27:106-110. 
Yamamoto,  K.,  K.  Yamauchi  and  S.  Kasuga.    1975a.    On  the  de- 
velopment of  the  Japanese  eel,  AnguiUa  japonica.  Bull.  Jpn.  Soc. 
Sci.  Fish.  41:21-28. 


LITERATURE  CITED 


745 


, AND  T.  MoRiOKA.    1975b.    Pre-leptocephalic  larvae  of 

the  Japanese  eel.  Bull.  Jap.  Soc.  Sci.  Fish.  41:29-34. 

Yamamoto.  M.,  and  K.  Ueda.  1978.  Comparative  morphology  of 
fisholfaclory  epithelium  — 2.  Clupeiformes.  Bull.  Jpn.  Soc.  Sci.  Fish. 
44:855-859. 

Yanagawa,  H.  1978.  Embryonic  development  of  the  kyuriuo,  Os- 
merus  eperlanus  mordax  (Mitchill).  Bull.  Fac.  Fish.  Hokkaido  Univ. 
29(3):195-198. 

Yang,  W.  T.,  and  U.  B.  Kjm.  1962.  A  preliminary  report  on  the 
artificial  culture  of  grey  mullet  in  Korea.  Proc.  Indo-Pac.  Fish  Counc. 
9:62-70. 

Yanulov,  K.  p.  1962.  On  the  reproduction  of  the  roughhead  grena- 
dier. Zool.  Zh.  41:1259-1262. 

Yapchionges,  J.  V.  1949.  Hypomesus  pretiosus  ix^  <ie\e\opmtnX  and 
early  life  history.  Nat.  Appl.  Sci.  Bull.  9. 

Yarberry,  E.  L.  1965.  Osteology  of  the  zoarcid  fish  Melanostigma 
pammelas.  Copeia  1965:442-462. 

Yashouv,  A.,  AND  E.  Berner-Samsonov.  1970.  Contnbution  to  the 
knowledge  of  eggs  and  early  larval  stages  of  mullets  (Mugilidae) 
along  the  Israeli  coast.  Bamidgeh  22:72-89. 

Yasutak£,  H.,  G.  NisHi  and  K.  Mori.  1973.  Artificial  fertilization 
and  rearing  of  bigeye  tuna  ( Thunnus  ohesus)  on  board,  with  mor- 
phological observations  on  embryonic  through  to  early  post-larval 
stage.  Bull.  Far  Seas  Fish.  Res.  Lab.  (Shimizu)  (8):71-78. 

Yefremenko,  V.  N.  1976.  Morphological  characteristics  of  Proto- 
myclophum  anderssoni  larvae  (Myctophidae).  J.  Ichthyol.  (Engl. 
Transl.  Vopr.  Ikhtiol.)  16:958-963. 

.    1977.   Morphological  features  of  the  larvae  of  Gvm«ojcopf/wi 

opisthopterm  (Mynophidae,  Pisces).  J.  Ichthyol.  (Engl.  Transl.  Vopr. 
Ikhtiol.)  17:797-800. 

.    1979a.    A  descnption  of  the  larvae  of  Balhylagus  anlarcticus 

(Bathylagidae).  J.  Ichthyol.  (Engl.  Transl.  Vopr.  Ikhtiol.)  1 9(  1 ):  1 56- 
160. 

.    1979b.    Description  of  the  larvae  of  six  species  of  the  family 

Chaenichthyidae  from  the  Scotia  Sea.  J.  Ichthyol.  (Engl.  Transl. 
Vopr.  Ikhtiol.)  19(3):65-75. 

.    1979c.   The  larvae  of  six  species  of  the  family  Nototheniidae 

from  the  Scotia  Sea.  J.  Ichthyol.  (Engl.  Transl.  Vopr.  Ikhtiol.)  19(6): 
95-104. 

.     1982.    The  finding  of  fla/Zu'/agwi  a«/arrt;cMi  (Bathylagidae) 

eggs,  their  characteristics  and  distribution  in  the  Scotia  Sea.  J.  Ich- 
thyol. (Engl.  Transl.  Vopr.  Ikhtiol.)  22(2):  135-1 37. 

.    1983.   Atlasof  fish  larvae  of  the  Southern  Ocean.  Cybium  7(2): 

1-74. 

Yevseyenko,  S.  a.  1974.  New  data  concerning  eel  eggs  on  Georges 
Bank.  J.  Ichthyol.  (Engl.  Transl.  Vopr.  Ikhtiol.)  14:294-298. 

,  AND  V.  P.  Serebryakov.    1974.    Larvae  of  Diplospinus  nnil- 

listnatus  Maul  (Pisces,  Gempylidae)  from  the  norihwestem  Atlan- 
tic. J.  Ichthyol.  (Engl.  Transl.  Vopr.  Ikhtiol.)  14:92-98. 

YuSA.  T.  1954.  On  the  normal  development  of  the  fish,  Theragra 
chalcogramma  (Pallas),  Alaska  pollock.  Bull.  Hokkaido  Reg.  Fish. 
Res.  Lab.  10:1-15.  (In  Japanese,  English  summary.) 

.     1957.    Eggs  and  larvae  of  flatfishes  in  the  coastal  waters  of 

Hokkaido.  I.  Embryonic  development  of  the  starry  flounder  Pla- 
tichthys  stellalus  (VaWas).  Bull.  Hokkaido  Reg.  Fish.  Res.  Lab.  15: 
1-14. 

.     1958.    Eggs  and  larvae  of  flatfishes  in  the  coastal  waters  of 


Hokkaido.  III.  Post-embryonic  development  and  larvae  of  the  flat- 
fish Lepidopsella  mochigarei  Snyder.  Bull.  Hokkaido  Reg.  Fish. 
Res.  Ub.  18:1-10. 


.    1 960a.    Diflferences  of  structures  of  eggs  and  larvae  between 

Limanda  yokohamae Gunther  and  Limanda  schrenki Schmidt.  Bull. 
Mar.  Biol.  Stn.  Asamushi  10(2):127-131. 

.    1 960b.    Eggs  and  larvae  of  flatfishes  in  the  coastal  waters  of 

Hokkaido.  IV.  Embryonic  development  of  mud  dab  Limanda  yo- 
kohamae  GunlhtT .  Bull.  Tohoku  Reg.  Fish.  Res.  Lab.  17:15-30. 

.    1960c.    Embryonic  development  and  larvae  of  the  greenling 

fish,  He.xagrammus  olakii  Jordan  and  Starks.  Groundfish  Rep.  To- 
hoku Reg.  Fish.  Res.  Lab.  26:76-80.  (In  Japanese.) 

.    1961.    Eggs  and  larvae  of  flatfishes  in  the  coastal  waters  of 

Hokkaido.  V.  Embryonic  development  of  the  flatfish.  Xyslrais 
ghgorjewi  (Herzenstein).  Bull.  Tohoku  Reg.  Fish.  Res.  Lab.,  19: 
109-118. 

.    1967.    Embryonic  development  of  the  Atka  mackerel,  F/fu- 

rogrammus  aronuj  Jordan  et  Metz.  Sci.  Rep.  Tohoku  Univ.  Fourih 
Ser.  (Biol..)  33:301-318. 

,  C.  R.  Forrester  and  C.  Ioka.     1971.    Eggs  and  larvae  of 

Limanda  yokohamae  {GuMher).  Fish.  Res.  Board  Can.  Tech.  Rep. 
No.  236. 

Yuschak,  P.  1982.  Description  ofthe  eggs  and  larvae  and  osteological 
development  of  the  northern  searobin,  Pnonotus  carolinus  (Pisces: 
Triglidae).  Unpubl.  MS  thesis.  Univ.  Connecticut,  Storrs. 

Zahradica,  K.,  and  S.  Frank.  1976.  Observations  sur  la  reproduction 
de  Nothobranchius  korthausae.  Aquarama  10(36):16-24. 

Zahuranec,  B.  J.  1980.  Zoogeography  and  systematics  ofthe  lan- 
temfishes  ofthe  genus  Nannobrachnun  (Lampanyctini:  Myctoph- 
idae). Unpubl.  PhD  dissertation.  George  Washington  Univ..  Wash- 
ington, D.C. 

Zaitsev,  Y.  p.  1970.  Marine  neustonology.  Israel  Program  for  Sci- 
entific Translations,  Jerusalem. 

Zama,  A.,  M,  AsAi  and  F.  Yasuda.  1977.  Changes  with  growih  in 
bony  cranial  projections  and  color  patterns  in  the  Japanese  boarfish, 
Pentaceros  japonicus.  Jpn.  J.  Ichthyol.  24:26-34. 

Zehren,  S.  J.  1979.  The  comparative  osteology  and  phylogeny  of  the 
Beryciformes  (Pisces:  Teleostei).  Evol.  Monogr.  (Univ.  Chicago)  1. 

Zheng,  W.  1981.  Comparative  studies  on  the  scales  of  Chinese  Ca- 
rangidae.  Trans.  Chin.  Ichthyol.  Soc.  1:33-45.  (In  Chinese,  English 
summary.) 

Zhitenev,  A.  N.  1970.  Ecological  and  morphological  features  of  the 
reproduction  ofthe  lumpsucker  (Cycloplerus  lumpus  (L.).  J.  Ich- 
thyol. (Engl.  Transl.  Vopr.  Ikhtiol.)  10:77-83. 

ZiSMAN,  L.  1982.  Means  of  identification  of  grey  mullet  fry  for  culture, 
p.  17-63.  In:  Aquaculture  of  grey  mullets.  Intl.  Biol.  Programme 
26.  O.  H.  Oren  (ed.).  Cambridge  Univ.  Press,  Cambridge. 

ZuRLiNi,  G.,  I.  FERR.ARI  AND  A.  Nassogne.  1978.  Reproduction  and 
growth  of  Euterpina  aculifrons  (Copepoda:  Harpacticoida)  under 
experimental  conditions.  Mar.  Biol.  (Berl.)  46:59-64. 

Zviagina,  O.  A.  1961.  Data  on  the  propagation  and  development  of 
fish  of  Laptev  Sea.  1 .  East-Sibenan  cod  (Arcotogadus  bonsovi,  Gad- 
idae).  Tr.  Inst.  Okeanol.  Akad.  Nauk  SSSR  43:320-327.  (In  Russian 
Transl.  by  Fish.  Res.  Board  Can.,  1969,  Transl.  Ser.  1278.) 

Zvjagina,  O.  a.  1965a.  Malerialy  orazvitii  jashcherogolobykh  ryb 
(Pisces,  Synodontidae).  Tr.  Inst.  Okeanol.  Akad.  Nauk  SSSR  80: 
147-161. 

.    1965b.   On  the  development  of  Therapon  theraps  Cuv.  et  Val. 

(Pisces,  Theraponidae).  Tr.  Inst.  Okeanol.  Akad.  Nauk.  SSSR  80: 
162-166.  (In  Russian,  English  summary.) 

Zvyagina,  O.  a.,  and  T.  S.  Rass.  1977.  Materials  on  development 
of  tropical  carangid  fishes  (Carangidae,  Pisces).  Proc.  P.  P.  Shirshov 
Inst.  Oceanol.  109:211-227. 


INDEX 

Italicized  page  numbers  indicate  illustrations.  Index  includes  all  genera  and  only  higher  taxa  that  are  illustrated. 


Abbottina  3 1 

Ablabys  440 

Ablennes  336,  345,  347,  350.  352 

Abramis  131 

Absalom  513 

Abudefduf5'\^ 

Abyssocottus  409-10,  443 

Acanthistius  500-1 

Acanthocepola  474,  490 

Acanthochromis  542,  544,  547 

Acanthoclinus  467 

Acanthocybium  592.  600-1.  606,  609,  612,  617-8 

Acanthogobius  582 

Acantholiparis  429-30 

Acanthopagrus  471 

Acanthophthalmus  129 

Acanthopsetta  643,  654,  660 

Acanthosphex  440 

Acanthostracion  452,  453.  454,  459 

Acanthurus  548,  549 

Acentrophryne  333 

Achiropichthys  103 

Achiropsetta  674 

Achirus  23,  25,  644,  649,  657,  664,  668 

Acromycter  74,  98,  101 

Acropoma  464,  466 

Acrytops  629-30,  633 

Adelosebastes  440 

Adinia  363-6 

Adioryx  18.  392 

Adrianichthys  345,  352 

Advent  or  440 

Aeoliscus  403,  404 

Aesopia  644,  649,  657,  672 

Aetapcus  440 

Aethotaxis  563 

Agonomalus  408,  ¥25,  428,  442,  447 

Agonopsis  408,  442 

Agonostomus  54 1 

Agonus  408,  442 

Agrostichthys  368-9,  371-2,  379 

/iWw  71,  103,  105 

Ahliesaums  207-8,  21 1,  272 

^;<2A:a5  578 

Akarotaxis  563 

.4/i«/a  52,  60,  67.  62,  98-9,  126,  138 

Alcichthys  442 

Aldrovandia  94-5 

^/ecto511,  575,  518-9,  524 

.4/epe5  511,  518,  524 

Alepisaurus  200,  207-8,  212,  216,  245,  259 

Alertichthys  440 

Alestes  133-4 

Allanetta  360 

^//i/75  103 

Allolepis  580 

Allothunnus  592,  600-1,  670,  612,  617 


^/oifl  112,  118,  124-5,  129-130,  133 

Alphestes  500,  507-8 

/4/u;era  452,  455-6,  -^57 

Amanses  452.  455,  -^57 

Amarsipus  621.  622-3,  625-8 

Ambloplites  473 

Amblycirrhitus  47 1 

Amblygaster  109,  124 

Ammodytes  574,  575 

Ammotretis  643-4,  660,  673 

Amphiprion  544 

Anableps  363 

Anacanthus  452,  -#54,  455 

Anaora  637 

Anarchias  65,  71 

Anarhichas  568,  577 

Anarrhichthys  568 

/l«r/!oa25,  29.  115,  117,  775,  125 

Anchoviella  116-7,  125 

Ancistrus  131,  7ii 

Ancylopsetta  640.  651,  682 

Andriashevia  557-8 

Anguilla  56,  65,  69,  72,  94,  126 

Anguillidae  87 

Anisotremus  49,  489 

Anodontostoma  111,  7  7(5,  124 

Anomalops  392 

Anoplagonus  442 

Anoplarchus  57 1 

Anoplogaster  7,  387,  i90.  392,  490 

Anoplopoma  408,  '/7'#,  421-2,  442,  444-5 

Anotopterus  207-8,  272,  216,  245 

Antennahus  15,  44,  320,  322.  i2i 

/lrtr;!/a5  500,  503,  505,  509 

Antigonia  393,  i96,  405,  460,  489-490 

Antimora  266,  280 

Antipodocottus  442 

Anyperodon  500 

/4pe/to  i99,  404 

Aphanius  362,  365 

Aphanopus  599-601,  604 

Aphredoderus  483 

Apistius  440-1 

Aplatophis  71,  103,  106 

Apletodon  629-30,  6ii 

Aploactis  440 

Aploactisoma  440 

Aplocheilus  364,  366 

Aplochiton  150-1,  752.  153,  202-4 

Apogonidae  -#70 

Apogonops  464 

Apolecius  530 

Aporops  501,  508 

Aprognathodon  103 

Aplehchtus  71,  103 

Aptocyclus  429-3 1 ,  437 


746 


INDEX 


747 


Araiophos  185-6,  188,  190,  197 

Archaulus  442 

Archistes  442 

Arctogadus  266,  268,  280,  284,  286-7,  290,  294-5,  297 

Argentina  14.  156-7.  158-9.  161,  163^,  168-9 

Argyripnus  185-6,  188,  190,  797 

Argyrocetlus  442 

Argyropelecus  185-8,  190-1,  197 

Anoinma  622-3,  624.  625-8 

Ariosoma  67,  69-70,  74 

Aristostomias  171-2,  175-6,  7 79,  181,  183 

Arnoglossus  19.  642-3,  648,  652,  656.  657 

Arrhamphus  342-3,  352-3 

Artedidraco  562.  563-4 

Ariediellichthys  442 

Artediellina  442 

Artedielliscus  442 

Artediellus  442 

Artedius  408,  ^/9,  421,  422.  427,  442,  446-7 

Ascelichthys  419.  442 

Asemichthys  442 

Aseraggodes  664.  668,  672 

Aspasma  629-30,  633,  6i5-<5 

Aspasmichthys  629-30,  633,  6i5 

Aspidophoroides  426.  428,  442 

Aspredo  130 

/ls5«r,ger  599-601,  604 

Asterorhombus  642,  652,  686 

Asterropteryx  582 

/l5/rafcf  5«4,  587 

Astrocottus  442 

Astronesthes  171-2,  174-5,  183 

ASTRONESTHIDAE   772 

Astroscopus  560 

Ateleobrachium  269 

Atheresthes  643,  654,  659,  660.  684 

Atherina  355-6.  360.  533 

Atherinason  360 

Athennomorus  355.  357,  i59,  360 

Atherinops  44.  48,  355-6.  360.  i67.  i67 

Athehnopsis  44.  48,  355-6,  i57,  360 

Atherinosoma  360 

Atherion  355-6,  360 

Atractoscton  25 

Alrophacanthus  452-4 

.4/rapu5  511,  518,  524 

rl/«/e511,  5/5,  517-8,  520.  524 

Atypichthys  469 

Auchenoceros  266,  268 

Aulacocephalus  500 

/lM/op«s  207-8,  2//.  258 

Aulorhynchus  399.  400-4 

Aulostomus  401 

Austroglossus  644.  657,  665 

Austrolychus  580 

Austrolycichthys  578 

Austromenidia  355-6 

/lM.v/5  592.  600-1.  606.  670,  612,  617,  619 

^zy,?*?^^^  643.  673,  684 


Bagahus  129-30, 
Bairdiella  24 


34.  137 


Batistes  452.  457 

Balistapus  452 

Barbatula  129 

Barbourisia  382 

Bar^Mi  729 

BascanichthyslX,  98.  707.  103.  107 

Bathophilus  171.  174-6,  779,  181,  183 

Bathyagonus  442 

Bathyaploactes  440 

Bathycallionymus  6311 

Bathyclupea  474,  479 

Bathydraco  563 

Bathvgadus  274-5 

Bathylagus  11,  156-7,  755,  161,  762,  163-4,  765,  167.  169 

Bathyleptus  199.  201 

Bathylychnops  156-7.  161,  163-4,  767 

Bathymicrops  207-9,  211,  272,  256,  258 

Bathymyrus  70,  73 

Bathypterois  207-9,  211,  272,  256,  258 

Bathysauropsis  207-8,  258 

Bathysaums  207-8,  272 

Bathysolea  657 

Bathvstethus  469 

Bathytyphlops  207-9,  272,  256,  258 

Batrachocottus  410,  422,  443 

Bec?c-r;a  355-7,  i59,  360 

7Sf'//<2/or419 

Be/o«f  336,  342,  i4i,  347,  350,  352 

Belonion  342,  i45,  347,  350,  352 

Belonoperca  500 

Bembradium  44 1 

Bembradon  44 1 

Bembras  44 1 

Bemhabella  245,  2-^6,  247-50,  256 

Benthenchelys  103,  106 

Benthocometes  3 1 2 

Benthodesmns  599-601,  602,  604,  606-7 

Benthosema  218-9,  221-2,  226,  227,  229,  241-3 

Bero  442 

Bm'x  392 

Bilabria  578 

Bleekeria  574 

Blepsias  408,  ^^25,  428,  442,  447 

Bodianus  544 

Bolinichthys  218,  220-2,  226,  2i5,  236,  240-3 

Bonapartia  182,  185-6,  188-9.  190-1.  79i,  195.  198 

Boreogadus  266,  268,  280.  284.  286-7.  290.  294-5,  297 

Borichthys  563-4 

Borophryne  327,  i2<S,  329,  333-4 

Borostomias  171,  174-5,  183 

Bostockia  469-70 

BOTHIDAE  641 

Bothragonus  408,  -^26,  428,  442 
Bothrocara  578.  580,  557,  582 
Bothrocarina  578 
5or;;M,s  642.  652.  655,  657,  672 
Bo/;a  131,  137 
Brachaluteres  452,  455 
Brachydanio  131 
Brachygalaxias  150,  153 
Brachymystax  143-4,  7-^6,  148-9 
Brachyopsis  442 


748 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Brachypleura  640,  645.  648,  651,  674,  678,  681-4 

Brachysomophis  103,  106 

Brama  474 

Branchiostegus  495 

Bregmaceros  40,  42,  260-3,  266,  268,  280-2,  300.  301-8,  319 

Brevoortia  40,  113,  123,  125 

Brinkmanella  469,  479 

Brosme  260-1,  266,  268,  277,  278.  280,  282,  284,  285,  286-9, 

294-5,  297-9 
Brosmiculus  266,  280 
Brosmophycis  309,  3 1 1 
Brotula  282,  308,  311,  319 
Brotulotaenia  308,  3 1 1 
Brycinus  131 

Buglossidium  644.  657,  660 
Bythitidae  309 


Caecula  103 

Caesio  468 

Caesioperca  500,  503 

Caesioscorpis  465,  476 

Cairnsichthys  360 

Callanthias  464,  -/dZ,  499 

Callechelys  71,  100.  103,  107 

Callionyinus  44,  48,  637,  6i9 

Callturwhthys  637-9 

Calotomus  544,  546 

Campogramma  511,  518,  523-4,  527 

Cantherines  452,  455 

Canthidermis  452,  ^55 

Canthigaster  448,  -^49 

Capoeta  129 

Caprodon  500 

Capromimus  405 

Caproi  393,  i95,  405,  460-1 

Caralophia  103,  107 

Cara«^o/fi^«  511,  513,  518,  524 

Cara^A:  510,  513,  522,  524,  52S-9 

Carapus  19,  309,  312-4,  i;6,  317-9 

Careproctus  429-32,  -/i-^.  437 

Carinotetraodon  448 

Caristius  474 

Carolophia  1 1 

Caspialosa  114,  125 

Catesbya  100 

Catostomus  53,  129 

Caulolatilus  468,  495 

Caulolepis  1 

Caulophryne  321.  316-1,  328.  334 

Centriscidae  407 

Centroberyx  392 

Cenlrobranchus  218-9,  221-2,  229,  234,  240-3 

Centrolabrus  544 

Centrolophus  622-3,  626,  628 

Centrophryne  329 

Centropogon  440 

Centropomus  473 

Centropristis  499-50 1 ,  50i,  509 

Centropyge  468 

Cephalopholis  500,  507-8 

Cephalopsetia  641,  651,  674,  682 


Ceratias  327,  i29 

Cera/05cope/M<r  218,  220-2,  226,  2i5.  236,  240-3 

Ceratostethus  361 

Cerdale  587 

Celengraulis  117,  /2/,  125 

Cetonurus  261 ,  274-5 

Chaenichthys  563 

Chaenocephalus  563 

Chaenodraco  563 

Chaenogobius  582 

Chaenophryne  326,  iii,  333 

Chaetodiptems  465,  465,  487,  489-90 

Chaetodontidae  474 

Chalinura  274-5 

Champsocephalus  563 

Champsodon  559,  560 

Channomuraena  72 

C/;aA!o<r24,  126,  129,  130,  133,  7i5.  139,  206 

Chascanopsetta  642,  652,  656 

Chaultodus  29,  30.  170,  171-2,  174,  181,  198 

Chaunax  321 ,  322 

Cheilinus  544 

Cheilopogon  337-9,  341,  i42,  343,  345,  347,  i49,  352,  354 

Chelidonichthys  407,  419 

Chelidoperca  500 

Chelmon  474 

Chelonodon  448 

Cheroscorpaena  440-1 

Cherublemma  313,  318 

Chilara  3\2-4,  319 

Chilatherina  360 

Chilomycterus  448,  450,  463 

Chilonetus  408 

Chionobathyscus  563 

Chionodraco  563 

Chirocentrodon  1 1 4 

Chirocentrus  108-9,  //5.  720,  124 

Chirostoma  355-6 

Chirostomias  171-2,  183 

Chitonotus  419.  442 

Chlorophthalmus  207-9,  2//,  256,  258 

Chloroscombrus  511,  574,  518,  524 

Choerodon  544 

Chonerhinos  448 

Choridactylus  440 

Chnodorus  343,  352-3 

Chromis  544 

Cichlasoma  542,  54i 

0//a/a  266,  268,  279,  284-90,  294-5,  297-9 

Cirrhilabrus  544 

Cirrhimuraena  103 

Cirricaecula  103 

Citharichthvs  641,  643,  650.  651-4,  670,  675,  657,  682,  685 

Citharoides  641,  672-3,  676,  67<S,  681-2 

C/anoi  130-1,  7ii,  134 

Claringer  586 

Cleidopus  392 

Cleisthenes  642-3,  654,  660 

Clidoderma  643,  660 

Clinocottus  409,  421,  422,  427,  442,  447 

Clupanodon  111,  122 

C/Mpra  75.  24-6,  1 10,  118,  122-4 


INDEX 


749 


Clupeichthys  1 1 2 
Clupeoides  1 1 3 
Clupeonella  110,  114 
Cobitis  129,  131,  137 
Coccorella  250,  257-254 
Coccotropsis  440 
Cocotropus  439-41 
Coelorhynchus  44,  269,  274-5,  279 
Co/Via  114,  118.  121.  125 
Colistium  643-4,  660,  672 
Coloconger  70 

COLOCONGRIDAE  5i 

Cololabts  14.  29,  iO,  336,  338,  ii9,  i^2,  347,  349.  350,  352, 

354 
Colpichthys  360 
Comephorus  408,  ^^76,  442 
Conger  65 

Congiopodus  407,  -^74.  419,  440 
Congothrissa  1 1 1 
Congresox  94 

CONGRIDAE  .57,  (SJ 

Congrogadus  467 

Conidens  629.  630-1,  6ii 

Conodon  473.  489-90 

Coraanus  465 

Coradion  474 

Coregonus  142-4,  /^^i,  148 

Coreoperca  469,  47i 

Corica  1 1 3 

Cora  544 

Coryphaena  474.  489-90,  -^96-7,  498 

Coryphaenoides  44,  48,  267,  269,  272,  274-5 

Coltapisti4s  440 

Co?//«W/a410,  443 

Cottiusculus  442 

Cottocomephorus  4\0,  416.  443 

Cottoperca  563 

Cottunculus  444 

Co»M5  422,  427,  442,  446-7 

Craterocephalus  360 

CraUnus  500-1 

Crenichthys  362-4 

Cromeria  138 

Cromileptes  500 

Crossias  442 

Crossorhombus  642,  652,  655.  657 

Crossostomus  578 

Cry  draco  563 

Cryothenia  563 

Cryptopsaras  'ill,  329.  333 

CryslalUas  429 

Crystallichthys  429 

Clenochaetus  548 

Ctenochirichthys  327 

Ctenolabrus  544 

Ctenolucius  127.  130 

Ctenopharyngodon  129,  131 

Ctenotrypauchen  587 

Cubaitichthvs  363,  365 

Cubiccps  622-3,  62-/,  625-8 

Cybiosarda  592,  600-1,  606,  612,  617 

Cyclopsetta  641,  650.  651-3,  657,  670,  673-5,  682-3,  685-6 


Cyclopsis  429-30 

Cyclopteropsis  429-30,  437 

Cydopterus  429-30.  4i2.  4i6.  437 

Cyclothone  182,  185-6.  188-91.  193,  795.  198 

O'f'wa  69,  72,  97 

Cyematidae  9i 

Cygnodraco  563 

Cynoglossus  640,  6-^7.  643-4,  664,  667,  669,  670,  672,  675 

Cynolebias  363-4 

Cynomacrurus  ll'^-S 

Cynoponticus  94 

Cynothrissa  1 1 3 

Cyprinodon  359,  362-3,  i65 

Cyprinus  131,  133-4 

Cypselurus  338,  341,  345,  347-8,  352,  354 

Cw/wi  379,  394 

Dactylanthias  500 

Dactyloptena  408,  441 

Dactylopterus  408.  '^27,  441 

Dactylopus  637 

Dadyanos  578.  580 

Daicocus  408 

Z)a//;a  140,  142,  202 

Dalophis  103,  107 

Dampierosa  440 

Danaphos  185-6,  188,  190,  797 

Z)a«;o  131 

Daruma  442 

i3a.rKO""-?410,  42<  428,  444 

Datnioides  465,  478 

Davidjordama  578 

Z)aye//a  113 

Decapterus  510,  512.  57-^,  518-20,  522,  524 

Dendrochirus  407.  416 

Dentatherina  357 

Denticeps  108-9 

Derepodichthys  578,  580 

Derichthyidae  57 

Derichthys  96 

Dermatolepis  500,  507 

Dermogenys  336-8,  347,  352-3.  i6/ 

Desmodema  368-9.  371-2,  377-9 

Dexistes  643 

Diaphus  207-8.  218-9.  221-2,  229,  236,  241-3 

Dicentrarclms  26,  469.  509 

Diceratias  330 

Dicologoglossa  644,  657 

Dicotylichthys  450 

Dinematichthys  3 1 1 

Dinoperca  465,  478 

£)/oc?o«  448,  '/49 

Diogemchlhys  218-9,  221-2,  226,  227,  229,  241-3 

Diplecogaster  629,  6ii 

Diplectrum  500-1,  50i,  509 

Diplocrepis  629-30,  6ii 

Diplodus  14 

Diplogrammus  637 

Diplomvstus  1 22 

Diplophus  182,  185-6,  188-9.  190-1,  792,  193.  198 

Diploprion  500,  510 

Diplospinus  593,  599.  600-1.  603-4,  615-7 


750 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Dipterygonotus  49 1 

Diretmoides  379 

Diretmus  379,  390.  392,  490 

Dissostichus  563 

Doderleinia  464,  466 

Dolichopteryx  156-7,  161,  163-4,  767 

Dolichosudis  207-8 

Dolliodraco  563 

Dolopichthys  111,  331 

Doratonotus  546 

Dormitator  585.  586-7 

Dorosoma  111,  118,  124 

Dorsopsetta  657 

Dracodraco  563 

Drepane  465 

Dwto  500-1 

Dussumieria  110,  ;/«,  720,  123-4 

Dysomma  101 

Dysommatidae  75,  96 


Ebinania  4 1 4 

£c/!e/«5  102-3,  106 

Echeneis  474.  497 

Echinomacrurus  274-5 

Echiodon  40.  42,  309.  312,  314,  i/5.  317 

Echiophisll,  103,  106 

Echiostoma  171-2,  174-5,  181,  183^ 

Ectreposebastes  407.  413.  417 

frfeZ/fl  469-70 

Edriolychnus  333 

Ehirava  1 1 3 

Eigenmannia  127,  131,  ;ii.  134,  137 

Elagatis  5\2,  517-20,  523-4.  526-7 

Elapsopis  103 

Elassichthys  335,  342,  350,  352 

Elassodiscus  429 

Elassoma  465-6,  469 

Electrona  218-9,  221-2,  225.  226,  229.  240-243 

Eleginops  563 

Eleginus  266,  268,  280,  284,  286-7,  290,  294-5,  297 

£/wrm  582,  584,  586-7 

Eleutherochir  637 

Eleutheronemus  540 

EUerkeldia  500 

Elops  60,  61.  62,  98-99,  126 

Elytrophora  6 1 7 

Embassichthys  642-3,  654,  659,  660,  66'^ 

Embolichthys  575 

Emmelichthyops  49 1 

Encheliophis  312,  314,  317-9 

Enchelyopus  266,  268,  279-80.  284-9,  290.  294-5,  297.  319 

Enchelyurus  570 

Encrasichelina  116.  123 

EngraulislA,  28.  108,  114,  //S,  ;2;,  122,  125 

Engyophrys  642,  652,  65i,  657,  672,  674,  676,  682-3,  685-6 

Engyprosopon  642,  652,  655.  657 

Enneapterygius  570 

Enophrys  409,  -^2/,  427,  442 

Eocallionymus  637 

Eophycis  264 

Eopsetta  642-3,  654,  659.  660,  666 


Ephippus  465 

Epigonus  469.  ^^70,  487 

Epinephelus  18.  499.  500.  507-8,  509 

Epmnula  593,  594,  596,  600-1,  603 

Epiplatys  359,  364 

Eretmophorus  266,  268,  272,  280 

Ereunias  443,  445 

Erilepis  442,  444 

£np5  103 

Erisphex  44 1 

froifl  440 

fra/efc  584,  586-7 

Erythrocles  473 

Eschmeyer  441 

Escualosa  109,  124 

Etox  7-^0-;,  142,  202 

Ethadophis  103 

Ethmalosa  113,  125 

Ethmidium  113,  122 

Etropus  642,  650,  651,  674-5,  685 

Etrumeus  15.  28.  29,  102,  108,  110,  //.S,  720,  122-4 

Eucinostomus  473 

Eucitharus  641,  673-4,  676,  681,  684 

Eudichthys  260-1,  263,  26^^.  266,  280 

Euleptorhamphus  336-7,  352 

Eumecichthys  368-9.  371-2.  277-9 

Eumicrotremus  iO.  31.  429-30.  ■^i/-2,  434,  437 

Eupleurogratnmus  599-601.  604 

Euryglossa  657,  664,  665 

Eurymen  4AA 

Eurypegasus  402 

Euryslole  355-6,  i57-5 

Eustomias  171-2,  175-6.  /79.  181.  183 

Eutaeniophorus  380-1,  iiS2 

Euthynnus  542.  600-1.  606.  6/0,  612,  617,  619 

Evermannella  207-8,  257,  252-4 

Eviota  587 

Evorthodus  582 

Evoxymetopon  599-601,  604 

Exechodontes  578 

Exocoetus  338,  i-^/,  345,  i'#9,  352 

Expedio  584.  586 

Filimanus  540 

Fistularia  398.  400-1,  403-4 

Flagelloserranus  508 

Flagellostomias  171-2,  174,  /76,  181,  183 

Florenciella  469 

Floridichthys  365 

Fluviphylax  362 

Foa  -^70 

Fodiaior  335,  338-9,  i-//,  343,  i'/9.  352,  354 

Foetorepus  637 

Forcipiger  21.  474 

Franzia  500,  503 

Fw^  448.  '/'^9 

Fundulopanchax  364 

Fundulus  43.  53,  359,  362-5,  i65-7 

Furcina  442 

Ga£^e//a  266.  268.  272,  279-80 

Gadiculus  266.  268,  279-80.  284,  286-7,  290.  292,  294-5,  297 


INDEX 


751 


Gadomus  267,  272.  274-5 
Gadopsis  469-70,  473,  482 
Gadus  15,  29,  31,  260,  266,  268,  278.  279,  281,  284,  286-7, 

290,  292.  294-7 
Gadusia  1 1 2 

Gaidropsarus  266,  269,  279-80,  284-90,  294-5,  297-9 
Galaxias  150-1,  152.  153 
Galaxiella  150-1,  153 
Galeoides  540,  547 
Gambusia  363-4 
Gargariscus  4 1 9 
Gargaropteron  21 
Gargilius  279 
Garialiceps  7 1 

Gasterochisma  591-2,  600-1,  617-8 
Gasteroclupea  122 
Gasterosteus  398.  399 
Gastrocyathus  629-30,  6ii 
Gastropsetta  642,  651.  674,  682 
Gastroseyphus  629-30,  633.  6i5 
Gempylus  593-4,  599,  600-1,  603-4.  615-7 
Genioliparis  429 
Ge«yp/erM5  309,  312,  318 
Gephyroberyx  392 
Gerlachea  563 
Gibber ichthys  7,  15,  391-2 
Gigantactis  325.  i29.  333 
Giganthias  500 
Gigantura  199,  201 
Gilbertidia  410,  424,  427,  447 
GHchnstella  113,  122,  125 
G(re//a  47/ 
Glenoglossa  103 

Glossanodon  156-7,  759,  163-4,  168-9 
Glossolepis  360 
Glyptauchen  440 

Glyptocephalus  28.  29,  642-3,  659.  660.  664 
Gnathanacanthus  440 
Gnathanodon  512,  J/5,  518,  524 
Gnathophis  66,  70 
Gnathopogon  131 
Gobiesox  629-31,  633,  6i5 

GOBIIDAE  55i 

Gobioides  585.  587 

Gobiomorphus  589-91 

Gobionellus  584.  587 

Gobiosoma  584-5 

Gonialosa  1 1 1 

Gomchthys  218-9,  221-2,  229.  234,  240-3 

Gomoplectrus  500,  508,  510 

Gonorhynchus  138.  139 

Gonostoma  182,  185-91,  794.  195,  198 

GordiichthyslX.  103,  107 

GracUia  500 

Grahamichthys  590 

Gramma  465 

Grammatobothus  642,  652,  657,  683 

Grammatonotus  464-5,  467 

Grammatorcynus  592,  600-1,  606,  605.  612,  617-9 

Grammatostomias  171-2,  183 

Grammicolepis  393 

Grammistes  501,  508 


Grammistops  501 

Grasseichthys  138 

Gulaphallus  356-1 ,  361 

Gunnellichthys  582,  587 

Gymnammodytes  574-5 

Gymnapistes  440 

Gymnelopsis  578 

Gymnelus  578-80,  557.  582 

Gymnocaesio  468 

Gymnocanthus  409,  427.  442 

Gymnocorymbus  129 

Gymnodraco  563 

Gymnosarda  592,  600-1,  606,  609.  612,  617-8 

Gymnoscopelus  218,  220-2,  226,  229.  236,  241-3 

Gymnothorax  72-3 

Gyrinichthys  429 

Hadropareia  578 

Hadropogonichthys  578 

Haemulon  51,  53 

Halargyreus  262,  266 

Halichoeres  544 

Halosauropsis  94-5 

Halosaurus  94-5 

Hapalogenys  465,  466,  480,  485,  487,  489-90.  493 

Haplophryne  326-7,  i2«.  334 

Harengula  110,  775.  122-4 

Harpadon  206-8,  212 

Harpagifer  562.  563-4 

Helicolenis  406,  410,  472.  439-40 

Hemanthias  500,  503,  505,  507 

Hemerocoetes  557 

Hemerorhinus  103 

Hemtcaranx  5\1,  518,  520.  524 

Hemilepidotus  409,  47  7.  425,  427,  442,  446 

Hemilutjanus  465,  480 

Heminodus  4 1 9 

Hemirhamphus  336-9,  i40.  i46.  347,  i57,  352 

Hemirhamphodon  337,  339,  345,  352-3 

Hemitripterus  409,  421,  425,  428,  442,  446-7 

Herklotsichthys  109.  123 

HermosiUa  469 

Herpetoichthys  103 

Heterandria  362 

Heterenchelyidae  5i 

Heteromycteris  657,  665.  687 

Helerophotus  171-2 

Heteropneustes  1 30 

Heterostichus  564,  570 

Heterothrissa  117,  125 

Hexagrammos  20,  410,  474.  421-2,  443-5 

Hildebrandia  101 

Hildebrandichthys  1 1 7 

//;75a  111,  123-4 

Himantolophus  325-7,  iiO 

HintoniallS.  220-1,  241-3 

///0£^oi7  126,  129,  133 

Hippocampus  53,  i95.  400,  402,  403 

Hippoglossina  642,  646.  651-2,  674,  657.  682 

Hippoglossoides  29,  640,  647.  642-3,  654,  659,  660 

Hippoglossus  14,  642-3,  654,  659,  660,  666 

Hirundichthys  338-9,  i47,  343,  352,  354 


752 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Histiodraco  563 

Histiophryne  'il^ 

Histrio  321 

Holanthias  500,  505.  507 

Holocentrus  18,  384,  i85.  387,  392 

Hoplias  133^,  136-7 

Hoplichthys  408,  414,  421 

Hoplolatilus  468,  495 

Hoplostethus  389,  392 

Hoplunnis  94-95,  96 

Horaichthvs  335,  339.  345,  347,  352 

Howella  465,  470,  480,  483,  487,  493 

Hozukius  A\Q,  AAQ 

Hucho  143-4,  ;¥6,  148-9 

Hygophum  218-9,  221-2,  226,  228.  234,  241-3 

Hymenocephalus  269,  274-5,  279 

Hyoinacrurus  274-5 

Hypentelium  129 

Hyperlophus  1 12 

Hyperoglyphe  611,  625-6,  628 

Hyperoplus  574.  575 

Hyphalophis  103 

Hyphessobrycon  133-4 

Hypoatherina  360 

Hypomesus  1 54 

Hypophthalmichthys  129,  131 

Hypoplectrus  500-1,  503 

Hvpoptychus  400-4 

Hyporhamphns  337-9,  i42,  345,  347,  352-3 

Hypsagonus  425,  426,  428,  442,  447 

Hypsoblennius  3 1 

Hypsopsetta  29,  643,  654,  659 

Icelinus  419,  442 

/ct/!/j409,  479.  442,  445 

Ichthyapus  7 1 

Ichrhyococcus  185-9,  191,  792.  198 

Ichthyupus  103 

Icichthys  29,  621-3,  625-6,  628 

Icosteus  28,  29,  576-7 

Ictalurusl9,  130,  134.  137 

Idiacanthus  11,  171-2,  175,  /«0,  181,  183-4 

//(5/ia  112,  117,  720,  123,  125 

lluocoetes  578-80 

Indostomus  404 

Inermia  49 1 

Inimicus  405,  407,  419,  440 

Inistius  544 

/;7/jop<r  207-9,  2/2,  256,  258 

Irialherina  360 

/50  355,  357,  i59,  360 

Isopsetta  15,  642-3.  654,  660 

Istiblennius  570 

Istiompax  604,  605 

Istiophorus  600-1,  604,  605,  607-8,  614-7 

Japonolaeops  642,  686 
Jeboehlkia  500,  508 
Jenkinsia  33,  111,  122,  123-4 
Jenynsia  363 
Jordanella  363,  365 
Jordania  442 


7i:a/(20,  555 

Kamoharaia  642,  652 

Kanazawaichthys  322 

Kanekonia  441 

Kareius  642,  654 

Kasidoron  7 

Katsuwonus  24,  26,  592,  600-1,  606,  670.  612,  617,  619 

Kaupichthys  100 

Kertomichthys  103 

Kneria  138 

Knightia  122 

Konosirus  111,  124 

Korogaster  389 

Kraemaria  588 

Krefflichlhys  218-9,  221,  223,  225,  229,  241-3 

Krohnius  269,  279 

Krusensterniella  578 

A:«w6a  274-5 

Kyphosus  469,  47i 

Labeotropheus  542 

Labracoglossa  469,  471,  49 1 

Labrisomus  53 

Labroides  544 

Labrus  544 

Lachnolaimus  544,  545 

Lactophrys  458 

Lactoria  44.  48,  452 

Laemonema  261,  266-7,  279-80 

Laeop5  642,  652,  656 

Laeviscutella  112,  125 

Lagocephalus  448,  449 

Lamnostoma  103 

Lampadena  218,  220-2,  226,  229,  236,  240-3 

Lampanyctodes  218,  220-1,  223.  226,  229,  236,  241-3 

Lampanvctus  218,  220-1,  223,  226,  2i7-,S.  239,  240-3,  257 

Lampichthys  218,  220-1,  223,  226,  229,  236,  241-3 

Lampris  368-9,  371-2,  i74-5.  377-9 

Lateolabrax  465,  469,  47i.  480,  485,  487,  493,  509 

La/e5  469 

Latimeria  55 

Le/«a  131 

Leiocottus  443 

Leiognathidae  465 

Leiostomus  40,  42 

Leiuranus  103 

Lepadichthys  629-30,  633,  6iJ,  636 

Lepadogaster  629-30,  6ii 

Leptdion  261,  266-7,  280 

Lepidoblepharon  640,  676,  678,  681-2,  684 

Lepidocephalus  1 3 1 

Lepidocybium  593.  594-6,  600-1,  603 

Lepidogalaxias  139,  202,  205 

Lepidophanes  218,  220-1,  223,  226,  2i5,  236,  240-3 

Lepidopsetla  15,  642-3,  654,  659 

Lepidopus  599-601,  602,  606-7 

Lepidorhombus  640,  643,  646 

Lepidorhynchus  274-5 

Lepidotrigia  407.  419 

Lepophidium  311.  312-4,  318 

Leptenchelys  102-3 

Leptobrama  464.  477.  479 


INDEX 


753 


Leptocephalus  62,  63.  69-70,  94,  96 

Leptocottus  409,  422,  All,  443,  447 

Leptolucania  363-4 

Leptophilypnus  589-91 

Leptosynanceia  440 

Lepturacanthus  599-601,  604 

Letharchus  71,  103 

Lethogoleos  103 

Lethotremus  429-30 

Lethrinus  474 

Leucichthys  143,  148 

Leuciscus  133-4 

Leuresthes  355-6,  360,  362 

Leuropharus  103 

Lhotskta  345 

Lichia  512,  518,  523-4,  526,  528-9 

Lile  110,  124 

Limanda  640,  641.  642-3,  654,  660 

Limnichlhys  557 

Limnothnssa  1 1 3 

Linophryne  325-6,  i2<S,  329,  333-4 

Liocranium  440 

Lioglossina  642,  651,  674,  682 

Lionurus  274-5 

Liopropoma  20,  42,  500,  507.  508,  510 

Liopsetta  642-3,  654,  660 

L/pans  429-30,  ^iA  432,  433-4.  436-7 

Lipariscus  429,  432 

Lipogenys  94-5 

Lipogramma  465 

Lobianchia  218-9,  221,  223,  226,  229,  234,  241-3 

Lobotes  466.  490 

Lophiodes  320 

Lophius  269,  320,  i2/ 

Lophodolos  333 

Lopholatilus  18.  495 

Lophonectes  642,  652,  655.  657,  683 

Lophotus  19.  368-9,  i70,  371-2,  377,  i79 

Loricaria  129-30 

Loricariichthys  1 30 

Lorn  261,  266,  268,  280,  284,  255.  286-9,  294-5.  297-9 

Lo?e//a  266,  279-80 

Loweina  19.  218-9,  221,  223,  229,  234,  240-3 

Lucama  359,  363-4,  i65,  368 

Luciogobius  582,  55i.  585-6 

Lutjanus  468.  469 

Lwvan«547.  550,  591 

Luzonichlhys  500,  503 

Lycenchelys  578,  582 

Lycengraulis  117-8,  125 

Lvcodapus  578-80,  582 

Lycodes  578,  580,  582 

Lycodichthys  578 

Lycodonus  578 

Lycogrammoides  578 

Lyconectes  57 1 

Lyconema  578 

Z,yconM.s  261-3,  267,  269,  272 

Lycothhssa  1 1 7 

Lvcozoarces  578 

Lyopsetta  642-3,  654,  659,  660.  666 

Lyosphaera  448,  450 


Maccullochella  469,  473 

Macquaria  469 

Macristiella  209 

Macristium  2 1 1 

Macroparalepis  207-8,  259 

Macropinnal 56-S,  161,  163-4,  767.  168 

Macrorhamphosodes  452-3 

Macrorhamphosus  398.  399-400,  407,  -^Oi,  404-5 

Macrosmia  274-5 

Macrostomias  171-2,  174,  183 

Macrouridae  272,  276 

Macrouroides  274-5 

Macrounis  269,  274-5 

Macrozoarces  578-80,  55/.  582 

Macrurocvttus  394,  i96 

Macruronus  262-3,  267,  269,  272,  281 

Magalespis  5\1,  518,  524 

MflA/a  274-5 

Makaira  600-1,  604,  605.  607-8 

Malacanthus  495 

Malacocephalus  269,  274-5 

Malacocottus  410,  -^24,  428,  444,  446-7 

Malacosteidae  779 

Malacosteus  171-2,  183 

Malakichthys  464 

Mallotus  154-5 

Malvolwphis  103,  106 

Manacopus  36 1 

Mancopsetta  642-4,  649,  652,  670,  672-4,  678,  681,  683-4, 

686-7 
Manducus  185-90,  193,  195,  198 
Margrethia  182,  185-6,  188-91,  79i.  195,  198 
Marlyella  643 
Marukawickthys  443-5 
Mastums  448,  450,  463 
Mataeocephalus  274-5 
Maulichthys  207-8 

Maurolicus  43.  44.  185-6,  188,  190,  797.  269 
Maynea  578,  580 
Medialuna  469,  ^^77 
Megalocottus  443 
Megalomycter  382 
Megalops  60,  67,  62,  98-99,  126 
Melamphaes  386-7.  389.  392 
Melanocetus  327,  iiO 
Melanogrammus  31,  267-8,  279,  281,  284,  286-9,  292.  294- 

5,  297 
Melanonus  260-3,  266,  268,  270.  280 
Melanostigma  578-80,  582 
Melanostomias  171-2,  174,  775.  181,  183-4 
Melanotaenia  355-7,  360-1 
Melapedalion  337-43,  352-3 
Membras  355-6 
A/e«e  -#65.  479 
Menidia  26,  355-7,  362 

Merlangius  267-8,  284,  286-9,  292,  294-5,  297 
Merluccius  25,  29,  iO.  51,  260-5,  267,  269,  272,  275.  279,  281, 

283,  294,  297-9 
Mesobms  269,  274-5,  277 
Mesocottus  443,  445 
Metacottus  443 
Metavelifer  368-9,  372,  i76.  378-9 


754 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Metelectrona  218-9,  221,  223,  225.  229,  240-3 

Metynnis  1 3 1 

Microcanthus  469,  473 

Microchints  644,  649,  657,  660,  664,  668.  611 

Microcottus  443 

Microdesmus  585.  586-7 

Microgadus  266-8,  281,  284,  290,  294-5,  297 

Microgobius  583,  587 

Micwlabrichthys  500 

Microlepidium  266 

Microlophichthys  327,  331 

Micromesistius  267-8,  281,  283-4,  286-7,  290,  294-5,  297 

Micropercops  590 

Micropterus  5 1 

Microspathodon  543,  544 

Microstoma  156-7,  75,S,  161.  163-4,  169 

Microstomus  640,  fi-/;.  642-3,  654,  659,  660,  664 

Microthrissa  1 1 2 

Mimasea  599 

Minous  407.  414.  419 

Minysynchiropus  637 

Mirapinna  380,  i«/ 

Mirophallus  361 

Misgiirnus  129.  131,  137 

A/o/fl  448,  450,  463 

Molacanthus  448,  450 

Mo/va  266,  268,  280-1,  284-289,  255,  294-5,  297-9 

Monacanthus  452,  455 

Monocenirus  392 

Monodactylus  467 

Monolene  (,A2.  652,  657,  672-4,  682-3 

A/ora  261,  266-8,  272.  281 

Moringua  64,  70,  72,  95-6 

MORINGUIDAE  93 

Moroco  131 

Moronf'  25,  29,  49,  469,  473.  485,  489,  508,  615-7 

Mugil  24-5,  29,  iO.  -^^  48,  531.  532-3,  541 

MURAENESOCIDAE  81 

Muraenichthys  103,  105 

MURAENIDAE  89 

Muraenolepis  260-3,  265-6,  268,  270.  279-80,  282,  319 

Mycteroperca  500,  508 

Myctophum  11,  75.  218-9,  221,  223,  226,  225-9,  234,  241-3 

Myoxocephalus  409,  419,  421.  427,  443,  446-7 

MyrichthyslX,  73,  103 

Myripristis  384-5.  392 

Myrophinae  77 

Myrophis  69-70,  103,  yO-/,  105 

Mysthophis  103 

Mystus  133-4 

Nalbantichthys  578 

Nanichthys  335,  340,  342,  350,  352 

Nannatherina  469 

Nannoperca  469-70 

Nansema  156-7,  755,  161,  163-4,  169 

A'aio  548,  5-^9 

Naucrates  498,  510,  512,  5/6,  518,  523-4,  526,  527 

Nautichthys  409,  -^25,  428,  443,  447 

Navodon  452,  455 

Nealotus  593-4,  596,  599,  600-1,  603 

Neaploactis  44 1 


Neat y pus  469 

Nectoliparis  429,  432 

Neenchelys  103,  704,  106 

Nemacheilus  129,  131,  137 

Nemanthias  500 

Nematolosa  111,  114 

Nematistius  496-8,  522-3,  526,  527,  528,  529-30 

Nematonurus  274-5 

Nematops  643,  684 

Nemichthyidae  55 

Nemichthys  69,  72,  97 

Nemipteridae  471 

Neoachiropsetta  674 

Neobythitinae  i09 

Neocentropogon  440 

Neoceratias  326-7,  i25 

Neochanna  150-1,  153 

Neoconger  69,  95,  96 

Neocyema  72 

Neoepinnula  593-4,  596,  600-1,  603 

Neolaeops  642-3,  652,  686 

Neonesthes  171-2,  183 

Neoodax  544 

Neoopisthopterus  112,  125 

Neopagetopsis  563 

Neopataecus  440 

Neophos  185 

Neophrinichthys  444 

Neoscombrops  464 

Neoscopelus  207-9,  27  7,  243,  257 

Neoscorpis  465,  469,  482 

Neostethus  36 1 

Neosynchiropus  637 

Nesiarchus  593-4,  599,  600-1,  603,  615-7 

Nesogalaxias  150,  151 

Nettastoma  96 

Nettastomatidae  69,  55 

Nezumia  267,  269,  274-5 

Nicholsina  544 

Mp^zow  500,  507,  510 

Nomeus  621,  622-3,  625 

Nomorhamphus  337-8,  352-3 

Normamchthys  4\0,  416.  422,  443,  445 

Notacanthus  94-5 

Notemtgonus  1 3 1 

Notestes  440 

Nothobranchius  362-4,  368 

Notolepis  207-8,  277,  218,  256 

Notoliparis  429,  430 

Notolychnus  218-9,  221,  223,  226,  229,  234,  240-4 

Notolvcodes  578 

Notoscopelus  218,  220-1,  223,  226,  229,  236,  241-3 

Notothenia  563-4 

Notropis  29,  131 

Novaculichthys  544 

Novumbra  140,  142,  202 


Oce//a  '^26,  428, 
Ocosia  440-1 
Ocyanthias  500 
Ocynectes  443 
OrfflX  544,  546 


442 


INDEX 


755 


Odaxothrissa  1 1 2 

Odontanthias  500 

Odontesthes  355-7,  359 

Odontognathus  114,  125 

Odontolipahs  429 

Odontornacnirus  269,  274-5 

Odontopysix  442 

Odontostomias  174,  181,  183 

Odontostomops  251,  254 

Odonus  452 

Ogcocephalus  321,  322 

Oidiphorus  578 

Oligocottus  409,  422,  427,  443,  447 

Oligopliles  510,  512,  5/7,  518,  522-4,  528-9 

Oligopus  3 1 1 

Omosudis  200,  207-8,  272,  216,  245,  259 

Ompak  130-1 

Oncoptems  643,  673,  684 

Oncorhynchus  20.  44,  143-4,  /-/d.  148-9 

Oneirodes  325,  327,  ii7,  333 

Onuxodon  312,  314,  i;6.  318 

Ophichthidae  79 

Ophichthinae  77 

Ophichthus  65-6,  65,  69,  71,  103,  705,  106-7 

Ophidian  28,  311.  312-4,  318 

Ophiodon  410,  -^/^  421-5 

Ophisurus  103 

Ophthalmolycus  578 

Opisthonema  109,  //<?,  122,  124 

Opisthoproctus  156-8,  161,  164,  767 

Opisthopterus  113,  118,  123,  125 

Opistognathus  467 

Oplegnathus  467 

Opostomias  171-2,  174,  776,  181,  183 

Opsarichthys  133-4 

Op?/vws  389,  i90 

Orbonymus  637 

Orcynopsis  592,  600-1,  606,  612 

Oreosoma  394,  i96 

Orthonopias  409,  421,  427,  443 

Onc/a<r  335,  339,  i42,  345,  347.  i49,  352,  357,  368,  533 

Osmerus  29,  129,  131,  133,  153,  154 

Osteodiscus  429-30 

Ostichlhys  392 

OSTRACIIDAE  iO,  ¥5i 

Ostracion  452,  459 

Otophidium  311.  312-4,  318 

Oxyjulis  544 

Oxylebius  410,  '#7'^.  422,  443-5 

Oxyporhamphus  44.  48.  337,  i-^O.  343,  345,  i46,  347,  353 

Pachycara  578 
Pachystomias  171-2,  183 
Pagetopsis  562.  563 
Pagothenia  563 
Pa^rws  489-90 
Paleogadus  262-5 
Pallasina  408,  442 
Palunolepis  47 1 
Pampus  622-3,  625-8 
Pangasius  1 30,  1 34 
Paniolabus  5  \  2-3,  518,  524 


Papuengraulis  117,  122 

Paraaploactis  44 1 

Parabothus  643,  652,  686 

Parabramis  133-4 

Parabembas  44 1 

Paracallionymus  14,  44,  48,  637,  6i9 

Paracentropogon  440 

Paracentropristis  500 

Paracetonnrus  274-5 

Parachaemchthys  563 

Paraclinus  570 

Paraconger  70 

Paracottus  410,  443 

Paradiplogrammus  637 

Paradiplospinus  593-4,  600-1,  603-4 

Paragalaxias  150-1 

Parahemi nodus  4 1 9 

Parakneria  138 

Parakumba  274-5 

Paralabrax  499,  500-1,  509 

Paralepis  201-S,  217 

Paraletharchus  103 

Paralichthodes  643,  673,  684 

Paralichthys  24,  i9,  640.  6-^7,  642-3.  646,  648,  651-2,  674, 

682,  685-6 
Paraliparis  429-30,  434,  437 
Paranthias  500,  507,  508,  510 
Parapercis  560 
Paraplagusia  643,  664 
Parapsettus  465 
Parapterygotrigla  4 1 9 
Parasalmo  143-4,  746.  148-9 
Parascorpis  469 
Parasilurus  1 30 

Parastromateus  5\2,  518,  527,  526.  525.  530 
Parasudis  206-8,  258 
Parataeniophorus  380,  i5/.  382-3 
Paratrachichthys  392 
Paraxenomvstax  94 

Parexocoetus  335,  338-9,  343,  i49,  350,  352,  354 
Pancelinus  409,  479.  443 
Panl-fl  452,  455,  457 
Parkraemaria  590 
Parana  512,  518,  522-4,  528 
Parophidion  i 77,  312-3,  318 
Parophrvs  642-3,  654,  659,  660 
Parvilux  218,  220-1,  224,  2i5.  236,  240-3 
Pataecus  440 

Patagonotothen  562,  563-4 
Paxanovia  578 
Pegasidae  400 
Pegasus  404 

Pf,gM5a  640,  647.  644,  649.  657,  660,  664 
Pelagocyclus  429.  432 
Pelecanichthys  643,  652,  656,  673,  683 
Pe//ona  1 1 3 
Pellonula  112,  125 

Pelotretis  640,  647.  643-4,  660,  667.  673,  684 
Peltorhamphus  640,  647,  643-4,  660,  673,  676 
Pempheris  467 
Pentanemus  540 
Pentherichthys  327,  ii7 


756 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Pepnlus  622-3,  624,  625,  627-8 

Perca  29 

Perca  473 

Percolates  469 

Percichthys  469 

Percilia  469 

Percis  442 

Percopsis  55 

Percottus  582 

Perissias  642,  652,  657,  672,  674,  682-6 

Penstedion  407,  419 

Peristrominous  44 1 

Perryena  440 

Petalichthys  342,  350,  352 

Phaenomonas  1  \ ,  103 

Phallocottus  443 

Phallostethus  361 

Phenacostethus  361 

Philypnodon  590 

F/!o/;5  577 

Phottchlhys  185-8,  193 

Photoblepharon  392 

Photocorynus  326,  333 

Photonectes  171-2,  174,  181,  183-4 

Photostomias  171-2,  175-6,  779,  181,  183 

Phractolemus  138 

Phractura  130 

Phrynichthys  333 

Phrynorhombus  640,  643,  6-^6,  651 

Phucocoeles  578-80 

F/j.vm  264,  266,  268,  279-80,  284-9,  290,  292,  294-5,  297-9 

Phyllophichthus  103 

Physiculus  266-8,  272,  279-80 

Ptedrabuenia  578 

/'/Area  501,  508 

Pisodonophis  103 

Plancterus  363-4 

Platamchthys  1 10 

P/ara.x-  465 

Platichthys  44,  642-3,  654,  659,  660 

Platybelone  342,  i45,  350,  352 

Platycephalidae  ^/-^ 

Platycephalns  407,  421 

Plecoglossus  154.  202-4 

Plectranthias  500,  503,  505 

Plectrogenium  439,  441 

Plectroplites  469 

Plectropomus  500 

Plectrypops  392 

Pleuragramma  563 

Pleurogrammus  410,  47-^.  421-2,  443-5 

Pleuronectes  25,  260,  642-3,  660 

Pleuronichthys  16,  29,  44,  48,  640,  6-^7,  643,  648,  654,  659. 

660,  666 
Pliosteostoma  1 1 4 
Poecilia  364 
Poeciliopsis  364 

Poecilopsetta  643,  660,  667,  684 
Poecilothrissa  1 12 
Pogonolycus  578 
Pogonoperca  50 1 
Pogonophryne  563-4 


Polistonemus  540 

Pollachius  261,  267-8,  279,  281,  283-4,  286-9,  292,  294-5, 

297 
Pollichthys  185-89,  792,  193,  198 
Polyacantiwnotus  94-5 
Polydaclylits  540-1 
Polyipnus  185-91,  797 
Poly  met  me  185-9 
Poly  mixta  379 
Polynemus  540 
Poly  per  a  429 

Polyprion  465,  ¥66,  482,  485,  488,  490,  494 
Potnacentrus  544 
Pomatomns  47 1 
Ponttmis  406-7,  413.  414 
Popondetta  361 
Portchthys  324 
Porocottus  443 
Porotnttra  387.  389.  391-2 
Potamalosa  113,  122-3 
Potatnorrhaphis  350 
Potamothrissa  1 1 2 
Priacanthidae  47-/ 
Prtacanthus  474 
Prtonedraco  563 
Prionotus  28.  29,  407,  ^Z-^,  419 
Pristigaster  1 1 3 
Procatopus  364 
froco»!/i  4 10,  422,  443 
Profundulm  366 
Progmchthys  338-9,  349,  354 
Protnethtchthys  594,  596.  599-601,  603 
Promtcrops  500,  507 
Pronotogrammus  500,  503,  505.  507 
Prosopium  143-5,  148,  204-5 
Prosoproctus  A  A 1 
Prosphyraeria  534 
Proteracanthus  465 

Protomyctophiim  218-9,  221,  224,  225,  229,  241-3,  256 
Protosphyraena  534 
Prototroctes  151,  752.  153,  203-4 
Psatntytodiscus  643,  684 
P^ews  622-3,  62-/.  625.  627-8 
Psenopsis  623,  627-8 
Psettichthys  642-3,  654,  659,  660 
Psettina  643,  652,  65i.  657 
Psettodes  640,  6-^5,  649,  672-3,  675,  676.  678 
Pseiidalutarts  452,  '^55.  456 
Pseudamia  470 
Pseudamiops  470 
Pseudanthtas  500 
Pseudobalistes  452 
Pseudoblennius  409,  443 
Pseudocalliurichthys  637 
Pseudocaranx  512,  518,  526 
Pseudochaentchthys  563 
Psendochetltnus  544 
Pseudochromidae  -^67 
Pseudocyttus  394 
Psendogramma  501,  508,  509 
Pseudolabrus  544 
Pseudomugil  356-7,  361 


INDEX 


757 


Pseudomyrophis  1  \ ,  103,  105-6 

Pseudonezumia  274-5 

Pseudopentaceros  474,  490 

Pseudophrites  563 

Pseudophycis  266 

Pseudopleuronectes  24,  642-3,  654,  660 

Pseudopristopoma  473 

Pseudorhombus  642-3,  646.  648,  651-2,  674,  682-6 

Pseudoscopelus  557 

Pseudosynanceia  440 

Pseudotylosaurus  336,  350,  352 

Psilodraco  562.  563 

Psychrolules  4\0,  424.  427-8,  444,  446-7 

Ptarmus  441 

Pteragogus  544 

Pterengraulis  1 1 7 

/•?ero;5  416,  405 

Pterosmaris  47 1 

Pterosynchiropus  637 

Pterothnssus  60,  67,  62,  99 

Pterygotngla  4 1 9 

Ptilichthys  565,  570 

Pyramodon  312,  314,  i  77.  318-9 

Quassiremus  103,  106 

Rachycentron  474.  489-90,  -/Sd,  497-8 

Raconda  1 1 3 

Racoviizia  563 

Radiicephalus  368-9,  371-2,  i76,  378-9 

Radulinus  409.  ^2/,  427,  443 

Rainfordia  501 

Ramnogaster  1 10 

Ramceps  261-2,  266,  268,  279,  284-90,  294-5,  297-9 

Ranzania  448.  -/^P,  450,  463 

Rasirelliger  592,  594,  600-1,  606,  612.  618-9 

Regalecus  368-9,  371-2,  i7i.  375,  377-9 

Reinhardtius  642-3.  654,  659,  660 

Repomucenus  637 

Reporhamphus  353 

Retropinna  150-1.  752,  153,  203-4 

/?e.vea  593,  596,  599-601.  603 

Rhadtnesthes  171-2,  183 

Rhadinocentrus  36\ 

Rhadulinopsis  443 

Rhamphocottus  409,  '/77,  422,  425,  443,  446 

Rhinesomus  452.  ^55.  459 

Rhinocephalus  262-5 

Rhinogobius  584 

Rhinoliparis  429-30 

Rhinoprenes  465 

Rhinosardina  1 10 

Rhizophryne  328 

Rhodichthys  429-30 

Rhomboserranus  464 

Rhombosolea  643-4,  660,  673,  676,  6.^0,  684 

Rhyactchthys  586.  588,  590 

Rhynchactis  325-7,  i29.  333 

Rhynchogadus  266,  268,  272 

Rhynchohyalus  156-7,  163-4,  767,  168 

Rhynchorhamphus  337-8,  340,  352 

Richardsomchlhys  440 


Ricuzenius  443 

Rimicola  629-30.  633.  6i5 

i?;Va  133-4 

7?/vM/i«  362-3,  i65-6.  368 

/?oft;a  333 

Roccus  469 

Rondeletia  382 

Ronquilus  57 1 

Rosaura  7,  199,  200,  201 

Rosenblattia  469 

Rosenblaltichthys  246.  247-250.  256 

Rudarius  452 

Ruvettus  591,  593-4,  596.  600-1.  603 

i?ypncw5  501.  507.  508 


Saccopharyngoidea  93 

Sacura  -#99.  500,  507 

Salangichthys  154.  155 

Sa//7ora  266,  268,  280 

Salmo53.  143-4.  146.  148-9 

Salvelinus  143-4.  7-^6.  148-9 

Samaris  643.  660,  667.  672-3 

Samahscus  643,  660,  667,  657,  684 

5ar^a  592^,  600-1.  606,  609,  612.  617-8 

Sardina  108-9,  123-4 

Sardinella  109.  118,  720,  122-4 

Sardinops  18,  29,  iO,  108,  110,  77«.  123^ 

Sarritor  442 

Satyrkhthys  4 1 9 

Saurenchelys  96 

Saurida  207-8,  272,  258 

Sauromuraenesox  94 

Scalanago  65 

SCARIDAE  546 

Scarus  544 

Scatophagus  474 

Schedophilus  621.  622-3,  625-8 

Schmdleria  1 1,  552,  55i,  554 

Schismorhynchus  103 

Schultzea  500-1 

Schultzidia  103 

Scomber  24,  592,  594,  600-1,  606,  605,  615-9 

5cowZ)eresox  44,  48,  335,  ii6,  ii9,  340,  347,  350,  352,  354 

Scomberoides  5\2,  516.  518,  522-4,  528-9 

Scomberomorus  591-2,  594,  600-1,  606-7.  609,  612.  617-8 

Scombrolabrax  591-2.  59i,  594.  599-601.  615-7 

Scombrops  474.  490 

Scopelarchoides  245.  2-^6,  247-250.  256 

Scopelarchus  207-8,  211,  245,  246,  247-50,  256 

Scopelengys  207-8,  277 

Scopeloberyx  387.  389.  392 

Scopelogadus  387.  392 

Scopelopsis  218,  220-1,  224,  226,  229,  236,  241-3 

Scopelosaurus  207-8,  272 

Scophthalmus  27,  640,  647,  643,  646,  651,  676 

Scorpaena  405,  407,  47i.  414,  439,  445 

Scorpaenichthys  409,  477,  425,  427,  443,  446 

Scorpaenodes  407,  47i.  414 

Scorpis  469 

Scytalichthys  103 

Sebastes  13.  23,  44,  405-6.  410,  47  7-2,  438-40,  445 


758 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Sebastiscus  440 

Sebastolobus  20.  405,  407,  413.  416-7,  439-40,  445 

Sectator  469 

Selar  5\2,  514.  518-9,  522,  526 

Selaroides  512,  518,  526 

Selenanthias  500,  507 

Selene  512,  518,  520.  522,  526 

Semicossyphus  544 

Semotilus  129 

Seriola  497.  498,  510,  512-3,  517.  518,  520,  523-4,  528-9 

Seholella  622-3,  627-8 

Seriolina  518-9,  528 

Serraniculus  501,  503.  509 

Serranus  500-1,  50i,  509 

Serrasalmus  129,  ;iO,  131,  133,  134 

Serrivomeridae  87 

Serhvomer  69 

Setipinna  117,  ;;,S,  125 

Sicydium  55 

Sierrathrissa  108,  114 

Siganus  26,  548,  549 

Sigmistes  443 

Silonia  137 

Siniperca  465,  469,  -^Zi.  484-5,  489,  495 

Snydehdia  312,  314,  i/ 7,  318-9 

Snyderina  440-1 

5o/fa  644,  649,  657,  660,  664,  665,  672,  678 

Solenophallus  361 

Solenostomus  400.  402.  403-4 

Solivomer  207-8,  256,  258 

Sorto^a  185-8 

Sparisoma  544 

Spectrunctulus  3 1 1 

Sphagebranchus  269 

Sphagemacrurus  274-5 

Sphoeroides  448,  450 

Sphyraena  26,  533-4,  5i5-6,  537,  5iS-9,  541 

Sphyraenodus  534 

Sphyraenops  469,  ¥70,  487,  490 

Spinicapitichthys  637 

Spiniphryne  333 

Spirinchus  154.  155 

Sprat  el  loides  111,  122-4 

Spratellomorpha  1 1 2 

Sprattus  108,  110,  122-4 

Squalidus  131,  133-4 

Squalogadus  274-5 

Stegastes  544 

Steindachneria  261-4,  267,  269,  277,  282-3 

Stelgistrum  443 

Stellerina  426,  442 

S/e////i>r  -^70 

Stemonosudis  201-8,  217.  218,  259 

Stenatherina  360 

Stenobrachius  218,  220-1,  224,  2i5,  236,  241-3 

Stenodus  142-5,  148 

Stephanoberyx  392 

Stephanolepis  452,  -/f^,  455-6 

Stephanophyrne  327 

Stereolepis  465,  -^66,  484,  489-90,  495 

Sternias  443 


Sternoptyx  185-6,  188,  190-1,  797 

Stethojulis  544 

Sthenopus  44 1 

Stictorhinus  1\.  103 

Stigmatonotus  465 

Stlegicottus  443 

Stlengis  443 

Stokellia  150-1,  203-4 

Stolephorus  116,  7;S,  72/.  123,  125 

Stolothrissa  1 1 3 

Stomias  30.  31,  770,  171-2,  174,  181.  183 

Stromateus  622-3,  627-8 

Strongylura  336,  339,  i-^0,  i42.  i45.  i50.  352 

Stylephorus  368-9,  371-2,  375,  i77,  378-9 

Stylophthalmus  181 

5«a'(5  20,  155,  207-8,  216,  277.  256 

Sufflamen  452,  454 

Sundasalanx  204 

Suttonia  50 1 

Svetovidovia  7,  266-8,  270,  279-81 

Syacium  642,  651-3,  670,  674-5,  682,  685 

Symbolophorus  218-9,  221,  224,  226,  22<S,  229,  234,  241-3 

Symphodus  544 

Symphurus  28.  29,  643,  664,  667,  669.  670,  672 

Symphysanodon  464-5,  -^66,  483,  489-90,  495 

Symphysodon  542 

Synagrops  464,  ¥66,  490 

Synanceia  440 

Synaphobranchidae  75 

Synaptura  644,  657 

Synchiropus  ^Til 

Synchirus  443 

Syngnathus  400.  402.  403 

Synodus  16,  29,  48,  207-8,  216,  256,  258 

Taaningtchlhys  2\».  220-1,  224,  229.  236,  240-3 

Tactostoma  10,  171-2,  174,  776.  181,  183-4 

Taenwides  582,  586-7 

Taeniopsetta  642,  652,  65i,  657,  672,  682-3,  685-6 

Tanakius  642-3,  654,  659,  660,  66¥ 

Tandanus  129 

Taranetzella  578 

Tarletonbeania  218-9,  221,  224,  229.  234,  240-3 

Tarphops  642,  6¥6,  651,  674,  682 

Tarpon  98 

Taurocoltus  443 

Taurulus  409 

Tautoga  544 

Telmatherina  355-6,  361-2 

Temnocora  429 

Tentoriceps  599-601,  604 

Tenualosa  108,  111,  124 

Tephrinectes  642,  672,  674,  678,  650.  681 

Tetragonurus  622-3,  62¥,  625-8 

Tetraodon  448 

Tetrapturus  600-1,  604,  605,  607-8 

Tetraroge  440 

Tetrosomus  452 

Thalassenchelys  69-70 

Thalassoma  544,  5¥5 

Thaleichthys  154-5 

Thaumatichthvs  326-7,  iiO.  333 


INDEX 


759 


Thecopterus  443 

Theragra  29,  267,  269,  281,  284,  286-7,  290,  294-5,  297. 

298-9 
Therapon  473 
Thorophos  185-8,  190-1 
Thrattidion  108,  114 
Thrvssa  116,  121,  125 

Thunnus  592,  600-1,  606,  610,  612,  613,  615-7,  619 
Thymallus  142-4,  148,  204-5 
Thvriscus  443 

Thyrsites  591,  593-4,  596,  599-601,  603-4 
Thyrsitoides  593-4,  600-1,  603 
Thyrsitops  593,  59^^,  596,  600-1,  603 
Thvsanactis  171-2 

Thysanopsetta  642,  672,  674,  678,  681-2,  686 
Tilesina  442 
r/7Mru5  62,  63.  94 
Tiluropsis  62,  6i,  70,  94 
Tomeurus  364 

Tongaichthys  591,  593-4,  596,  600-1,  603,  612,  615-7 
ror^!<i,ge«er  448,  450 
Tosana  500 

Tosarhombus  643,  652,  686 
Trachelochismus  629,  630-1,  6ii 
Trachicephalus  440 
Trachichthyidae  i90 
Trachichthys  389,  392 
Trachidermus  443 
Trachinocephalus  lQl-9>,  212,  216 
Trachinolus  498,  513,  5/6,  518,  523-4,  526.  528-9 
Trachinus  31,  559,  560 

Trachipterns  30,  368-9,  i70,  372,  375,  377-9 
Trachonurus  274-5 

Trachurus  12,  14.  23,  510,  511,  513^,  518,  520,  526 
Trachycorystes  130 
Trachydermis  445 
Trachyrhynchus  1()1,  269,  274-5 
Trachyscorpia  440 
Tragulichthys  448,  '#'^9 
Triacanthodes  453 
Triacanthus  452,  '#5i,  454 
Tribolodon  129 

Trichiurus  600-1,  602.  604,  606-7 
Tnchodon  3\,  557.  564 
Trichonotus  557 

Trichopsetta  642,  652,  657,  672,  674,  682-3,  685-6 
Tng/a  419 

Thglops  409,  419.  443 
Trigonolampa  171-2 

Trinectes  640,  6-//,  644,  649,  657,  66<S,  650 
Tnphoturus  218,  220-1,  224,  226,  2i5,  236,  241-3 
Tnplophos  182,  185-90,  193,  198 
Tripterodon  465 
Tripterophycis  261,  266 
Tripterygion  570 

Trisopterus  267,  269,  284,  286-9,  292,  294-5,  297-9 
Trisotropis  500 

Trypauchen  584,  555.  587.  589-90 
TMftWa  622,  626-628 
Tvlosurus  336,  i40.  i45.  347,  350,  352 


t/Zua  513,  518,  526 

(7w*ra  140,  7-^7.  142,  202 

Uncisudis  207-8 

[/ra^;a419 

t/ra5p(5  513,  518,  527.  526 

Urophycis  264,  266,  268,  27<S,  279,  280-1.  284-9,  290,  294 

Uropterygius  72 

Valencia  364 

Valenciennellus  185-6,  188,  190,  797 

Fano/a  500 

Velambassis  467 

Velifer  368-9,  375 

Ke///ror  443 

Ventrifossa  274-5 

Verasper  642-3,  654,  659.  660 

Verecundum  642,  651,  674,  682 

Verilus  464 

Vinciguema  28.  29,  185-6,  188,  190,  792,  193,  797 

Vomeridens  563 

Winteria  156,  163 

PKoo£75;a  185,  188-9,  191,  792.  193,  198 

Xaniolepis  50 

Xantichthys  452,  -^56 

Xenaploactis  44 1 

Xeneretmus  408,  -^26.  442 

Xeneniodon  336,  i'^2.  347,  350,  352 

Xenisthmus  590 

Xenistius  473 

Xenocongridae  97 

Xenolepidichthys  393 

Xenornv'^/aA'  94 

Xenopoecilus  345,  352 

Xenopthahmchthys  156,  767,  163-4 

Xi/j/j/as  49,  489,  495,  591,  600-1,  604,  606.  612,  615-7,  619 

Xiphophorus  363 

Xyrichthys  544,  545 

A>r/a5  103 

A'^'^rreun'^  642,  646.  651,  674,  682 

Yarella  185-6,  188-9,  191,  792.  193 
Yirrkala  103 
Yonogobius  584-5 
Yo~ia  402 

Zanchlorynchus  440 

Zanclus  548,  550 

Zaniolepis  410,  '^7'^.  422,  443-5 

Zaprora  57 1 

Zebrasoma  548 

Zebrias  644,  657 

Zenarchopterus  337-8,  352-3 

Zenopsis  393 

Zesticelus  443 

Zeugopterus  640,  643,  646.  65 1 

Ze«5  398 

Zoarces  578-9 

Zm  368-9,  i77.  372,  i7i.  375,  378-9 


760 


ONTOGENY  AND  SYSTEMATICS  OF  FISHES-AHLSTROM  SYMPOSIUM 


Photo  of  symposium  attendees.  La  Jolla,  California,  August  17,  1983.